Two twin series of carbon replicas were synthesized by the acid-catalyzed precipitation polycondensation of various amounts of furfuryl alcohol in SBA-15 suspensions using water and toluene as reaction media. The textural and structural parameters, as well as the morphology of the polymer/silica carbonizates and corresponding replicas, were investigated comprehensively.
Trang 1Available online 1 November 2021
1387-1811/© 2021 The Authors Published by Elsevier Inc This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
Effect of solvent polarity in formation of perfectly ordered CMK-3 and
CMK-5 carbon replicas by precipitation polycondensation of
furfuryl alcohol
aAGH University of Science and Technology, Faculty of Energy and Fuels, Al A Mickiewicza 30, 30-059, Krak´ow, Poland
bAGH University of Science and Technology, AGH Centre of Energy, Ul Czarnowiejska 36, 30-054, Krak´ow, Poland
cJagiellonian University, Faculty of Chemistry, Ul Gronostajowa 2, 30-387, Krak´ow, Poland
A R T I C L E I N F O
Keywords:
CMK-3
CMK-5
Carbon replica
Poly(furfuryl alcohol)
Nanocasting
SBA-15
A B S T R A C T Two twin series of carbon replicas were synthesized by the acid-catalyzed precipitation polycondensation of various amounts of furfuryl alcohol in SBA-15 suspensions using water and toluene as reaction media The textural and structural parameters, as well as the morphology of the polymer/silica carbonizates and corre-sponding replicas, were investigated comprehensively It was found that the polarity of the reaction medium plays an essential role in the scenario of the deposition of poly(furfuryl alcohol) (PFA) onto the surface of the silica matrix Namely, the water-based environment results in propagating PFA chains radially from the pore centres to the wall thereof, while in the case of toluene its growth progresses in the reverse direction The spectroscopic studies disclosed that this is due to the competitive adsorption of monomer and solvent on the superficial silica silanol groups In the case of the water-furfuryl alcohol system, H2O is adsorbed preferentially,
hindering the formation of a homogenous polymer layer, thus precluding the formation of a hollow-type replica
Contrarily, for the toluene-furfuryl alcohol mixture, the monomer adsorption is favored Furthermore, the forming polymer anchors to the silica surface covalently and clads it evenly, therefore facilitating the formation
of a high-quality CMK-5 structure
1 Introduction
Ordered Mesoporous Carbons (OMCs), also called carbon replicas,
pose a class of nanoporous materials offering unique structural and
surface beneficial properties They show such remarkable properties as a
long-range mesoscopic ordering, excellent homogeneity of pore shape
and size, highly developed specific surface area (up to ca 2500 m2 g− 1),
most desirable feature of OMCs is the opportunity of precise control of
their structure at the synthesis stage and ease of surface modification
[5–11] With this, it is not surprising that in the last two decades carbon
replicas have attracted extensive interest in the scientific community,
especially for these purposes in which a well-defined porosity with a
long-range ordering is required The favorable properties of OMCs
resulted in their successful applications as functional materials in a
va-riety of fields, including catalysis (e.g hydrocarbons dehydrogenation,
esterification and transesterification, oxidative degradation) [10, 12–18], adsorptive hydrogen storage [9], purification (e.g removal of
well as CO2 capturing) [19–23], electrochemistry (as electrical double layer (super)capacitors) [6,24–26], and medical purposes (mainly as intracorporeal drug delivery carriers) [7,8,27–29] Moreover, carbon replicas are excellent model materials for a theoretical study of diffusion
patterns simulation/prediction [34] Another interesting application involves their use in the synthesis of mesoporous inorganic materials (commonly metal oxides) featuring the structure of original silica matrices (secondary replication of carbon structures) [35,36] Attempts were also made to synthesize metal oxides exactly imitating the
replication may be an ingenious tool for investigation of structures of porous materials [1,38–42,47] Recently, we reported on the elucidation
* Corresponding author AGH University of Science and Technology, Faculty of Energy and Fuels, Al A Mickiewicza 30, 30-059, Krak´ow, Poland
E-mail address: rjanus@agh.edu.pl (R Janus)
Contents lists available at ScienceDirect Microporous and Mesoporous Materials
https://doi.org/10.1016/j.micromeso.2021.111542
Received 20 September 2021; Received in revised form 26 October 2021; Accepted 29 October 2021
Trang 2of the mechanism of pseudomorphic transformation (PT) of porous
sil-icas by non-direct investigation of the daughter carbon structures of the
The pioneering synthesis of carbon replicas has been published in
1999 by the group of researchers from the Korea Advanced Institute of
al-lows the preparation of negative carbon structures (inverse replicas) cast
from porous silica materials (matrices) based on a so-called hard
tem-plating strategy It consists in filling the pore system (either partial or
complete) of a mineral matrix with a proper carbon precursor followed
by carbonization of the composite and removal of the inorganic template
by etching with alkali or hydrofluoric acid The first replica, called
CMK-1 (Carbon Mesostructured by KAIST) was synthesized by
impreg-nation of MCM-48 silica with an acidified sucrose solution as a carbon
source [2] Inspired by Ryoo, other researchers put efforts to synthesize
