Effect of the active phase-support interaction on the electronic, thermal and catalytic properties of [H–Pyr]+[HSO4]?/support (support = rice husk ash; corundum)

Journal Pre-proof

Effect of the active phase-support interaction on the electronic, thermal and cata-

+

]^–/Support (Support = rice husk ash; corundum)

lytic properties of [H–Pyr] [HSO

4

1

Ivaylo Tankov , Rumyana Yankova , Anife Veli ,

1

2

Radoslava Nikolova , Zilya Mustafa , Ivanka Stoycheva

¹University “Prof. Dr. Assen Zlatarov” Burgas, 8010, Bulgaria

²Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Scienc-

es, Bulgaria

Abstract

Influence of the surface active phase-support interaction on the electronic, thermal and

+

–

catalytic performance of pyridinium hydrogen sulfate ([H–Pyr] [HSO , PHS] ) ionic liquid

4

immobilized by a wetness impregnation method on rice husk ash (RHA) and corundum (Al-

NA) carriers has been investigated in the current work. For that purpose, both the devel-

oped supports and the corresponding catalysts (PHS/Al-NA and PHS/RHA) were character-

ized in details by means of analytical methods such as N

2

adsorption-desorption

measurements, X-ray diffraction, X-ray photoelectron spectroscopy, spectroscopy in the

ultraviolet and visible regions, infrared spectroscopy and thermogravimetric analysis. To

compare the catalytic activity of PHS/Al-NA and PHS/RHA, a process of butyl acetate syn-

thesis was applied as a test reaction. The physicochemical characterization revealed that

the active phase-support interaction in the case of PHS/RHA catalyst is stronger than that

between PHS phase and Al-NA. Based on this, the catalytic performance of PHS/RHA sam-

ple (expressed by substrate conversion degree) in the reaction of acetic acid esterification

with butanol was found to be more pronounced with respect to that of Al-NA supported

pyridinium hydrogen sulfate.

Keywords: Supported ionic liquids; Physicochemical characterization; Thermal behavior;

Catalytic properties

1

Journal Pre-proof

1. Introduction

Over the past few decades, the ionic liquids (ILs) investigation is a major field in the

modern chemistry due to their unique properties – low vapor pressure, high thermal

stability, tunable polarity and high conductivity [1]. Ionic liquids are molten salts (melt at

or below 100°C), consisting of organic cations associated with organic or inorganic anions

[2]. The anion and cation variety allows changing of the ILs properties according to a

specific application. Besides as a “green” alternative to volatile organic solvents, the ILs

application as homogeneous catalysts for various organic reactions is widely reported

[

3,4]. Tao et al. [3] present acetic acid esterification with C

1

–C alcohols in the presence of

4

+

]^–. In the field of organic synthesis, the alkylation of phenol

the ionic liquid [C

1

Im] [HSO

4

+

–

with tert-butyl alcohol as an alkylating agent is studied in presence of [C

4

C

1

Im] [X] (X = Cl,

BF , PF ) [5]. Maximum phenol conversion of 93% at molar ratio of phenol:TBA:IL of 1:2:1,

4

6

reaction time of 3 h and reaction temperature of 80°C is achieved. However, the ILs

application as homogeneous catalysts have several drawbacks: (i) usage of large amounts

of ILs for high reactants conversion, (ii) mass transfer limitations due to high ILs viscosity,

(

iii) difficult IL catalyst separation from the reaction products [6]. To overcome the

aforementioned disadvantages, the ILs immobilization on solid supports is extensively

studied [7].

Immobilisation of ILs can be carried out by several methods: (i) solid supports

impregnation with IL, (ii) creating covalent bonds between the IL and the support, (iii) a of

IL confinement in the support material or (iv) copolymerisation of the IL with a suitable

monomer. Covalent bonds may be formed between the IL and the support even when the

SILP is obtained by impregnation – in case of chloroaluminate ILs where anions are

attached through the metal centre of the anion to a surface OH group [8]. Non-covalent

−

interactions of the functional groups of anions, such as the SO

surfaces were also detected by IR measurements [9].

3

group of [OTf] , with solid

The supported ionic liquid (SILs) catalysts combine the advantages of the ILs and those

of the solid support. Compared with pure ILs, SILs offer further benefits such as easy

separation, reusability and application in fixed or fluidized bed reactors. It should be

further emphasized that even a very low solubility of the catalyst in the flowing liquid

media will make long term stability of the SILC catalyst system impossible due to slow but

steady leaching. Moreover, almost identical catalytic activity between the SILs and their

2

Journal Pre-proof

homogenous counterparts was shown as well [10]. Different materials have been used as

solid supports for SILs preparation [11]. Brønsted acidic SILs can be prepared by the use of

[

HSO

4

]^–or [H

2

PO

]^–anions or by the introduction of alkane sulfonic acid groups as side

4

chains of the cations. Combination of these two approaches leads to the so called “dual

acidic” ionic liquids. For example, Wang et al. [10] investigate the catalytic activity of

−

]^–, [H

]^–and CF

–

[

C

4

SO

, ZrO

that the [C

3

HC

1

Im]+[X] (X = [HSO

4

2

PO

4

3

COO ) supported on various solids (SiO ,

2

Al

2

O

3

2

, TiO

2

) in esterification of pentaerythritol with caproic acid. The results showed

+

]⁻/SiO

sample possesses the best catalytic activity, expressed

4

SO HC

3

1

Im] [HSO

4

2

by ester yield value of 99%. In another work, alkylation of benzene with 1-dodecene over

+

–

SBA-15 supported [C

4

C

1

Im] [TFSI] is investigated [11]. A promising alternative of the SiO

2

support for SILs catalysts could be rice husk ash [12], a major waste product of the rice

+

]⁻/RHA in 1-(benzothiazolylamino)-

milling industry. Usage of [C

1

(CH

3 3

O) SiC

3

Im)] [HSO

4

phenylmethyl-2-naphthols and pyrimido[1,2-a]benzimidazole synthesis showed product

yields of 97 and 93%, respectively [13].

It is well known that the support nature significantly influences the catalytic behavior of

a heterogeneous catalyst due to an active phase-support interaction [14]. Hence, to predict

the properties of SILs catalysts and choose a more suitable one for a specific application, a

detailed analysis on the IL-support interaction is needed. The degree of ILs immobilization

on the solid support surface and the ILs-support interaction as well can be successfully

studied using different analytical techniques such as specific surface area measurements,

X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), spectroscopy in the ultra-

violet and visible regions (UV-vis), Fourier transform infrared spectroscopy (FT-IR) and

thermogravimetric analysis (TGA).

The aim of the present work is to study the effect of the surface active phase-support

interaction on the catalytic properties of new SILs, namely pyridinium hydrogen sulfate

immobilized on rice husk ash and corundume carriers. For that purpose, a combination of

analytical techniques (N adsorption-desorption measurement, XRD, XPS, UV-vis, FT-IR,

2

TGA/DSC) is applied for a physicochemical characterization of the samples. To compare the

catalytic activity of the obtained SILs, acetic acid esterification with 1-butanl is used as a

test reaction. The results are expected to be useful for future experimental and theoretical

studies on the structure-activity relationship in different types of supported ionic liquids.

2. Experimental section

3

Journal Pre-proof

2

.1. Preparation of pyridinium hydrogen sulfate

Pyridine (2.13 g, 0.0269 mol) and ultra clean water (1.45 g, 0.0807 mol) were charged

into a 0.020 l three-necked flat bottom reactor, equipped with a magnetic stirring, a reflux

condenser and a thermometer. The aqueous solution of pyridine was stirred at ambient

temperature for 0.5h, where after the temperature was reduced to 5°C. Then, H

2

SO (2.64

4

g, 0.0269 mol) was added dropwise over a period of 1.5 h under vigorous stirring. The re-

action mixture was stirred at reflux for an additional period of 5 h at 80°C. For solvent re-

moving, the aqueous solution of pyridinium hydrogen sulfate was distillated at 70°C under

reduced pressure (0.050 MPa) and the product was dried at the same conditions for 8 h

until the residue mass remained constant. After cooling, the colorless solid was washed

tree times with diethyl ether 0.100 l and dried under vacuum (0.050 MPa) at 80°C for 12 h.

Pyridinium hydrogen sulfate is denoted as PHS.

2.2. Preparation of rice husk ash

Rice husk ash, abbreviated as RHA, is obtained by controlled rice husk combustion in

air through a fluidized bed reactor lab-scale installation at a temperature of 320°C and a

heating rate of 10°C/min.

2.3. Synthesis of corundum

Corundum synthesis includes a pre-formation of ammonium alum (NH

4

Al(SO

4

)

2

.12H O)

2

from ammonium and aluminum sulfates in water solution at 70°C. After cooling for 2 h, a

white precipitate was formed. It was filtered and washed in deionized water several times.

The powder produced was calcined at 1000°C for 1 h to obtain α-Al

2

3

O (denoted in the pa-

per as Al-NA)

2.4 Preparation of supported pyridinium hydrogen sulfate

The supported ionic liquid is obtained by wetness impregnation of the synthesized

supports (RHA and Al-NA) (3g) with an aqueous solution (0.0271 mol/l, 66 wt.% PHS with

respect to the support mass) of PHS in a round bottom flask (0.100 l). The mixture is inten-

sively stirred (1000 rpm) at ambient temperature for 30 h. The water is removed in a rota-

4

Journal Pre-proof

ry evaporator at 70°C and reduced pressure (0.010 MPa). The solids (PHS/RHA and

PHS/Al-NA) are dried at 110°C and 0.1 MPa for 12 h.

2

.5. Sample characterization

XRD analysis is performed by a computerized Seifert 3000XRD diffractometer using Cu

Kα (λ = 0.15406 nm) radiation and a PW 2200 Bragg–Brentano θ/2θ goniometer equipped

with a bentgraphite monochromator and an automatic slit. Diffraction peaks are recorded

in a 2θ range of 10–90° with step size of 0.033 and step scan of 10.0 s. Phase identification

is carried out by comparison with database cards. The crystallite size of PHS nanoparticles

is calculated from the full width at half maximum (FWHM) of the diffraction peaks using

Debye–Scherer’s equation (Eq. (1)).

kλ

(1)

D =

βcosθ

where, “D” is the average crystalline dimension perpendicular to the surface of the

specimen, “λ” is the wavelength of used X-ray, “k” is Scherer’s constant (0.9), “β” is the full

width at half maximum intensity of a Bragg reflection excluding instrumental broadening

and “θ” is the Bragg diffraction angle. The values of the β and θ parameters from the XRD

peak are estimated by Gaussian fitting.

A porous textural characterization of the synthesized samples is evaluated from the N

2

adsorption-desorption isotherms, determined at -196°C on an automatic apparatus Surfer

sorption analyser (Thermo Scientific). The surface area of the samples is determined by the

BET method using data from the adsorption isotherms in the range of relative pressures

from 0.05 to 0.28 (for isotherms type IV). The total pore volume, known as volume of

Gurvich, is determined based on the volume of adsorbate V0.98, recorded on the desorption

branch of the adsorption isotherm at a relative pressure (P

i

/P

0

) of 0.98. The micropore

volume is calculated using the Dubinin-Radushkevich equation up to P

i

/P ≤0.15. The pore

0

diameters are obtained by applying the Non-local density functional theory (NLDFT) on N

adsorption data.

2

UV-vis spectra of the samples are recorded in the range of 200–800 nm at room tem-

perature using a Thermo Evolution 300 UV-Vis spectrophotometer equipped with a pray-

ing mantis device. Spectralon is used as a reference material.

5

Journal Pre-proof

–

1

FT-IR spectra are recorded in the frequency region of 4000-400 сm by means of a Ni-

colet iS 50 Thermo Scientific FT-IR spectrophotometer equipped with DTGS KBr detector

–

1

(

4 cm ) at scan's number of 32.

XPS analysis is carried out using ESCALAB MkII (VG Scientific) electron spectrometer at

a base pressure in the analysis chamber of 510^–¹⁰mbar (during the measurement 10^–⁸

mbar), using AlKα X-ray source (excitation energy 1486.6 eV). The pass energy of the hem-

ispherical analyzer was 20 eV, 6 mm slit widths (entrance/exit). The instrumental resolu-

tion measured as the full width at a half maximum (FWHM) of the Ag3d5/2, photoelectron

peak is 1 eV. The energy scale is corrected to the C1s – peak maximum at 285.0 eV for elec-

trostatic charging. The processing of the measured spectra includes a subtraction of X-ray

satellites and Shirley-type background [15]. The peak positions and areas are evaluated by

a symmetrical Gaussian-Lorentzian curve fitting. The accuracy of the BE values was ± 0.2

eV. The relative concentrations of the different chemical species are determined based on

normalization of the peak areas to their photoionization cross-sections, calculated by Sco-

field [16].

Thermal stability of the obtained samples is investigated by means of a thermal gravi-

metric analyzer Netzsch STA 449C. The experiment conditions are as follows: initial sample

mass – 5.0mg, heating rates – 10°C/min, temperature range from 30 up to 700°C, nitrogen

flow – 0.020 l/min, platinum pans. During the heating, variation of the weight loss with re-

spect to time and temperature is collected automatically by the instrument. Differential

scanning calorimetry (DSC) measurements are carried at the same experimental condi-

tions.

2

.6. Catalytic test

In a typical procedure, acetic acid (8.34 g, 0.1388 mol) and catalyst (10 wt.%, with re-

spect to the mass of substrate) are placed in a three-necked round bottom flask of 0.100 l

capacity. The latter is equipped with a condenser, a port for sample withdrawal and a

thermometer. Mixing of the reaction fluids was performed by a magnetic stirrer. Once the

desired reaction temperature (80°C) is reached, an equimolar amount (10.29 g, 0.1388

mol) of 1-butanol, separately heated to the set reaction temperature (80°C) is added into

the reactor. This time is considered as zero reaction time (t = 0). The reaction progress is

monitored by GC analysis. At regular intervals of time (10 min), a sample (1 ml) is with-

drawn and immediately cooled in an ice bath. The reaction mixture is analyzed by GC

6

Journal Pre-proof

7890A (Agilent Technologies) equipped with a flame ionization detector and a capillary

column HP-INNOWAX, 30 m0.32 mm0.25 μm. High purity helium at a flow rate of

.0015 l/min is used as a carrier gas. The oven temperature is programmed at 70°C for 2

0

min and then it is raised from an initial value 70°C to 190°C at a ramp rate of 5°C/min. The

injector and detector temperature are maintained at 250°C and 300°C, respectively. The

analytical uncertainty of GC is less than 4%, which is in accordance with the acceptable lim-

its. The acetic acid conversion is calculated according to Eq. (2). It should be noted that the

occurrence of side reactions such as butylene synthesis due to alcohol dehydration is not

registered.

C

⁻^CA

A₀

(2)

X_A=

C_A₀

where, X is acetic acid conversion, C_A₀is initial concentration (mol/l) of the acetic acid

A

and C corresponds to the acetic acid concentration (mol/l) at a given reaction time (t,

A

min).

3. Results and discussion

3.1. Phase composition and textural characterization

For a description of the crystalline phases present in the obtained supports and cata-

lysts, a XRD analysis is conducted. The XRD patterns of Al-NA, RHA, PHS, PHS/Al-NA and

PHS/RHA are presented in Fig. 1.

The X-ray diffraction analysis of Al-NA showed a number of strong XRD patterns at 2θ

values of 25.50° (d = 3.4965), 35.12° (d = 2.5532), 37.68° (d = 2.3854) and 43.23° (d =

2

.0911) which undoubtedly confirmed the presence of α-Al

296). The aforementioned 2θ values correspond to the (012), (104), (110) and (113) lat-

, respectively [17]. It was established that our results completely coin-

cide with those presented by other authors [18,19]. For example, electric field gradients

and charge density in corundum, α-Al were established by Lewis et al. [19]. XDR analy-

2

O

3

phase (ICDD No. 00–001–

1

tice planes of α-Al

O

2 3

O

2 3

sis of the compound under investigation showed an appearance of strong peaks related to

d-spacing values of 3.4913, 2.5545, 2.3910 and 2.0928. In another work, the thermal con-

ductivity and viscosity of synthesized α-alumina-based nanofluids are investigated [20].

7

Journal Pre-proof

The authors report that the calcination temperature of 1100°C for 2 h results in the for-

mation of a single-phase of corundum powder, being confirmed by XRD patterns at 2θ =

25.71°, 35.18°, 38.90° and 43.44°. Except corundum, patterns (2θ values of 12.40° (d =

7.1324), 16.03 (d = 5.5245), 17.66° (d = 5.0181), 20.45° (d = 4.3394), 21.54 (d = 4.1222),

25.41° (d = 3.5025), 27.13° (d = 3.2842) and 29.07° (d = 3.0693) related to ammonium al-

um (ICDD No. 00–042–1430) were detected in the X-ray diffractogram of Al-NA as well.

These are related to the (100), (102), (012), (011), (101), (110), (111) and (015) lattice

planes of the NH

4

Al(SO

4 2

)

.12H O, respectively [21-24]. This indicates that a part of the

2

ammonium alum has not been converted to corundum during the calcination process.

The X-ray diffraction analysis of the unsupported ionic liquid revealed strong peaks ap-

peared at 2θ = 10.43° (d = 8.4748), 13.45° (d = 6.5779), 14.31° (d = 6.1845), 18.32° (d =

4.8388), 20.96° (d = 4.2349), 22.83° (d = 3.8921), 23.23° (d = 3.8260), 23.96° (d = 3.7110)

̅

and 24.59° (d = 3.6174), which correspond to the (010), (011), (011), (002), (020), (111),

̅

(

111), (111) and (111) lattice planes, respectively. The registered XRD positions and the

respective d-spacing values for the PHS sample in this paper are identical to those reported

in a work of Rogers and Bauer [25], where the crystal structure of pyridinium hydrogen

sulfate was presented for the first time. Indeed, for the aforementioned XRD patterns, d-

spacing values of 8.4706, 6.5816, 6.1898, 4.8437, 4.2353, 3.8927, 3.8277, 3.7111 and

3.6184 are listed.

The X-ray patterns of the pure rice hush ash showed a typical broad diffraction peak in

the range between 2θ of 13 and 40° with a maximum at 21.8°, which implies the presence

of a high amount of an amorphous phase of silica (Fig. 1). Almost identical results can be

found in different works, where an appearance of a broad diffraction peak around 2θ = 22°

due to amorphous silica in RHA is clearly presented [26]. Except a phase of amorphous sili-

ca, the X-ray diffractogram of pure RHA in the current work contains a single XRD peak

centered at 2θ = 29.48° (d = 3.0275). The latter is indication that a crystalline phase of cal-

cite (ICDD No. 01–072–1652) present as a component in the synthesized RHA. Based on

these results, it could be proposed that a number of Si–O–Ca interactions exist in pure RHA

support. A XRD analysis of the crystalline calcium carbonate conducted by Harris et al. [27]

showed a strong reflection at 2θ value of 29.50°, which corresponds to the (104) lattice

plane in CaCO .