a family of replicas employing silicas with different pore system
ar-rangements and a variety of carbon precursors used in various amounts
The ultimate solids featured symmetry elements identical to the
matrices, although they were exact structural negatives thereof
Furthermore, the partial pore filling of the silica matrix with a carbon
precursor may lead (but needs not) to the formation of open-work
hol-low-type frameworks, whereas total filling results in obtaining rod-type
replicas The materials were marked with the common acronym CMK-n,
where n = 1–9 and differs depending on the matrix used and refers to the
The efficiency of carbon precursor deposition in the pores of silica
plays a crucial role in the quality of the resulting final replica (i.e the
fidelity of matrix structure replication) Besides the aforementioned
impregnation, early methods of carbon precursor incorporation
included chemical vapor deposition (CVD) However, this approach
requires the use of an advanced apparatus and is time- and energy-
consuming Moreover, prior to the deposition of carbon precursor, the
matrix surface has to be properly modified (generation of active centres
such a sophisticated synthesis path precludes the possibility of utilizing
carbon replicas on a technical scale Therefore, the reported application
tests, although gave very promising results, did not pass beyond the
laboratory scale, and attempts to synthesize high-quality hollow-type
structures (in fact, more challenging than the rod-type ones) have been
scarcely reported [3,18,47,48]
In our former study, we put efforts to develop a simplified route for
consisted on employing the precipitation polymerization of carbon
precursor’s monomer onto the silica matrix walls in liquid media Based
on this strategy, we successfully synthesized the CMK-3 replica by
nanocasting of the SBA-15 silica by the acid-catalyzed polycondensation
of furfuryl alcohol in an aqueous suspension of the rigid template This
allowed to shorten the synthesis time while using a green reaction
me-dium Moreover, we managed to eliminate the step of preliminary
modification of silica
These encouraging findings gave rise to undertaking attempts to
employ the same procedure to obtain the corresponding hollow-type
CMK-5 replica Unexpectedly, the intended material has finally not been
obtained, although another interesting structure with bimodal
meso-porosity was created (the so-called pseudo-CMK-3) [15,50] It was found
that the chemical nature of the medium used for the decoration of silica
with a polymer governs the manner of the carbon source deposition (the
homogeneous coating of the silica walls with polymer performed in the
water environment is not feasible) Furthermore, we hypothesized that
the solvent’s polarity and its possible interaction with the superficial
SBA-15 silanols may affect the mechanism of PFA deposition (e.g due to
the competitive solvent-monomer adsorption hindering the
homoge-neous distribution of carbon precursor) This may influence (either
deteriorate or improve) the structural quality of the ultimate OMC
Noteworthy, since our first report on PFA deposition in an aqueous
medium [12], there is a lack of research on the use of other media in the literature We have found this issue worth investigating as it is plausible that the deposition of the carbon precursor in liquid media is more ho-mogeneous than that of impregnation, being the most common pro-cedure In fact, the impregnation may be influenced by the local fluctuations in the monomer concentration caused by the evaporation of the solvent Contrarily, the polymer precipitation in liquid media is a self-regulating process driven by the affinity of the monomer to the silica’s surface
In this work we elucidate the role of the polarity of the medium used for the precipitation of poly(furfuryl alcohol) onto SBA-15 silica matrix walls on the mechanism of its deposition This was feasible by the
and low-angle XRD, respectively), morphology (TEM), and spectro-scopic study (FT-IR and XPS), which were carried out for two twin series
of replicas synthesized in water and toluene It was found that using polar solvent results in propagating polymer chains radially from the bulk monomer solution to the silica pore wall, while in the case of nonpolar medium their growth progresses in the reverse direction As a
result, the polar medium precludes the formation of a hollow-type
replica, whereas the nonpolar solvent facilitates the formation of an excellent CMK-5 structure This finding may be a cornerstone to the development of a simple and versatile method for the synthesis of other carbon replicas
2 Experimental section
2.1 Synthesis
All chemicals were commercially available and used without further purification Tetraethyl orthosilicate (TEOS, 98.0%) was purchased
from Acros Organics, poly(ethylene oxide)-block-poly(propylene oxide)-
block-poly(ethylene oxide) triblock copolymer (Pluronic P123), furfuryl
alcohol (FA, 98%), hydrofluoric acid (40–45%), potassium bromide (≥99.0%), and isopropanol (≥99.5%) were supplied by Sigma-Aldrich, whereas hydrochloric acid (35–38%, pure p.a.), tartaric acid (TA, pure p.a.), toluene (pure p.a.), and sodium sulphate anhydrous (99.0%) were purchased from Avantor Performance Materials Poland
2.1.1 SBA-15
SBA-15 silica matrix was synthesized under acidic conditions at a molar gel composition of 1.00 TEOS: 0.02 Pluronic P123: 2.94 HCl:
to 4 ◦C) was slowly instilled (2 drops s− 1) and hydrolyzed at 35 ◦C for 22
h in a mixture containing 40.00 g of Pluronic P123 dissolved before in