3

Comparing the XRD profiles of the pure supports with the corresponding catalysts, sig-

nificant alterations are detected (Fig. 1). The latter is much more pronounced in the case of

PHS/Al-NA sample. Hence, the following phenomena are observed after impregnation of

8

Journal Pre-proof

Al-NA with the ionic liquid pyridinium hydrogen sulfate: (i) reflections, previously assigned

to NH

er available in the XRD of PHS/Al-NA, (ii) diffraction peaks due to formation of new crystal-

line phases such as (NH (15.68°, 17.08°, 26.71° and 27.53°, ICDD No. 00–024–0012)

and Al (SO (12.00° and 17.83°, ICDD No. 00–023–0767) are registered, (iii) presence of

4 4 2

Al(SO ) .12H

2

O phase (12.40°, 14.31°, 16.03°, 17.66°, 21.54 and 27.73°), are no long-

4 2

) S

O

2 5

2

)

4 3

agglomerated PHS particles on the Al-NA surface is observed via appearance of XRD pat-

terns at 2θ values of 10.42°, 22.84° and 24.59° and (iv) the XRD patterns (25.50°, 35.12°,

37.68° and 43.23°) characterizing corundum in Al-NA support are slightly decreased in

intensity. Using Eq. (1), a value of 31.5 nm is calculated as an average size of the PHS parti-

cle on the Al-NA support surface. We believe that the observed phenomena are a conse-

quence of a surface ionic liquid-support interaction, probably due to a formation of bonds

between the oxygen atoms in the hydrogen sulfate anion and the metal ions in the

3+

NH

α-Al

Our statement that the [HSO

tion is supported by a work of Sobota et al. [28], where the interaction mechanism of

4

Al(SO

and the oxygen atoms in PHS should not be excluded as well.

]^–fragment is mainly involved in the IL-support interac-

4

)

2

.12H O. In addition, a surface interaction between the aluminium sites (Al ) in

2

O

2 3

4

+

–

[C

4

C

1

Im] [Tf

2

N] on a well-ordered alumina surface is investigated. It is established that the

+

–

[C

4

C

1

Im] [Tf

2

N] adsorbs in a slightly tilted orientation with respect to the surface, mainly

interacting with the support via sulfonyl groups. In another paper, the surface interactions

+

]^–and γ-Al

]^–anion

Im] cation interacts with

between [C

4

C

1

Im] [BF

4

O

2 3

are studied [29]. Results reveal that the [BF

4

3+

+

is significantly polarized on the acidic Al sites, while the [C

C

4 1

–

the superficial HO groups on the γ-Al

2

O

3

surface. Guo and co-workers [30] impregnated of

(pH ≤3) and reported for a decreased

crystallization degree of alumina due to breaking of the Al–O–Al bonds. Considering that

the active phase in the current work is a Brønsted ionic liquid with a pK value of 3.2 [31],

Al O

2 3

with an acidic nitrate solution of ZrO(NO

)

3 2

A

lower intensity of the characteristic XRD patterns for corundum in PHS/Al-NA sample in

comparison with those in pure Al-NA carrier could be attributed to the acidic aqueous solu-

tion of IL (see pt. 2.1.4) used for Al-NA impregnation.

+

]^–on RHA carrier does not affect

phase in

Results showed that the deposition of [H–Pyr] [HSO

4

the amorphous silica phase. However, the XRD peak (2θ = 29.48°) related to CaCO

3

pure RHA cannot be detected in the XRD profile of PHS/RHA sample (Fig. 1). Instead of this,

a crystalline phase of α-quartz (ICDD No. 01–070–2535) is registered by reflections at 2θ

values of 25.42° (d = 3.5011), 35.95° (d = 2.4961), 38.56° (d = 2.3329), 41.06° (d = 2.1970)

and 48.71° (d = 1.8679). These are related to the (101), (110), (102), (200) and (112) lat-

9

Journal Pre-proof

tice planes, respectively [32]. The observed results are trying to be explain using the above

presented statement that the pure RHA contains oxygen atoms, linked to silicon and a

charge balancing cation in the form of Si–O–Ca bonds. Knowing that the Si–O interaction

(

799.6 kJ/mol) is notably stronger than Ca–O (383.3 kJ/mol) [33], it could be assumed that

during RHA impregnation with IL, the Ca ions will be removed easier than Si. In other

words, a greater number of Si–O bonds is expected to present in the PHS/RHA sample in

comparison with the number of Si–O bonds in RHA. However, IL-RHA surface interaction

with the participation of Si atoms can be also proposed. A detailed analysis of the XRD pro-

file of PHS/RHA revealed an absence of diffraction patterns related to PHS particles on the

RHA surface. It proposes that the PHS particles are quite small to be identified by XRD. Tak-

ing into account that agglomerated PHS particles (31.5 nm) are registered on the Al-NA

surface, it seems that the IL–Al-NA and IL–RHA interactions are different in nature.

In order to investigate the textural properties of the synthesized supports and catalysts,

2

N adsorption-desorption analysis (Fig. 2) is carried out. Results revealed that type IV-A

isotherms characterize the obtained samples, according to the IUPAC classification [34].

Type IV isotherms are usually given by mesoporous materials, where the mechanism of

adsorption of simple gasses (such as nitrogen) begins with the formation of a monolayer on

the surface, followed by multilayer formation and finally by capillary condensation [35]. In

the case of a type IV-A isotherm, the capillary condensation is accompanied by hysteresis

loop. For the samples studied in the current work, type H-1 hysteresis is registered. The

latter is found in materials which exhibit a narrow range of uniform mesopores [36].

A detailed analysis of the textural properties of the pure supports showed that a value

2

of 71 m /g characterizes the specific surface area (S_B_E_T) of Al-NA (Table 1). In this regard,

2

almost identical S_B_E_T(78 m /g) is obtained in the case of RHA carrier. However, the total

3

pore volume (V_t_ꢀ_t_a_l) of the Al-NA sample (0.5532 cm /g) is found to be 2.27 times higher

3

than that (0.2436 cm /g) of RHA. It is correlated to the significantly larger volume (0.5258

3

cm /g) of the mesopores of Al-NA with respect to the volume (0.2161 cm /g) of the meso-

3

pores in RHA. For comparison, the volume (0.0274 cm /g) of the micropores in the Al-NA

3

and RHA (0.0276 cm /g) carriers is practically identical.

2

N adsorption-desorption analysis established that both the specific surface area and

the total pore volume of the pure supports are notably decreased after the ionic liquid dep-

2

osition. For example, S_B_E_Tof PHS/Al-NA and PHS/RHA is established to be 4 and 2 m /g,

3

respectively. In addition, values of 0.0269 and 0.0083 cm /g are found as V

for the

tꢀtal

PHS/Al-NA and PHS/RHA, respectively. These effects are manly attributed to the plugging

1

0

Journal Pre-proof

of the support pores with PHS species. It is interesting to point out that the difference (3.24

times) between the V_t_ꢀ_t_a_lof the PHS/Al-NA and PHS/RHA is greater than that (2.27 times)

between V_t_ꢀ_t_a_lof the corresponding supports. This assumes that the plugging may not be

the only reason for the registered V_t_ꢀ_t_a_ldifference. It was already showed that the PHS

phase on the RHA surface is much more dispersed in comparison with those on the Al-NA.

Therefore, the difference between V_t_ꢀ_t_a_lof the PHS/Al-NA and PHS/RHA samples could be

also attributed to a process of pores filling (being observed in a greater extent in the case of

PHS/RHA sample), occurring simultaneously with the pores plugging.

3

.2. Surface analysis

In the previous subsection was observed that the PHS particles on the RHA carrier are

significantly smaller than those on the Al-NA surfaces, which allowed us to state that the

PHS–RHA and PHS–Al-NA interactions are different in nature. To clarify this, the surface

effects on the PHS/RHA and PHS/Al-NA are studied via FT-IR and XPS. Infrared spectra of

–

1

–1

the PHS/RHA and PHS/Al-NA samples in the regions 400-900 cm and 950-3800 cm are

presented in Fig. 3. For comparison the FT-IR profiles of Al-NA, RHA and PHS are added as

well. It should be pointed out that a detailed FT-IR analysis on the PHS sample was carried

out in a previous our work [37]. Nevertheless, for a better understanding the PHS–RHA and

PHS–Al-NA interactions, a brief discussion on the vibrational modes in PHS is present in

the current work as well.

3.2.1. Vibrational assignments in the pure supports and ionic liquid

FT-IR spectrum of the Al-NA support showed that well-defined infrared bands appear

at wavenumber values of 445.4 (bending Al–O), 502.4 (octahedral coordinated aluminium

–

1

of AlO

6

), 588.3 (Al–O stretching), 615.1 (Al–O stretching) and 632.1 cm (stretching Al–O–

Al), related to vibrational modes in α-Al

2 3

O

[38]. It was found that our observations corre-

spond well with these presented by Boumaza et al. [39], where the specific signatures of

corundum powders by FT-IR spectroscopy are investigated. The authors reported that FT-

–

1

IT spectrum of corundum contains infrared bands at 447, 594 and 640 cm due to Al–O

vibrational modes.