round-bottom flask placed in a silicone oil bath and equipped with a reflux condenser Subsequently, the milky reaction mixture was trans-ferred to a laboratory dryer and kept statically at 100 ◦C for 72 h
(pre-cipitate aging step) Then, the white product was recovered by filtration,
Finally, the structure-directing agent was removed by calcination of the silica/P123 composite in a muffle furnace under an air atmosphere at
550 ◦C for 10 h at a heating rate of β = 1 ◦C min− 1 The ultimate material was marked as SBA-15 A small portion of as-made SBA-15 was calcined using the identical thermal regime as for carbonization (850 ◦C for 4 h, β
2.1.2 Carbon replicas
Two twin series of carbon replicas were cast from SBA-15 by the acid- catalyzed precipitation polycondensation of various amounts of FA in suspensions of the matrix, according to the modified procedure reported
in our former works [12,15,49] The series differed in the liquid media used for the incorporation of PFA into SBA-15 pores These media were selected in such a way to be significantly different in polarity and to be
Trang 3miscible with the monomer For this purpose, water (dipole moment μ =
chosen The use of tartaric acid as a catalyst with lower acid strength
than in our earlier reports (hydrochloric acid) (for HCl pKa = − 6.3, while
for TA pKa1 =2.98, and pKa2 =4.34; each value given for water solution)
enabled the slower deposition of the polymer in the pores of SBA-15
This prevented clogging the pores by the rapid growth of the polymer
plugs near the pore entrances
In both series, the same intended monomer/silica mass ratios of 0.50,
1.10, 1.40, 1.70, 2.00, and 2.60 were adjusted using proper masses of
FA TA was used as the polyreaction catalyst at the constant molar ratio
of TA/FA = 0.50 The cumulative mass of the solvent together with the
monomer was kept constant at 50.00 g for each synthesis batch In the
case of the T-series, additionally, anhydrous sodium sulphate was added
as a desiccant agent at the constant molar ratio of Na2SO4/FA = 0.15 to
provide an anhydrous reaction environment It traps the traces of water
originating from toluene and monomer impurities as well as this one
released in the FA polycondensation reaction Briefly, an amount of
1.50 g of SBA-15 held before at 200 ◦C overnight was added under
vigorous stirring (800 rpm) to a mixture of FA, solvent (water or
toluene), TA, and Na2SO4 (solely in the case of the T-series) The mixture
oil bath placed on a magnetic stirrer and equipped with a reflux
condenser It was then agitated at room temperature for 30 min, and
next a heating was turned on After the temperature of the reaction
24 h under vigorous stirring (800 rpm) The resulting brownish
com-posite of poly(furfuryl alcohol) (PFA) and SBA-15 (PFA/SBA-15) was
then isolated, washed with distilled water or toluene (depending on the
reaction medium, respectively), and dried at 90 ◦C overnight
After-wards, to remove the TA and Na2SO4 (undissolved in the original
organic medium), the T-series materials were additionally washed with
90 ◦C This step prevented the damage of the carbonizate structure
caused by its high-temperature oxidation with sodium sulphate during
carbonization as follows: Na2SO4+ 4C →T Na2S + 4CO↑ The as-
synthesized composites were labelled as PFA/S-x_y, where x stands for
the real PFA/SBA-15 mass ratio (determined based on TG measurements
under an air atmosphere), and y refers to the series (y ≡ W and T for
water and toluene medium, respectively) Additionally, two samples of
bulky PFA were synthesized without using the silica matrix in water and
toluene following the same protocol as for the composites These
ma-terials were labelled as PFA_W and PFA_T, respectively The PFA/S-x_y
composites were carbonized in a tubular quartz furnace under an argon
atmosphere (40 cm3 min− 1) at 850 ◦C for 4 h using a heating rate of β =
with HF at room temperature for 90 min Namely, 1.00 g of carbonizate
again The carbonizates and corresponding carbon replicas were marked
as C/S-x_y and C-x_y, respectively
2.2 Characterization methods
Textural parameters of materials were investigated by means of low-
temperature adsorption-desorption of nitrogen (− 195.8 ◦C) The
iso-therms were collected using an ASAP 2020 sorptometer (Micromeritics)
under vacuum The specific surface areas (SBET) were calculated
while the micropore surfaces (Sμ) were assessed based on the t-plot
model (using the de Boer equation) at the same relative pressure range
The external surface areas of SBA-15 and carbonizates (Sex) were
according to the single-point approach (s-p) from the adsorption
respectively For the SBA-15 matrix and carbonizates, the foregoing parameters were assessed with respect to the macroporous silica LiChrospher Si-1000 (SBET =25 m2 g− 1) [51], while for the ultimate
curves (PSDs) In the case of SBA-15, the PSD was calculated using the non-local density functional theory model (NLDFT; adsorption branch; cylindrical pores assumption; software ASIQwin™ ver 1.11, Quan-tachrome Instruments), while for carbonizates and carbon replicas the two-dimensional non-local density functional theory model devised for carbons possessing heterogeneous surfaces was applied (2D-NLDFT; SAIEUS software, ver 3.0) [53,54]
Structural parameters were investigated by low-angle X-ray powder diffraction (XRD) using a Bruker D2 Phaser instrument equipped with a
step of 0.02◦ Transmission electron microscopy (TEM) imaging was performed on
an FEI Tecnai TF20 X-TWIN (FEG) microscope operated at an acceler-ating voltage of 200 kV Before measurements, samples were dispersed