–

1

Except infrared bands between 400 and 635 cm , absorbance of very low intensity at

–

1

7

36.6 and 830.3 cm is also registered in the FT-IR spectrum of Al–NA. The first one is due

to a torsional Al–O mode, while the second one expresses N–S bond [40]. These results

1

Journal Pre-proof

completely coincide with the XRD analysis, where a phase of ammonium alum is registered

in the Al-NA carrier. Another conformation for the presence of NH

4

Al(SO

4

)

2

.12H O is ob-

2

tained by the infrared bands at 1094.9, 1161.3 (shoulder), 1237.5 (shoulder), 1397.9 and

–

1

–1

1

447.1 cm . The bands between 1090 and 1240 cm characterize the S–OH and S=O vi-

2

ꢁ

–

1

brations in sulfate groups (SO ) [41], while that at 1397.9 cm is related to ammonium

4

+

4

–

1

cations (NH ) [42]. In addition, the absorbance at 1447.1 cm is recognized as N–H defor-

+

4

mation vibrations in NH . The infrared bands at wavenumber values of 1635.4 and 3418.4

–

1

cm indicates O–H bending and stretching vibrations, respectively, due to water [43]. It is

interesting to point out that the FT-IR spectrum of Al-NA contains bands at 2468.3 and

–

1

2

921.3 cm , correlated to impurities, adsorbed on the support surface. Hence, the band at

2

ꢁ

–

1

2

468.3 cm is assigned to CO species [37], whereas the second one is an indication for

3

carbonaceous residues (C–H) linked to the alumina network skeleton [44].

–

1

The vibrational spectrum of RHA carrier between 400 and 900 cm revealed infrared

–

1

bands at wavenumber values of 465.6, 572.2, 797.8 and 873.4 cm . The strong absorption

–

1

peaks at 465.6 (asymmetric Si–O–Si bending modes) and 797.8 cm (symmetric Si–O–Si

stretching modes) represent amorphous silica [45], while the infrared bands at 572.2

2ꢁ

–

1

(

symmetric CO₃deformations) and 873.4 cm (asymmetric CO deformations) corre-

3

spond to CaCO

3

[46]. These findings support the XRD results (Fig. 1), where amorphous

silica and calcite phases in the RHA samples are detected.

–

1

Analyzing the FT-IR spectrum of RHA in the region above 900 cm , additional infrared

peaks due to silica and calcite phases are registered. For instance, the intensive absorption

–

1

–1

band at 1092.2 cm with two well-defined shoulders at 960.6 and 1223.3 cm are referred

as asymmetric Si–O–Si stretching band, stretching mode of Si–OH and symmetric Si–O–Si

stretching vibration, respectively [47]. The presence of CaCO

3

in the RHA carrier is also

–

1

confirmed by a peak of low intensity at wavenumber value of 1424.6 cm , attributed to

2

ꢁ

–

1

asymmetric CO₃stretching mode [46]. Signatures of water (1637.9 and 3436.9 cm ) [43]

–

1

and adsorbed carbonaceous residues (2923.2 cm ) [44] are registered in the FT-IR spec-

trum of RHA as well. These are established to be less pronounced in the RHA spectrum in

comparison with the respective ones in the case of Al-NA support.

FT-IR spectrum of the unsupported pyridinium hydrogen sulfate in the low-frequency

–

1

–1

region (400-900 cm ) showed infrared peaks at 455.4 and 748.9 cm which represent

2

ꢁ

bending and stretching vibrations in the SO₄group, respectively [48]. The absorbance at

–

1

5

86.3 and 840.6 cm is correlated to a stretching O···H vibration between the pyridinium

cation and the [HSO

4

]^–fragment [37]. The bands at wavenumber values of 621.7, 677.9 and

1

2

Journal Pre-proof

–

1

8

79.4 cm is related to C=C and N=C vibrations in the aromatic ring [49]. Additional infra-

red bands due to stretching S–O vibrations in the PHS sample are registered in the region

–

1

–1

above 950 cm , namely 1002.3, 1171.5 and 1322.3 cm . In light of this, the peak at 1047.6

–

1

–1

cm and those from 1480 to 1620 cm are characteristic for stretching C=C and N=C vi-

+

brations in the [H–Pyr] cation.

Purely stretching C–H modes in the aromatic ring are detected at wavenumber values

–

1

between 3000 and 3230 cm [50]. An indication for protonated pyridine in PHS structure

–

1

is obtained via infrared bands at 1457.2, 1631.8 and 2850.9 cm , assigned to N–H vibra-

–

1

tions [40,41]. It should be noted that the band at 1631.8 cm is notably broaden, probably

due to a presence of vibrational O–H mode component in the form of N–H···O–S interaction

+

]^–moieties in PHS. A stretching O–H vibration in water is rec-

between [H–Pyr] and [HSO

4

–

1

ognized at 3409.3 cm .

3.2.2. Vibrational assignments in PHS/Al-NA and PHS/RHA samples

–

1

Results showed that the intensity of the IR bands at 445.4, 502.4 and 632.1 cm , previ-

ously assigned as vibrational modes in α-Al , slightly decrease when the ionic liquid is

immobilized on the Al-NA surface. In the same way, the peak intensity at 748.9 cm^–¹

O

2 3

2ꢁ

(

stretching SO₄vibration in pure PHS) is notably decreased in the spectrum of PHS/Al-NA.

3+

Consequently, a surface interaction between the Al sites in Al-NA support and the O at-

oms in [HSO

]^–is proposed. In other words, a greater number of Al–O bonds in the PHS/Al-

NA sample with respect to that in the pure support should be expected. This statement is

4

–

1

confirmed by the peak at 615.1 cm (related to stretching Al–O vibration) in the spectrum

of PHS/Al-NA, which is found to be more pronounces in comparison with the same band in

the IR spectrum of pure Al-NA.

–

1

On the other hand, the infrared pattern at 1171.5 cm in the spectrum of PHS under-

–

1

goes a considerable red shift to 1094.6 cm in the spectrum of PHS/Al-NA. As a result, it

2

ꢁ

completely coincides with the position of the band in pure Al-NA, characterizing SO

4

groups in the form of NH

4

Al(SO

4

)

2

.12H O. Furthermore, the intensity of the band at 1094.6

2

–

1

cm in the PHS/Al-NA spectrum is notably higher than that in the spectrum of pure Al-NA

support. Based on the aforementioned effects, we assumed that a surface interaction (Al–

O–S) between the metal ions in NH

4

Al(SO

4

)

2

.12H O and the hydrogen sulfate anion has oc-

2

curred during the impregnation process.

1

3

Journal Pre-proof

–

1

It is interesting to point out that the absorption band at 1447.1 cm (recognized as de-

+

4

formation N–H vibrations in NH ) does not present in the IR spectrum of PHS/Al-NA. In-

+

4

–

1

stead of this, the IR pattern at wavenumber value of 1397.9 cm (free NH ) in pure Al-NA

support possesses a higher intensity when the ionic liquid is immobilized on the Al-NA sur-

face. Considering that new phases of (NH

)

4 2

S O

2 5

and Al

2

(SO

4

) present in the PHS/Al-NA

3

structure (see pt. 3.1), it could be proposed that a cleavage of chemical bonds in the ammo-

nium alum structures is responsible for the increased intensity of the IR band at1397.9 cm^–

¹.

surface IL-support interactions in the form of Al–O–S and N–H···O–S bonds. The O–S com-

We assumed that the formation of (NH

)

4 2

S O

2 5

and Al

(SO )

2 4 3

phases is connected with

ponent in these bonds expresses the [HSO

]^–from the IL, while the Al and N–H denotes the

4

+

4

3+

Al and NH sites, respectively, in the Al-NA carrier structure.

Except a support-anion interaction (expressed as Al–O–S and N–H···O–S bonds) the FT-

IR spectrum of the PHS/Al-NA sample revealed that the pyridinium cation might be in-

volved in the overall PHS–Al-NA surface interaction as well. For example, the infrared band

–

1

+

]^–

at 586.3 cm (denoted as stretching vibration of O···H bond between [H–Pyr] and [HSO

4

moieties) in the spectrum of pure PHS cannot be clearly observed in the spectrum of Al-NA

–

1

supported ionic liquid. Moreover, it was established that the IR pattern at 588.6 cm (re-

lated to Al–O vibration in corundum) decreases in intensity when the PHS/Al-NA sample is

obtained. Considering these results, we stated a cleavage of the S–O···H–N bond in pure

PHS during impregnation and formation of a new hydrogen bond (Al–O···H–N) between the

3+

+

Al sites in Al-NA support and the N atom in [H–Pyr] cation. Knowing that the atomic ra-

dius of aluminium (0.125 nm) is higher than that of sulfur (0.100 nm) [51], it could be as-

sumed that the Al–O interaction in the Al–O···H–N bond (PHS/Al-NA sample) will be less

pronounced in comparison with the S–O interaction in the S–O···H–N bond (pure PHS sam-

ple). In other words, the oxygen atom in Al–O···H–N should possess a higher electron densi-

ty with respect to the oxygen atom in S–O···H–N. As a result, the N–H bond in the PHS/Al-

NA is expected to be more suppress than the same bond in the pure PHS. An evidence for

–

1

the latter is a higher intensity of the IR bands at 1457.2, 1631.8 and 2850.9 cm , (assigned

to N–H vibrations in pure PHS), than the intensity of the same bands in the spectrum of

PHS/Al-NA. Unlike intensity, the position of the bands characterizing the N–H vibrations in

pure PHS does not differ after the Al-NA impregnation with ionic liquid. Therefore, the in-

teractions S–O···H–N (in pure PHS) and Al–O···H–N (in PHS/Al-NA) are referred as very

similar in nature.

1

4

Journal Pre-proof

A detailed analysis on the FT-IR spectrum of the Al-NA supported pyridinium hydrogen

–

1

sulfate revealed that the bands in the region from 3000 to 3200 cm (C–H vibrations) is

significantly reduced in intensity in comparison with the same bonds in the spectrum of

–

1

pure PHS. In addition, the IR band at 1631.8 cm (mixed N–H and O–H vibrations) in the

spectrum of Al-NA supported ionic liquid is notably wider in comparison with the same

bond in pure PHS. Hence, it is proposed that the support-cation interaction is not limited to

the formation of Al–O···H–N bond and the carbon atoms in the aromatic ring are involved

as well. Based on the above presented results, we suggested that an additional hydrogen

bond in the form of C–H···O–Al interaction exists in the interface between pure PHS and Al-

NA support.