in isopropanol followed by sonication for 10 min and deposited onto carbon-coated copper TEM grids by the drop-casting technique Mid-infrared spectra (300 scans each) were collected in the spectral range of 650–4000 cm− 1 at a resolution of 4 cm− 1 using a Nicolet iS5 (Thermo Scientific) FT-IR spectrometer equipped with a DLaTGS de-tector A diffuse reflectance (DRIFT) device (EasiDiff™-Pike Technolo-gies) and attenuated total reflectance kit (iD7 ATR Accessory, Thermo Scientific) for solid and liquid samples analyses were used, respectively Prior to the measurements, the solid materials, held before at 105 ◦C for
72 h, were diluted with spectral grade dry KBr to 2 wt% and gently milled in an agate mortar, while the ATR spectra for the liquid samples were acquired without dilution
Average values of the ζ-potential (ZP) of SBA-15 immersed in pure reaction media (water and toluene) and respective FA solutions, were determined by using a Zetasizer Nano ZS instrument equipped with a maximum 4 mW He–Ne laser, emitting at 633 nm (Malvern Instruments Ltd., Malvern, U.K.) The measurements were carried out using a Uni-versal dip cell (ZEN1002) combined with a glass cuvette (PCS1115) Prior to the measurements, four suspensions containing 0.1 wt% of freshly calcined SBA-15 were prepared using distilled water, toluene, and corresponding 7.8 wt% solutions of FA The suspensions were son-icated in an ultrasonic bath for 15 min The analyses were performed at
was allowed to equilibrate in the instrument chamber for 2 min Each analysis was repeated three times
X-ray photoelectron spectroscopy (XPS) measurements were per-formed on a Prevac photoelectron spectrometer equipped with a
eV) as an X-ray radiation source at a constant pass energy of 100 eV for survey and high-resolution modes The powder composites were placed
on a sample holder and introduced by a load lock into an analytical
was calibrated using the Si 2p line of pristine SBA-15 silica at 103.6 eV
The surface composition was analysed on the base of the areas and
binding energies of Si 2p, C 1s, and O 1s core levels The spectra were
fitted using CasaXPS software version 2.3.23
An adsorptive interaction of the silica surface with FA in an aqueous medium was investigated employing total organic carbon (TOC) anal-ysis using a Shimadzu TOC-VCPH apparatus Briefly, 1.0000 g of freshly
of a 1.00 wt% FA-water mixture in a 100 cm3 round-bottom flask
Trang 4equipped with a magnetic stirrer Then, the suspension was vigorously
stirred (1000 rpm) for 30 min After separation of silica by filtration on a
Büchner funnel, the filtrate was subjected to the TOC analysis The
capability of silica towards monomer adsorption was estimated based on
a drop in the FA concentration during silica immersion compared to the
mother liquor
High-resolution thermogravimetric measurements (TG) were carried
out using a SDT Q600 analyzer (TA Instruments) An amount of ca 20
20 ◦C min− 1) at an air atmosphere (100 cm3 min− 1) The true amounts of
the carbon precursor incorporated into the silica matrices (i.e the real
polymer/silica mass ratios in the PFA/SBA-15 composites) were
calcu-lated based on the mass loss recalcu-lated to the burning-off of the polymeric
component regarding to the mass of the silica residue The silica’s pore
filling degree was computed as a ratio of PFA volume (density of bulky
PFA at room temperature, ρPFA =1.55 g cm− 3 [50]) with respect to the
mea-surement procedure was employed for the study on the
thermo-oxidative stability of the ultimate carbon replicas
Kinematic viscosity of the binary mixtures of FA with water and
toluene was determined using a suspended-level (Ubbelohde)
viscom-eter The measurements were carried out at 21 ◦C for the mixtures
containing 7.80 wt% of the monomer This concentration corresponds to
the mixtures used in the syntheses of the highest loaded composites
3 Results and discussion
3.1 Effectiveness of PFA incorporation into SBA-15 mesochannels
The efficiency of deposition of PFA inside the SBA-15 mesopore
system was investigated by thermogravimetric measurements
per-formed under the oxidative atmosphere (i.e air) The calculated yield of
polymerization as well as the true PFA/SBA-15 mass ratios and pore
filling degrees are presented in Fig 1, while the recorded TG mass
changes together with DTG and DTA curves are displayed in
Supple-mentary information section (Fig S1)
The FA polymerization effectiveness is evidently influenced by the
polarity of the reaction medium and the FA/SBA-15 ratio used Namely,
in the case of the toluene series, the yield of polymerization is roughly two-threefold higher compared to that of the W-series (excepting the materials with the highest polymer content) This is reflected in a similar trend observed for the degree of filling of the pores, which for the W- series varies between ca 10 and 73%, while for the T-series it spans in the range 28–68% It is pertinent to mention that the lower FA content, the higher polymerization efficiency is observed, notwithstanding the reaction medium (excepting the PFA/S-1.30_W composite; we reported
on a similar effect in our previous works [50,70]) For the composites with the lowest PFA loading, it attained 34 and 99% for the W- and T-series, respectively This means that the use of toluene as the reaction medium facilitates definitely the successful incorporation of PFA to the channels of the silica matrix It is reasonable to conjecture that the effectiveness of silica decoration with PFA is governed by the following circumstances: (i) behavior of silica itself under harsh hydrothermal conditions of PFA deposition (SBA-15 undergoes partial leaching fol-lowed by re-precipitation of silica resulting in the flattening of the inner surface corrugations [55]), (ii) state of the SBA-15 surface silanols in an aqueous and anhydrous environment and their likely role in the monomer pre-adsorption, (iii) mutual interaction between solvent and monomer molecules, (iv) catalyst acidic strength in these media, and (v) viscosity of the FA-solvent mixtures, which may play a crucial role in the kinetic of infiltration of the matrix pore system with carbon precursor [56] Surprisingly, this parameter was not discussed in the literature as far Herein, we have measured the viscosity of the studied synthesis systems The kinematic viscosity of the 7.80 wt% FA-water mixture at
21 ◦C is equal to 1.14 mm2 s− 1, whereas for the FA-toluene mixture of the same concentration is 0.72 mm2 s− 1 (for pure FA it equals 4.73 mm2
s− 1) Thus, this may be a hint unraveling the higher PFA loading within the T-series
Interesting insights are provided by the analysis of the TG, DTG, and
the polymer content, the materials feature similar decomposition pro-files with two distinctive stages (Fig S1, DTG profile) with the maxima
curves recorded for the composites of the W-series, the high-temperature maximum dominates, while the opposite situation is observed for the T- series This suggests a slightly higher thermal stability of the W-series materials
Furthermore, the differences in the manner of PFA deposition find reflection in the macroscopic images of the materials Namely, one can clearly see also the differences in the colors of the as-made composites (Fig S2) The brighter tints of the W-series materials may be seen at a glance even when the polymer content is higher than that one of the T- series On one hand, this is indicative of a higher level of T-series PFA crosslinking On the other hand, this suggests another mechanism of polymer chain growth favoring the formation of chromophoric species (conjugated π-bond systems) in the T-series [57–60] It should be noted that this in turn may influence the carbonization of the polymer and the structural ordering of the final carbon material
3.2 Textural and structural characteristics of C/S-x_y carbonizates and C-x_y replicas
Textural and structural parameters of the parent silica matrix SBA-15
as well as the carbonizates and corresponding carbon replicas were investigated by low-temperature adsorption of nitrogen and low-angle X-ray diffraction The collected isotherms together with the corre-sponding PSDs are depicted in Fig 2, while Fig 3 displays the relevant XRD patterns The respective textural and structural parameters are gathered in Table 1 For better readability, all results are presented along with the ascending real polymer loading
The N2 adsorption isotherm for pristine SBA-15 is a textbook example of a IV(a) type with H1 hysteresis loop featuring steep parallel adsorption and desorption branches (cf Fig 2) [61] This evidences the
Fig 1 Efficiency of FA polymerization and effectiveness of PFA deposition in
the pore system of SBA-15 expressed as true PFA/SBA-15 mass ratio and pore
filling degree (the shaded areas refer to the real PFA contents required for
obtaining the respective replicas)
Trang 5presence of open-ended main mesopores uniform in diameter and
long-range ordering of the architecture thereof These mesochannels are
accompanied by a minor fraction of micropores which act as
inter-connecting channels The textural and structural parameters of the silica
matrix are coherent with the typical values reported for SBA-15 in previously published papers [4,12,13,15,35,43,44,50,55] A very similar isotherm was recorded for the SBA-15@850 material Indeed,
Fig 2 Nitrogen adsorption-desorption isotherms (A, B) and respective PSDs (A′, B′) for C/S-x_y carbonizates (red lines and symbols) and corresponding C-x_y
replicas (black lines and symbols) of W-series (A, A′): x = 0.17 (a), 0.28 (b), 0.35 (c), 0.39 (d), 0.43 (e), 1.30 (f), and T-series (B, B′): x = 0.49 (a), 0.73 (b), 0.94 (c),
1.02 (d), 1.10 (e), 1.22 (f) For clarity, the PSDs were offset of 0.25 (A′), and 0.60 cm3 g− 1 nm− 1 (B′) each (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Trang 6in narrowing mesopores by 0.8 nm and extinction of microporosity This
in turn caused a drop in both SBET and Vt (cf Table 1) [15,50]
In the case of the carbonizates of the W-series, except the material
loaded with the highest amount of PFA (viz Fig 2A–a-e), the deposition
of PFA followed by carbonization did not influence the nature of the
isotherm The only differences are a slight shift of the hysteresis loop
towards lower relative pressures caused by the thermal shrinkage of the
SBA-15 structure during carbonization of the PFA/SBA-15, and a
gradual decrease in both specific surface area and total pore volume
with increasing PFA content (cf Fig 2A, Table 1) This is not surprising
in view of the progressive filling of the silica’s pore system with carbon
Interestingly, notwithstanding the pore filling degree, the pore size of
Table 1) The N2 adsorption isotherm for the carbonizate containing the
0.4 This points to the effect of cavitation of the adsorptive in partially
the shift in the main mesopore size to ca 5.2 nm (Fig 2A’–f, Table 1)
Expectedly, the accumulation of carbonaceous material entailed a
gradual decrease in the SBET and Vt, while increasing in the micropore
volume This is a cumulative effect of the development of inherent
microporosity in the carbonized PFA as well as the formation of
slit-shaped micropores between the carbon material and silica wall due
to their uncapping caused by discrepancies in the shrinkage effect dur-ing carbonization [13–15,50]