–

1

Similar to the bands between 3000 and 3200 cm , the absorbance at wavenumber val-

–

1

ue of 677.9 cm (C=C vibrations) in PHS/Al-NA also possesses a lower intensity with re-

spect to the same band in the spectrum of pure PHS. Moreover, the IR pattern at 879.4 cm^–¹

(

observed in PHS) practically does not present in the spectrum of PHS/Al-NA. In contrary, a

–

1

well-defined infrared peak at 1254.6 cm appears in the IR spectrum of PHS/Al-NA sam-

ple. According to the literature data, the IR absorbance around 1250 cm^–¹is a typical

stretching vibration of a single C–O bond [52]. Thus, we believe that the peak at 1254.6 cm^–

1

in the IR spectrum of PHS/Al-NA is a consequence of cation-anion interaction in the sup-

ported PHS due to formation of C–O–S bond. It should be noted that a similar interaction

(C···O–S) was observed in the pure PHS and it was denoted as strongly electrostatic in na-

ture [31]. The latter correlates well the IR spectrum of pure PHS, where a IR band due to a

–

1

stretching C–O vibration is hard to be found. The clearly registered band at 1254.6 cm in

+

]^–

the IR spectrum of PHS/Al-NA implies that the distance between [H–Pyr] and [HSO

4

fragments considerably decreases after the IL immobilization on the Al-NA. It agrees in

some extent with the XRD analysis, where a phase of agglomerated PHS particles was regis-

tered on the Al-NA carrier surface.

The vibrational spectrum of PHS/RHA sample indicated that the IR band at 465.6 cm^–¹

(bending Si–O–Si modes in pure RHA) notably decreases and completely overlaps the ab-

2

ꢁ

–

1

sorbance pattern at 455.4 cm (vibrations of the SO group in pure PHS). In addition, the

4

–

1

band at 797.8 cm (stretching Si–O–Si modes in pure RHA) is almost disappeared when

the IL is immobilized on the RHA surface. These findings are attributed to a support-anion

interaction via formation of Si–O–S bonds. Similar effects are registered in the PHS/RHA

–

1

–1

spectrum between 950 and 1230 cm , where the bands (960.6, 1092.2 and 1223.3 cm ) in

–

1

the spectrum of pure RHA and those (1002.3 and 1171.5 cm ) in the spectrum of pure PHS

1

5

Journal Pre-proof

appear as a double peak, correlated to stretching modes of Si–O and S–O bands which are

overlapped together.

–

1

It is worth noting that the absorbance bands (621.7, 677.9 and 879.4 cm ) related to

vibrations in the aromatic ring of PHS are notably decreased in intensity after the IL immo-

bilization of RHA. However, these bands are still well-defined (regardless of their low in-

tensity) in the IR spectrum of PHS/RHA. Lower intensity of the aforementioned patterns in

PHS/RHA with respect to their intensity in PHS is associated with a surface support-cation

interaction, similar to that presented earlier in the case of PHS/Al-NA sample. The fact that

the IR spectrum of PHS/RHA follows the IR profile of the pure ionic liquid (Fig. 3) could be

interpreted as a RHA surface which is significantly covered by PHS species. The latter cor-

responds with the XRD of PHS/RHA, where the presence of highly dispersed PHS particles

on the RHA is assumed. A conformation that the PHS–RHA interaction is stronger than

PHS–Al-NA is obtained comparing the IR spectra of PHS/RHA and PHS/Al-NA in the region

–

1

above 1430 cm . Hence, the absorption bands in the spectrum of PHS/RHA are found to be

more suppressed (expressed by their lower intensity) than the same bonds in the IR spec-

trum of PHS/Al-NA. Another important effect in the IR spectrum of PHS/RHA is the absence

2

ꢁ

–

1

of the absorption bands at 572.2, 873.4 and 1424.6 cm , previously connected with CO

3

vibrations of calcite in pure RHA. The latter agrees well with the XRD data, where no CaCO

3

phase is registered in PHS/RHA.

3.2.3. Core-electron levels description

In order to confirm the XRD and FT-IR results, XPS analysis of the pure carriers and the

corresponding supported ionic liquid is carried out. XPS of O 1s, C 1s, N 1s, Al 2p and S 2p,

core-electron levels for Al-NA and PHS/Al-NA are presented in Fig. 4. In this regard, XPS of

O 1s, C 1s, N 1s, Si 2p, S 2p and Ca 2p core-electron levels for RHA and PHS/RHA are pre-

sented in Fig. 5. In addition, the surface atomic concentrations of the elements in the inves-

tigated samples are depicted in Table 2.

Results showed that the O 1s spectrum of pure Al-NA support contains three compo-

nents with binding energy (BE) values of 530.6, 531.7 and 533.2 eV. The first component is

2

–

attributed to lattice O species [53], most probably in corundum and ammonium alum

2ꢁ

structures. The second one (531.7 eV) correspond to a surface oxygen in SO₄[54] and Al–

O [55] environment, while the component with the highest BE expresses oxygen-containing

species in the form of hydroxyl (O–H) groups [56]. However, the O 1s profile of pure Al-NA

1

6

Journal Pre-proof

carrier differs when pyridinium hydrogen sulfate is immobilized on the Al-NA surface. In

details, the fragments with BE of 530.6 and 533.2 eV are not observed in the O 1s spectrum

of PHS/Al-NA but only a single peak at 531.5. eV. Obviously, the position of the latter is

2ꢁ

practically identical to that (531.7) of the O 1s peak associated with SO₄and Al–O in pure

Al-NA. These results coincide with the XRD, where a transformation of NH Al(SO .12H

phase into (NH and Al (SO is observed. On the other hand, the data presented in

4

)

4 2

2

O

4 2

) S

O

2 5

2

)

4 3

Table 2 revealed that the oxygen concentration on the Al-NA (55 at.%) and PHS/Al-NA (53

at.%) surfaces does not differ. Based on this, it could be proposed: (i) absence of lattice ox-

ygen migration to the Al-NA surface during impregnation and/or (ii) formation of agglom-

erated PHS particles on the Al-NA.

Analyzing the O 1s line of pure RHA support (Fig. 5), two components with a binding

energy of 530.6 and 532.8. eV due to Ca–O in calcite [57] and Si–O–Si unit [58], respective-

ly, are observed. Although the position of the intensive peak at 532.8. eV is typical for Si–O–

Si group, the presence of Si–O–Ca fragment due to the relative electronegativities of the

component elements should not be excluded. Silicon is more electronegative than calcium,

thus the oxygen atom experiences decreased atomic shielding and has a high binding ener-

gy. The presence of calcite phase in the pure RHA carrier (also observed by FT-IR and XRD)

is confirmed by appearance of a doublet peak of Ca ꢂp_ꢃ_/₂(351.3 eV) and Ca ꢂp₃_/₂(347.6

2+

eV) with a spin-orbit splitting of 3.7 eV (Fig. 5), which is representative of Cꢄ in CaCO

59].

The ionic liquid deposition on the RHA carrier caused the low-intensive O 1s peak due

3

[

2+

to Ca–O (and/or Si–O–Ca) and these related to Cꢄ (Ca ꢂp_ꢃ_/₂and Ca ꢂp₃_/₂) to be not de-

tected in the XPS spectrum of PHS/RHA sample. These effects completely agree with the

XRD and FT-IR studies, where a phase of CaCO in the RHA supported ionic liquid is not

3

distinguished. On the other hand, the dominant peak (532.8. eV) in pure RHA is shifted to a

binding energy value of 531.7 eV in the spectrum of PHS/RHA, which is 0.9 eV lower than

Si–O–Si binding energy. The appearance of the peak at BE of 531.7 eV is recognized as Si–O

bonds formation [60] due to loss of a bond with oxygen atom in the first neighbor or the

changes in the second neighbors. Therefore, it can be concluded that Si–O–Si covalent

bonds in the RHA support are broken during the PHS/RHA synthesis. This confirms the FT-

IR investigation, where the infrared bands at 465.6 and 797.8 cm^–¹(related to Si–O–Si

modes in pure RHA) are notably decreased when PHS is immobilized on the support sur-

face.

1

7

Journal Pre-proof

XPS analysis showed that the surface oxygen concentration on the pure RHA is reduced

with 15% (from 60 to 51 at.%) when the PHS/RHA sample is obtained (Fig. 5, Table 2). Dif-

ferent reasons could be introduced as a possible explanation: (i) removing of the CaCO

3

phase during the RHA impregnation with pyridinium hydrogen sulfate, which lowers the

surface oxygen content and (ii) a significant amount of PHS species present on the support

surface. The latter implies a predominant formation of Si–O–S bonds between PHS and

RHA, being already registered by FT–IR. However, knowing that only 2 at.% characterizes

2+

the Cꢄ concentration on the pure RHA surface (Table 2), we assumed that PHS particles

present on the RHA surface are mainly responsible for decreased surface oxygen concen-

tration on the PHS/RHA carrier in comparison with that on RHA.

Comparing the oxygen concentrations of PHS/Al-NA (53 at.%) and PHS/RHA (51 at.%)

surfaces, it can be seen that these are almost identical. However, the XRD study clearly

demonstrated that the PHS particles present on the RHA surface are significantly smaller in

size in comparison with these on Al-NA. This implies that a greater number of O-containing

sites (expressed by [HSO

]^–fragments in IL) should present on the PHS/RHA surface with

4

respect to that on Al-NA supported ionic liquid. To explain this contradiction, we proposed

that during the RHA impregnation with an ionic liquid, a number of the PHS species are

–

situated perpendicular to the support surface. As a result, [H–Pyr] moieties in the immobi-

lized IL are predominantly registered by XPS. It correlates with the oxygen concentration

on the RHA surface which was found to be lower than that on PHS/RHA. The statement that

a number of PHS species are probably placed perpendicular to the RHA surface is con-

firmed to some extent comparing the C 1s profiles of PHS/Al-NA and PHS/RHA (see below).

Another indication that the PHS particles on the RHA surface possess a smaller size

than these on Al-NA and is obtained by analysis of the C 1s, S 2p and N 1s core-electron lev-

els for PHS/Al-NA (Fig. 4) and PHS/RHA (Fig. 5) samples. Prior to that, the corresponding

core-electron levels for pure supports are described. Hence, the XPS investigation showed

that C 1s spectrum of Al-NA carrier contains peaks centered at 287.6 and 285.1 eV, related

to C=O and C–C/C–H groups, respectively [61]. It agrees well with the FT–IR study of Al-NA

–

1

(

Fig. 3), where bands at 2468.3 and 2921.3 cm assigned to adsorbed impurities are de-

tected. In this regard, the C 1s line of RHA also showed presence of carbonaceous residues

C–C/C–H). However, a component at BE of 289.2 eV due to O–C=O in calcite is detected

59], in accordance with the XRD and FT–IR analyses. It should be noted that the carbon

(

[

concentration on the Al-NA surface is about 17% more pronounced than that on the RHA

support (Table 2). This supports the FT–IR data, where the infrared bands between 2800

1

8

Journal Pre-proof

–

1

and 3200 cm in the spectrum of Al-NA sample are established to be more intensive in

comparison with same bands in the spectrum of RHA. The IL deposition on Al-NA and RHA

caused the peaks at 287.6 and 289.2 eV to be not present in the C 1s spectra of PHS/Al-NA

and PHS/RHA, respectively. Moreover, the peak at BE value of 285.1 eV increased consid-

erably in intensity. The latter was especially noticeable in the case of RHA supported PHS,

where the surface carbon concentration is found to be more than 40% higher than that on

the PHS/Al-NA. This assumes that the RHA possesses a greater number of exposed PHS

particles in comparison with Al-NA due to a stronger IL-RHA surface interaction.

We came to the same conclusion, evaluating the S 2p core-electron level. Thus, the S 2p

line of pure Al-NA (Fig. 4) showed two characteristic doublets at BE values of 170.1 (S

2ꢁ

ꢂp_ꢃ_/₂) and 168.8 eV (S ꢂp₃_/₂) (Fig. 5), which are typical for SO₄groups [62]. This is in ac-

cordance with XRD and FT–IR analyses, where presence of sulfate groups in Al-NA sample

(

due to NH

doublets considerably increase in intensity and slightly shift towards lower BE values

from 170.1 to 169.8 eV and from 168.8 168.5 eV, respectively) after the Al-NA impregna-

4

Al(SO

4

)

2

.12H

2

O phase) was detected. It was found that the aforementioned

(

tion with PHS. The first effect could be related to the presence of additional sulfate-

+

]^–) on the Al-NA surface. Considering that the BE

containing phase (such as [H–Pyr] [HSO

4

difference (0.3 eV) between S 2p core-electron levels of pure Al-NA and PHS/Al-NA is not

2ꢁ

noticeable, a weak surface IL-support interaction and/or SO₄in a slightly different envi-

ronment should be expected in that case. Earlier, the XRD data revealed agglomerated PHS

particles and (NH

4 2

) S

O

2 5

and Al

2

(SO

4

) phases in PHS/Al-NA. Results showed that binding

3

energy values of 169.2 and 167.9 eV characterize the S ꢂp_ꢃ_/₂and S ꢂp₃_/₂, respectively in the

case of PHS/RHA (Fig. 5), which is 0.6 eV lower than the corresponding ones for PHS/Al-

NA. It is connected with a stronger IL–RHA interaction (via Si–O–S bonds formation) in

comparison with the IL–Al-NA one (expressed by Al–O–S connections). It is interesting to

point out that the sulfur concentration on the PHS/Al-NA and PHS/RHA surfaces are

strongly equivalent (Table 2). This seems to contradict the XRD, where the sulfate-

+

]^–) in PHS/Al-NA assume a higher

containing phases ((NH

4 2

) S

2 5

O , Al

2

(SO

4

)

3

, [H–Pyr] [HSO

4

sulfur concentration on the PHS/Al-NA surface than that on PHS/RHA. Hence, a greater

number of PHS sites on the RHA surface (due to stronger IL-support interaction) with re-

spect to these on Al-NA was supposed to clarify this discrepancy.

Figures 4 and 5 exposed that the N 1s core-electron level of PHS/Al-NA and PHS/RHA is

presented as a single peak at 401.1 and 400.9 eV, respectively, connected with a protonated

+

4

nitrogen in the form of C–N and/or NH [63]. Although a more dispersed PHS phase is

1

9

Journal Pre-proof

proposed on the RHA surface, the nitrogen concentration on PHS/Al-NA is found to be

identical to that on the RHA supported ionic liquid (Table 2). It is connected with the pres-

+

4

ence of increased amount of NH species on the Al-NA surface during impregnation, being

–

1

registered by FT–IR. Indeed, the band at 1397.9 cm in the infrared spectrum of PHS/Al-

NA sample is observed to be more intensive than the same band in the IR spectrum of pure

Al-NA support.

Similar to the N 1s core-electron level, the Al 2p line of pure Al-NA carrier is expressed

as a single peak as well (Fig. 4). Since the registered BE value of 74.3 eV correspond to alu-

minium species in octahedral coordination [53], the Al 2p peak characterizes the alumini-

um in α-Al O

2 3

and NH

4

Al(SO

4

)

2

.12H O phases [64]. While minor changes in the Al 2p peak

2

position (74.1 eV) are observed after the Al-NA carrier impregnation with pyridinium hy-

drogen sulfate, the aluminium concentration on the PHS/Al-NA surface is significantly re-

duced (from 37 to 22 at.%) in comparison with that on pure support (Table 2). Based on

the first effect, it could be assumed that the aluminium environment in Al-NA does not dif-

fer notably from that in PHS/Al-NA. This is associated well with the XRD study, where a

phase transformation of NH

4 4 2

Al(SO ) .12H

2

O into (NH

4 2

) S

O

2 5

and Al

2

(SO

4

) is detected. On

3

the other hand, lower surface Al concentration in Al-NA supported ionic liquid proposed

that the Al-NA surface is considerably covered by PHS particles, being already detected by

FT–IR.

Results showed that the effects in the RHA and PHS/RHA samples analyzed by Si 2p

core-electron level (Fig. 5) are identical to those described earlier by means of the corre-

sponding O 1s lines. For example, the component at BE of 103.6 eV in the Si 2p spectrum of

4+

pure RHA corresponds to Si (expressed as Si–O–Si bonds) in silica [65]. However, both

the binding energy and the intensity of this component decreased notably when the

PHS/RHA is synthesized. As a result, a peak with a BE value of 102.7 eV due to Si–O interac-

tion is registered. In accordance with the Al 2p analysis described above, reduced intensity

of the Si 2p peak for RHA supported ionic liquid can be associated with the presence of PHS

species on the RHA surface. Comparing the surface Si and Al concentrations on the pure

supports with the corresponding catalysts (Table 2), it can be clearly seen that the first one

decreases in a greater extent (50%) with respect to Al (40%). It implies that that a greater

number of PHS particles present on the RHA surface in comparison with Al-NA one. In oth-

er words, a higher PHS dispersion on the RHA is observed, supporting the XRD and FT–IR

studies.

2

0

Journal Pre-proof

3

.3. Electronic transitions

Electronic spectra of the obtained Al-NA, RHA, PHS, PHS/Al-NA and PHS/RHA samples

are presented in Fig. 6. Following the above used model of discussion, the electronic transi-

tions in the pure supports and ionic liquid are described first.

Results showed that the UV-vis spectrum of Al-NA carrier contains absorption bands at

wavelength values of 245, 280, 311 and 357 nm, which are attributed to the charge trans-

fers SO → ꢅꢆ³⁺[66], O → ꢅꢆ³⁺[67], SO₄→ NH [68] and OH → ꢅꢆ [69], respective-

2

ꢁ

2ꢁ

+

ꢁ

3+

4

ly. Analyzing the electronic spectrum of RHA, absorption bands due to electron density dis-

tributions CO₃→ Cꢄ²⁺(231 and 359 nm) [70] and O → Si⁴⁺(298 nm) [71] are detected,

respectively. Obviously, these assignments agree with the XDR, FT–IR and XPS studies,

where corundum and ammonium alum phases are registered in Al-NA. In addition, calcite

and silica are clearly observed in the RHA.

2

ꢁ

2ꢁ

Considering that the ionic liquid pyridinium hydrogen sulfate is composed by an organ-

+

]^–) fragments, different electronic transitions are ex-

ic ([H–Pyr] ) and a inorganic ([HSO

4

pected to present in this structure. Hence, to describe the origin of these charge transfers,

the experimental wavelengths are compared with theoretical ones. For that purpose, the

theoretical UV-vis spectrum of PHS (added as a supporting material, Fig. 1S) is computed as

a vertical electronic excitation energy from the minima of the ground-state structure (add-

ed as a supporting material, Fig. 2S) in a condensed phase (water) via a time-dependent

density functional theory [72] using a conductor-like polarizable continuum model [73].