The two W-series samples with the lowest PFA contents (viz C-
C-0.28_W material reveals additionally a narrow H4 hysteresis loop typical of suchlike mixed-porosity solids Indeed, both materials feature relatively low total pore volumes of 0.23 and 0.29 cm3 g− 1 with 56 and
featureless XRD patterns for these carbons disclose the entirely disor-dered structures thereof (Fig 3A–a,b)
Other N2 adsorption isotherms of the W-series carbons may be classified as IV(a) type with H2(b) hysteresis loops For these samples, the adsorption branches show the presence of two inflections in the mesopore region (this is best seen in the case of the C-1.30_W replica), which confirm gradual development of two individual mesopore
Interestingly, simultaneous extinction of the microporosity is observed
primary mesopores originating from the removal of silica matrix walls centered at 3.1 nm are accompanied by far broader ones at ca 4.5–15.0
nm resulting from the coalescence of the adjacent pores of SBA-15,
Fig 3 Low-angle XRD patterns for C/S-x_y carbonizates (red lines) and corresponding C-x_y replicas (black lines) of W-series (A): x = 0.17 (a), 0.28 (b), 0.35 (c),
0.39 (d), 0.43 (e), 1.30 (f), and T-series (B): x = 0.49 (a), 0.73 (b), 0.94 (c), 1.02 (d), 1.10 (e), 1.22 (f) Reflections assignment: * ≡ (1 0 0), ^ ≡ (1 1 0), “ ≡ (2 0 0), # ≡
(2 1 0), $ ≡ (3 0 0) (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Trang 7which were previously either entirely empty or partially filled with the
carbon precursor The width of this peak should not be surprising given
the random distribution of PFA inside the SBA-15 pore system, which
yields the pseudo-CMK-3 structures [15,50] Due to the defective
struc-ture, this material exhibits a slightly lower specific surface area than
g− 1 is understandable Noteworthy, despite the non-ideality of these
frameworks, their XRD patterns gradually take shape of the pattern of
standard CMK-3 material along with increasing PFA loading, achieving
the maximum similarity for the highest PFA content (Fig 3B–c-f)
Recently, we reported on the formation of similar structures when
SBA-15 with a low degree of silica framework condensation was
employed as a hard template (therein, the silica matrix was detemplated
deposition of PFA was carried out in water medium [15,50] However, it
is pertinent to mention that using hydrochloric acid as a polyreaction
catalyst leads to the formation of the typical CMK-3 structure [12,49]
Another scenario was observed for the T-series carbonizates The two
carbonizates with the lowest PFA loadings (i.e C/S-0.49_T and C/S-
hys-teresis loops (Fig 2B–a,b) [61] It is worth noting that these loops are
shifted to lower relative pressures compared to both SBA-15 and
SBA-15@850, which suggests a progressive cladding of the inner walls
of pores with the polymer Indeed, considering the corresponding PSDs
(Fig 2B’–a,b), a gradual decrease in the diameter of the main mesopores
with increasing PFA content is evident The materials with moderate
PFA loadings (i.e C/S-0.94_T, C/S-1.02_T and C/S-1.10_T) feature the
isotherm of IV(a) type with a H2(a) hysteresis loop For these samples, the main pore size equals ca 4.0–4.1 nm regardless of the real content of
PFA loading (C/S-1.22_T) shows the maximum nitrogen uptake close to
clearly evidences its total pore filling with the polymer The XRD pat-terns collected for the T-series carbonizates show lower intensity of the characteristic reflections compared to the parent silica, which proves filling the pores of the hard template with organic material (Fig 3B) The analysis of the behavior of the N2 adsorption isotherms recorded for the final carbon replicas of the T-series provides particularly inter-esting conclusions The replicas derived from the two materials with the lowest PFA content disclose a micro-mesoporous character thereof (isotherm of type I(b) with a H4 loop), similar to the corresponding W- series replicas (see Fig 2A–a,b, vs Fig 2B–a,b) [61] The lack of a long-range ordering of the architecture of these materials is visible in the
when considering the OMCs synthesized from the carbonizates of moderate PFA loadings (real polymer/silica ratio of 0.94–1.10, viz Fig 2B–c-e) Namely, the isotherms are of type IV(a) with H1 hysteresis loop and well-distinguished two inflections in the adsorption branch at
ca p/p 0 =0.30–0.50, and 0.55–0.70 Apparently, this is reflected in the respective PSDs displayed in Fig 2B’–c-e, which show the bimodal mesoporosity of these materials featuring two maxima centered at 2.9
walls, while the broader ones are inherited from the carbonizate, in
Table 1
Textural and structural parameters of parent SBA-15, SBA-15@850, C/S-x_y carbonizates, and corresponding carbon replicas
Sample SBET (Sex) a [m 2 g − 1 ] S μ a [m 2 g − 1 ] Vt c [cm 3 g − 1 ] V μ a [cm 3 g − 1 ] Vme [cm 3 g − 1 ] Dp [nm] Dw [nm] a 0[nm]
aαs model
b Vme=Vt (s− p)− Vμ(αs)
cSingle-point at p/p 0 =0.98
dNLDFT for silicas; adsorption branch; cylindrical pores assumed
e2D-NLDFT for carbons with heterogeneous surfaces
fSilica wall thickness; Dw,sil. =a0− Dp
gCarbon nanorod diameter; Dw,carb. =c⋅d1 0 0
(
ρ carb.− 1+Vμ
Vme+ρ carb.− 1+Vμ
)1/
2
, c – constant; for cylindrical pores c = 1.213; d1 0 0 – interplanar spacing; d1 0 0 = 2⋅ d2 0 0;
ρcarb – amorphous carbon density; ρcarb =2.05 g cm− 3 [15,43,46]
hAverage thickness of carbon wall in the tube-type replicas; w C=(Dsil.850
p − Din
p)
2 , Dsil.850p is the mesopore diameter of SBA-15@850; Din
p means the inner diameter of
carbon tube