The electronic spectrum of pure ionic liquid possesses a typical shape of protonated

pyridine [74], containing absorption peaks at 207, 239, 245, 250, 255 and 263 nm. As will

∗

be shown later, the bands in the region 200-250 nm are due to π → π transitions, while

∗

these from 250 to 270 represent n → π charge transfers [75]. The corresponding theoreti-

cal wavelength values are calculated at 209, 227, 229, 230, 246 and 275 nm, respectively.

For a detailed analysis of the electronic transitions in the PHS, frontier molecular orbitals

for six excited states are studied. The frontier molecular orbitals which are active in the

electronic absorption spectrum of PHS are HOMO, HOMO–1, HOMO–2, HOMO–3, LUMO and

LUMO+1. The theoretical analysis revealed that HOMO and HOMO–1 are strongly delocal-

ꢁ

ized on the oxygen atoms in the [HSO ] moiety as the contribution 

= 14.4%

4

ꢇꢈMꢈ

5

4

5

3

p

y

(O ) + 11.1% 4p

y

(O ) + 10.6% 3p

x

(O ) + 8.3% 4p

x

(O ) + 7.2% 2p

y

(O ) ‒ 6.5%3p

x

(O ) and

3

2

5

_ꢇ_ꢈ_M_ꢈ_ꢁ_ꢃ= 12.5% 3p

z

(O ) + 10.0% 4p

z

(O ) ‒ 8.4% 3p

x

(O ) ‒ 8.0% 3p

z

(O ) ‒ 6.6% 4p

z

(O ) +

5

6.4% 3p

x

(O ). It could be seen that HOMO is expresses by p

y

and p orbitals of the oxygen

x

2

1

Journal Pre-proof

atoms, while p

z

orbitals characterize HOMO–1. In the opposite case, LUMO and LUMO+1 are

completely delocalized on the pyridinium ring in PHS as the contribution _L_U_M_ꢈ= 12.8%

1

0

10

13

8

10

3

p

z

(C ) + 9.9% 4p

z

(C ) + 9.7% 3p

z

(N ) ‒ 6.5% 4p

z

(C ) ‒ 6.5% 3p

z

(C ) + 6.5% 2p

z

(C ) and

9

11

12

9

8

_L_U_M_ꢈ₊_ꢃ= 9.6% 3p

z

(C ) ‒ 9.4% 3p

z

(C ) + 8.5% 3p

z

(C ) + 8.3% 4p

z

(C ) ‒ 8.3% 3p

z

(C ) ‒

11

8.1% 4p

z

(C ). It is exposed that HOMO–2, HOMO–3 are manly delocalized on the O atom in

4

the hydrogen sulfate anion as the contribution _ꢇ_ꢈ_M_ꢈ_ꢁ₂= ‒ 9.7% 3p

z

(O ) ‒ 7.7% 4p (O ) ‒

z

8

12

11

4

5.8% 3p

z

(C ) + 5.7% 3p

z

(C ) + 5.0% 3 p

z

(C ) ‒ 4.8% 2p

z

(O ) and 

= ‒ 8.5%

ꢇꢈMꢈꢁ3

4

12

8

9

11

3

p

z

(O ) ‒ 6.9% 4p

z

(O ) ‒ 6.2% 3p

z

(C ) + 6.0% 3p

z

(C ) + 5.1% 3p

z

(C ) ‒ 4.9% 3p

z

(C ).

However, the _ꢇ_ꢈ_M_ꢈ_ꢁ₂and _ꢇ_ꢈ_M_ꢈ_ꢁ₃demonstrate that the carbon atoms in the aromatic

ring are notably pronounced in HOMO–2, HOMO–3 as well.

It is established that the excited states at wavelength values of 263, 255 and 250 nm are

due to the following electron density distributions HOMO → LUMO, HOMO–1 → LUMO and

HOMO → LUMO+1, respectively. Considering the above presented molecular orbitals con-

tributions, we stated that these electronic transitions indicate an interaction between the

∗

hydrogen sulfate system and the pyridinium ring, i.e. n → π charge transfers. In light of

this, the absorption bands at shorter wavelengths (245, 239 and 207 nm) are recognized as

charge transfers HOMO–2 → LUMO, HOMO–3 → LUMO and HOMO–3 → LUMO+1, respec-

tively. These are denoted as electron density distribution in the aromatic ring of PHS, i.e.

∗

π → π charge transfers.

Figure 6 exposed that the absorption peaks in the UV-vis spectra of PHS/Al-NA and

PHS/RHA samples are identical to those that characterize the electronic transitions in pure

ionic liquid. In addition, the intensity of the bands of PHS/Al-NA and PHS/RHA is estab-

lished to be remarkably higher in comparison with that of the bands registered for unsup-

ported PHS. It suggests that a significant number of PHS sites available for an electron den-

sity distribution present on the Al-NA and RHA surfaces. It is interesting to point out that

the absorption peaks of PHS/Al-NA are slightly more pronounced in intensity with respect

to these of PHS/RHA, assuming that the PHS phase on the Al-NA surface should be in a

more dispersed state than that on RHA. However, the XRD analysis clearly indicated that

the PHS particles on Al-NA are significantly bigger in size than those on RHA. Moreover, the

XPS showed that the silicon concentration on RHA surface decreases in a greater extent

than aluminium one on Al-NA when they (supports) are impregnated with PHS. Hence, we

believe that higher hindrance (expressed by a lower absorption) of the charge transfers in

PHS/RHA sample in comparison with these in PHS/Al-NA is a consequence of a stronger

IL–RHA surface interaction with respect to that between the ionic liquid and Al-NA carrier.

2

Journal Pre-proof

This statement is supported by a work of Jia et al. [76], where a UV-vis characterization of

gold complexes with pyridine-based SNS ligands is presented. The authors reported that

the absorption peaks at 225 and 321 nm in the pure pyridine-based SNS ligand become

weaker when the gold complexes are produced, due to a charge transfer between the lig-

and and metal center.

3

.4. Thermal properties

The TG and DSC curves for the thermal decomposition of Al-NA, RHA, PHS, PHS/Al-NA

and PHS/RHA are shown in Fig. 7. As can be seen, the decomposition process of Al-NA

support includes three thermal intervals, up to 110°C, from 140 to 280°C and 310-560°C.

As it was already presented by XRD, phases of corundum, ammonium alunite and

lonecreeckite present in the Al-NA structure. Taking into acount that a thermal

decomposition of α-Al

2

O phase should not be expected, the observed temperature

3

intervals are connected with degradation of ammonium alum.

Results showed that the mass loss in the first temperature region is 6.37%. Thereafter,

a weight loss of 3.49% is detected in the temperature interval from 140-280°C. In addition,

a values of 3.67% characterize the sample mass decreasing between 310-560°C. The DSC

curve of Al-NA sample demonstrated the presence of three endothermic peaks – a strong

one at 84°C and two low-intensive ones at 227 and 503°C. The first temperature region is

due to removal of physisorbed water on the Al-NA surface and NH

dehydration to NH Al(SO [77]. In the temperature interval from 140-280°C, NH

undergoes dehydroxylation and deammoniation (2NH Al(SO → Al (SO + 2NH

4

Al(SO

4

)

2

.12H

Al(SO

+ SO

2

O

4

)

4 2

4

)

2

4

)

4 2

2

4

)

3

+

H

2

O), while the sample mass decreasing between 310-560°C is related to desulphation of

(SO to 3SO and Al

The TG profile of pure RHA revealed a decompostion process before 200°C,

Al

2

)

4 3

3

2 3

O .

accompanied by a minimal weight loss of 2.34%. Theafter, a monotonically mass loss

decreasing (3.05%) up to 690°C is registered. Similar to Al-NA sample, the recorded weight

loss in the first temperature region is due elimination of adsorbed water. Comparing the

first decomposition peaks for pure Al-NA and RHA supports, it can be seen that the

adsorbed water on the Al-NA is in a greater amount. The latter was undoubtedly confirmed

by means of the corresponding DSC peaks. It is interesting to point out that a broad

endothermic peak is observed in the temperature interval from 560 to 690°C. To explain

this peak, the RHA content is taking into acount. Considering that SiO

2

and CaCO present

3

2

3

Journal Pre-proof

as phases in RHA (see XRD, XPS and FT–IR analyses), calcite is expected to decompose in

some extent during a thermal theatment of RHA. Concequently, the observed weight loss of

3.05% in the current work is attributed to carbonates decomposition [78].

The PHS thermal behaviour showed decomposition steps in the intervals from 90 to

80°C and from 250 to 550°C (Fig. 7). While a value of 5.77% registered as a mass loss in

1

the first temperature region, the weight decreases fast with elevating the temperature up

to 550°C. At that temperature (550°C) the sample is completely decomposed with a total

mass loss of 95.03%. The DSC curve of pure PHS clearly indicates two endothermic peaks, a

broad one from 50 to 170°C and a very strong one centered at 390°C. In addition, an

exothermic peak of low intensity is observed at 564°C. It is intersting to point out that the

first endothermic peak starts from temperature of 50°C, which is notably lower than the

temperature (90°C) characterizing the initial mass loss. In other words, an endothermic

effect without a mass loss is detected. We believe that the endothermic effect within 50-

90°C is a consequence of PHS melting [79]. The weight loss between 90 and 180°C is

assigned to elimination of physisorbed water and/or carbonates, bieng already detected by

FT–IR analysis (Fig. 5). Comparing the mass loss due to removal of physisorbed water from

the Al-NA, RHA and PHS surfaces, it can be seen that adsorbed water amount decreases in

the order Al-NA > PHS > RHA.