i Due to the featureless XRD patterns in the (1 0 0) reflection region (Fig 3), the lattice parameters were calculated from (2 0 0) reflection; a0 = 4⋅ 3−1/2⋅ d2 0 0
Trang 8which the inner silica walls were covered with a PFA film (intra-tubular
carbon pores) Interestingly, the share of the latter one in the mesopore
volume decreases with increasing content of carbon precursor (see
Fig 2B’–c-e, decreasing the maxima at 4.0–4.1 nm), while the pore
diameter stays constant Indeed, the bimodal porosity contributes to the
g− 1 (cf Table 1) It should be underscored that such textbook examples
of CMK-5 isotherms and PSDs were rarely reported in the literature The
reflections including the dominating (1 1 0) one are indicative of the
excellent quality of the synthesized materials is evident It may be
surprising that the T-series sample with the highest PFA loading yielded
a high-quality rod-type CMK-3 replica, notwithstanding its pore filling
degree barely equals 68% (see Figs 1, Fig 2B–f, Fig 2B’–f) However, given the thermal shrinkage of the SBA-15 structure during composite carbonization, this is understandable (such shrinkage causes a reduction
in Vt roughly by ¼, cf Table 1) [15,50] This carbon replica features monomodal mesopores of 3.7 nm in diameter, a total pore volume of 1.17 cm3 g− 1, and SBET of 1222 m2 g− 1 Such textural parameters are in accordance with previous reports on CMK-3 materials [3,4,6,9,10,12,
13,15,35,43,49,50] The structural ordering is manifested in the XRD pattern with three distinguished reflections, also typical of suchlike structures (Fig 3B–f) [34]
Fig 4 TEM images and Fourier diffractograms of carbon replicas: C-0.43_W (A), C-1.30_W (A′), C-1.10_T (B), and C-1.22_T (B′)
Trang 9The phase purity of the chosen carbon replicas was studied by high-
for the T-series evidence the higher homogeneity (i.e lack of impurities
being disordered carbonaceous material, which could be formed onto
the external surfaces of silica matrix) of these materials compared to the
W-series Interestingly, the later ones exhibit the same temperature of a
in the materials In contrast, the CMK-5 sample shows the maximum
combustion rate at the temperature of ca 15 ◦C lower This is justified by
the open-work structure of this material
3.3 Morphology of carbon replicas
The structural ordering and morphology of the carbon replicas were
investigated by TEM imaging The micrographs taken for the chosen
materials of both series together with relevant Fourier diffractograms
are displayed in Fig 4
The images recorded for the carbon material based on the partially
dis-closes vestigial hexagonal architecture features This is coherent with
the textural parameters (see Fig 2A–e, Fig 2A’–e) The higher degree of
filling of the PFA matrix results in obtaining the pseudo-CMK-3 replica
[15,50] As mentioned above, in this case, the carbon precursor fills the
honeycomb pore system of SBA-15 randomly, i.e some channels remain
empty, while others are partially or completely filled with PFA This may
car-bon nanorods and cavities formed from empty pores, respectively
Noteworthy, despite these structural discontinuities, such material is
mesoscopically well ordered (cf Fig 2A–f, Fig 2A’–f, Fig 3A–f)
More interestingly, the carbon structure derived from the partially
filled composite of the T-series displays a fabulous TEM image taken
along the [1 0 0] direction (Fig 4B) This is typical of a high-quality
hollow-type CMK-5 replica with well-distinguished bimodal
mesoporosity and an excellent hexagonal arrangement (p6mm space group) This is in line with the N2 adsorption isotherms and XRD pattern (Fig 2B–e, Fig 2B’–e, Fig 3B–e) As expected, the higher loading of SBA-15 with PFA achieved in toluene results in the formation of a
reg-ular rod-type CMK-3 replica with a perfect hexagonal mesoscopic
ar-chitecture (Figs 4B’, 2B–f, 2B’–f and 3B–f) It should be emphasized that the analysis of the dozen TEM images of both materials from the T-series (not shown here) disclosed a lack of the effect of the formation of an external amorphous shell of the excessing PFA enveloping the PFA/silica composite particles Such a phenomenon was observed in the case of the highest-loaded materials synthesized in water as was reported in our previous works [12,15] Naturally, this influences positively the quality
of the carbon replicas synthesized in toluene in terms of both structural ordering and textural parameters
3.4 Mechanism of PFA deposition: a spectroscopic study
The substantial differences in the textural parameters of the W- and T-series OMCs were a premise suggesting different mechanisms of deposition of the carbon precursor depending on the polarity of the reaction medium This inspired us to deepen the study on the in-teractions of monomer and polymer with the silica surface We put ef-forts to unravel these issues by the investigation of FA-silica interactions (FT-IR) and analysis of the non-carbonized composites (FT-IR and XPS)
3.4.1 Monomer pre-adsorption
The adsorptive interactions of silica surface with monomer and both reaction solvents were studied by means of DRIFT spectroscopy For this
of 5 wt% solutions of FA in water and toluene, respectively, at room temperature for 3 h Additionally, to distinguish the silica-solvent in-teractions, the SBA-15 matrix was contacted in the same manner with the pure solvents After the contact, the materials were separated without washing, dried at room temperature overnight, and then evac-uated under static conditions at 50, 100, and 150 ◦C for 1 h The
Fig 5 DRIFT spectra collected for the SBA-15 after pre-adsorption of FA from water (A), and toluene (B) solutions (red lines) and after contact with pure solvents
(black) followed by desorption at room temperature overnight (a), and at 50 ◦C (b), 100 ◦C (c), 150 ◦C (d) for 1 h The ATR spectra of pristine SBA-15, pure FA and respective solvents are added in the bottom (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Trang 10collected DRIFT spectra are gathered in Fig 5
The contact of the SBA-15 template with both pure water and the FA-
water mixture resulted in rehydration of the silica surface This is
manifested by a pronounced drop in the intensity of free silanols
vibrations of hydrogen-bonded silanols, and shift of the stretching Si–O
mode from 979 to 962 cm− 1 (stretching Si–OH) (Fig 5A–a-d) [50,63] In
the case of the silica immersed in FA-water mixture, the spectra reveal
additionally the features of FA, namely, at 2929 and 2873 cm− 1
bending in furan ring) [64,65] Interestingly, the characteristic band
lower intensity for the FA-water mixture This is the effect of competitive
surface simultaneously attracts both water and FA with the engagement
of free silanols However, the accessibility of adsorption sites for alcohol
molecules is largely hindered by the shield of preferentially adsorbed
water The desorption at elevated temperatures resulted in gradual
extinction of the FA bands, but no surface dehydration was observed
Indeed, it is impossible to reverse the silica hydration process under
these conditions
The interaction between silica and FA before the polycondensation
reaction was proven by TOC analysis of the FA-water mixture after 30
min of contact with freshly calcined SBA-15 It was found that the
of the silica surface Such a negligible monomer adsorption suggests the preferential adsorption of water This is in line with the above FT-IR study as well as the reports published elsewhere [65]
Another scenario was observed when silica was immersed in pure
sample after the contact with pure solvent followed by evacuation at room temperature showed the complete loss of toluene Thus, the state
of the freshly calcined silica surface was restored even for such mild desorption conditions The complete evaporation of the solvent at the temperature of ca 90 ◦C below its boiling point (i.e 110 ◦C) indicates a
low affinity of toluene towards the silica surface Indeed, the phobic
character of silica towards aromatics adsorption is not surprising [66]
In contrast, the contact of SBA-15 with the FA-toluene mixture clearly revealed that the free silanols were involved in the attracting of the alcohol molecules This means that monomer adsorption is favored
when toluene is used With this in mind, the formation of hollow-type
carbon replicas as a result of polycondensation of the FA selectively adsorbed onto the silica surface appears understandable Moreover, the comparison of the intensity of FA bands adsorbed in water and toluene (see Fig 5A–a,d vs Fig 5B–a,d) confirms that the use of the aprotic medium promotes the adsorption of larger amounts of alcohol, which in turn is in line with the higher efficiency of PFA deposition in toluene (cf Fig 1) These findings were also proven by the measurements of the zeta potential of the parent silica immersed in pure reaction media and FA-solvent mixtures In contact with pure solvents, the SBA-15 silica revealed typical ZP values (− 31.3, and − 24.2 mV for water and toluene, respectively) [67,68] In contact with the FA solutions, the surface be-comes depleted in a negative charge; in the case of FA-water, the ZP equaled − 6.7 mV, while for FA-toluene the ZP reached +12.9 mV This
is due to the protonation of the free silanols by the FA molecules as
2 :OR [69]
3.4.2 Surface chemistry of PFA/silica composites
The DRIFT spectra of the as-made W- and T-series PFA/silica com-posites with the highest PFA loading and respective bulk polymers are
XPS spectra are displayed in Fig S4, while the concentrations of particular carbon- and oxygen-containing surface moieties are gathered
in Table S1 The spectrum recorded for the PFA/S-1.30_W composite is essen-tially a simple superposition of the silica and bulk PFA spectra, excepting the 2800–3750 cm− 1 region and the band at 979 cm− 1 (Fig 6a) [70]
the increase in the intensities of 2800–3750 cm− 1 and 962 cm− 1 modes may be assigned to the profound rehydration of the silica surface during the PFA deposition, which is not surprising taking into account its
respec-tively) confirms the lack of chemical anchoring of the monomer mole-cules before the polyreaction (i.e the absence of the Si–O–C bridges that
should be expected in this case) This ultimately proves the shielding role
of water molecules occupying the adsorption sites on the silica surface The spectrum of the PFA/S-1.22_T composite discloses an increase in
absorp-tion at 979 cm− 1 remains unaltered (Fig 6c) This suggests the engagement of isolated silanols in FA anchorage while lacking matrix rehydration, which is not surprising given the anhydrous conditions provided in the reaction system More interestingly, the PFA/S-1.22_T
ring) as well as 1560 and 1600 cm− 1 bands (furan ring vibrations) This
is due to the effect of the acid-catalyzed furan ring-opening leading to the formation of γ-diketone moieties, which is evidenced by the presence
of an intense band at 1715 cm− 1 (stretching vibrations of C––O species)
Fig 6 DRIFT spectra of non-carbonized PFA/SBA-15 composites and bulk PFA
samples: PFA/S-1.30_W (a), PFA_W (b), PFA/S-1.22_T (c), and PFA_T (d) The
green line represents the spectrum of pristine SBA-15 (For interpretation of the
references to color in this figure legend, the reader is referred to the Web
version of this article.)