As it was noted above, the temperature within 250-550°C (expressed by an

endothermic peak at 390°C) characterizes the complete PHS decay. To offer a plausible

mechanism of PHS decomposition, the chemical bonds involved in the PHS structure are

considered (see Fig. 2S). Previous our studies showed that the PHS formation includes an

] and [H–Pyr] via S –O ⋯H –N and S¹–

–

+

1

3

7

13

interaction between the fragments [HSO

4

18

12

O ⋯H –C hydrogen bonds [80]. It was established that these bonds are electrostatic in

1

4

18

12

nature, being observed in a greater extent for S –O ⋯H –C . Taking into account that the

electrostatic bonds are weaker than those with a covalent nature [81], we suggested that a

1

3

7

13

1

4

18

12

cleavage of the bonds S –O ⋯H –N and S –O ⋯H –C is the initial step of PHS

degradation. As a result, pyridine and sulfuric acid fragments are formed. It is widely

presented that the pyridine thermal decomposition proceeds at temepratures above 700°C,

where H , HC≡CH and HCN are the major products [82]. Since the thermal analysis in the

2

current work is in the temperature interval 30-700°C (see pt. 2.2.), the mass decreasing

above 250°C should not be associated with a cleavage of C=C and/or N=C bonds in the

aromatic ring. On the other hand, a value of 115°C is registered as a boiling point of

pyridine [83]. Hence, it could be proposed that the pyridine (44.64% from the total mass of

2

4

Journal Pre-proof

+

]^–→ Pyr + H

PHS) formed above 250°C ([H–Pyr] [HSO

4

2

SO ) undergo a rapid evaporation.

4

It causes the sample mass to decrease notably. Unlike pyridine, when sulfuric acid is above

00°C, it gradually decomposes to SO and H O [83]. Thereafter, the sulfur trioxide is

transformed into SO and 0.5O . We can summarize that the temperature interval between

50 and 550°C with the corresponding DSC peak express a PHS degradation to pyridine,

SO (or SO and 0.5O ) and H O. The exothermic peak detected at 564°C implies that some

oxidation process has ocurred, most likely with the participation of SO and/or O

3

2

3

2

3

2

.

Analyzing the thermogram of PHS/Al-NA, four temperature regions of mass loss are de-

tected – from 40 to 90°C (mass loss of 2.96%), between 120 and 200°C (mass loss of

3

.75%), within 210-320°C (mass loss of 4.25%) and from 350 to 570°C (mass loss of

.51%). The corresponding DSC peaks are centered at 80, 175, 260 and 530°C, respectively

6

(

Fig. 7). Comparing the TG prfiles of pure Al-NA and PHS/Al-NA in the temperature interval

up to 100°C, it can be observed that PHS/Al-NA sample is characterized with a lower mass

loss. It is related to (i) a decreasing of exposed Al-NA surface during the ionic liquid

deposition, as it was noted earlier by C 1s, Al 2p and S 2p core-electron levels and (ii)

transformation of ammonium alum in (NH

4 2

) S

O

2 5

and Al

2

(SO

4

) . In addition, the O 1s

3

analysis indicated that the peak at BE value of 533.2 eV (referred as adsorbed O–H groups)

in the XPS spectrum of Al-NA cannot be found in the O 1s line of PHS/Al-NA sample. On the

other hand, the same temperature region (up to 100°C) showed a lower mass loss of

PHS/Al-NA (2.96%) in comparison with that of pure PHS (5.77%). This is due to formation

of large in size PHS species on the Al-NA surface, being detected by XRD.

The mass loss between 120 and 200°C in the TG curve of PHS/Al-NA is associated with

a decomposition of the immobilized ionic liquid. Obviously, the PHS degradation in PHS/Al-

NA sample starts at a notably lower temperature with respect to that registered for pure

PHS (above 250°C). Nevertheless, the mass loss for PHS/Al-NA (3.75%) in comparison with

that for pure IL (95.03%) is found to be remarkably lower. The first effect is attributed to a

surface IL-support interaction, which causes the impregnated PHS to decomposes earlier

than pure ionic liquid. However, the PHS on the Al-NA surface is in the form of

agglomerated particles, being observed by XRD. It implies that these will be hardly

decomposed. The latter is confirmed in some extent, comparing the DSC profiles of PHS and

PHS/Al-NA. Thus, the DSC curve of pure PHS described above clearly showed an

exothermic peak, denoted as an oxidation process with the contribution of SO

However, no such exothermic peak is observed in the DSC profile of PHS/Al-NA. It suggests

that notably low amounts of SO are formed due to a hard decomposion behavoiur of PHS

3

and/or O .

2

3

2

5

Journal Pre-proof

when it is deposited on Al-NA. The last two degradation steps (within 210-320°C and from

350 to 570°C) in the TG profile of PHS/Al-NA correspond to those observed above in the

case of pure Al-NA support, namely deammoniation and desulphation, respectively. It

should be noted that the mass loss (4.25%) due to PHS/Al-NA deammoniation is more

pronunced than that registered in the case of Al-NA (3.53%). This is connected with an

+

4

incresed amount of NH (revealed by N 1s and FT–IR studies) on the Al-NA support surface

during the impregnation. The same tendency is also observed for the desulphation process.

Hence, the mass loss (6.51%) due to PHS/Al-NA desulphation is established to be higher

than that registered for Al-NA (3.67%). It is ascribed to additional sulphate groups present

on the PHS/Al-NA surface (except those from support) in the form of [HSO

]^–fragments.

4

The TG curve of RHA supported pyridinium hydrogen sulfate showed two well-defined

decomposition steps (Fig. 7). The first one is in the temperature interval 40-150°C, where a

value of 8.53% is detected as a mass loss. The main decomposition step, expressed by a

weight loss of 29.96%, is between 180 to 430°C. In this regard, two endothermic peaks,

centered at 90 and 385°C, respectively, are observed in the DSC profile of PHS/RHA

sample. The mass loss within 40-150°C (related to elimination of adsorbed water) revealed

that the adsorbed water amount on the PHS/RHA surface is about 1.5 times higher than

that on PHS (5.77%) and more than 3.5 times higher than that on RHA (2.34%). XDR

analysis revealed that PHS particles cannot be detected on the RHA surface. On the other

hand, the XPS study (C 1s, Si 2p and S 2p) showed that the exposed RHA surface decreases

after PHS deposition. Hence, the above presented results are connected with highly

dispersed PHS species present on the RHA surface, which leads to a higher adsorption

capasity PHS/RHA with respect to that of pure PHS and RHA. The temperature interval

from 180 to 430°C in the TG curve of PHS/RHA expresses a thermal degradation of the

immobilized ionic liquid. It is interesting to point out that the PHS decomposition is

characterized with an endothermic peak centered at a position (385°C), which is identical

to that (390°C) detected earlier in the DSC profile of pure PHS. However, the mass loss in

the case of PHS/RHA sample (29.96%) is established to be considerably lower than that

detected for PHS (95.03%). These findings are correlated with a surface IL-RHA

interaction, which prevents the PHS thermal decomposition in some extent.

Based on the XRD, XPS, FT–IR and UV-vis, it was established that the IL-RHA surface in-

teraction is stronger that that between PHS and Al-NA support. As a result, the PHS phase

on the Al-NA surface starts to decompose at lower temperatures than the PHS phase depos-

ited on RHA. However, the highly dispersed PHS species formed on the RHA surface (due to

2

6

Journal Pre-proof

strong IL-RHA interaction) undergo a thermal decomposition in a greater extent (ex-

pressed by mass loss of 29.96%) in comparison with larger ones on the Al-NA surface,

where a mass loss of 3.75% due PHS degradation is detected.

3

.5. Catalytic activity

The catalytic performance of the obtained PHS/Al-NA and PHS/RHA samples is evalu-

ated by a process of acetic acid esterification with 1-butanol at reaction temperature of

80°C, initial alcohol-to-acid molar ratio of unit and catalyst loading of 10 wt.% (Fig. 8). For

comparison, the substrate (acetic acid) conversion for non-catalyzed (blank) reaction is

presented as well. It is established that a value of 0.049 characterizes the acetic acid con-

version when the esterification process is conducted without a catalyst. However, the addi-

tion of PHS/Al-NA and PHS/RHA (10 wt.%) considerably increased the substrate conver-

sion up to 0.278 and 0.512, respectively. Comparing the substrate conversion in the pres-

ence of PHS/RHA and PHS/Al-NA, it is clear that the catalytic performance of the former

one in the butyl acetate synthesis is more pronounced than that of the Al-NA supported IL.

This is trying to be explain as follows.

It is well known that the activation of a carboxylic acid in the presence of a Brønsted

acidic ionic liquid proceeds via proton transfer (formation of a hydrogen bond) from the IL

particles towards the substrate [84]. As a result, an active complex due to a carboxylic acid-

catalyst interaction is obtained. In this sense, it could be stated that a greater number of IL

particles present on the support surface will lead to a formation of a greater number of ac-

tive complexes. XRD, FT–IR, XPS and UV-vis analyses in the current work clearly indicated a

stronger IL-RHA surface interaction than that between PHS and Al-NA. As a result, the PHS

phase on the RHA surface is established to be remarkably more dispersed in comparison

with that on the Al-NA. Based on this, more prominent activity of the PHS/RHA catalyst

(

with respect to that of PHS/Al-NA) in the process of acetic acid esterification with butanol

is related to the presence of a great number of active sites on the RHA surface, available for

substrate activation. Considering that the ionic liquids are materials with a relatively high

viscosity, PHS particles with a large size on the Al-NA surface (a consequence of a weak

surface IL-support interaction) imply a slow mass transfer of the substrate into the active

phase (PHS). In other words, a great part of a given ionic liquid catalyst droplet will not be

in contact with the substrate. The latter limits the overall reaction rate and reduces the ion-

2

7

Journal Pre-proof

ic liquid contribution in the reaction. A plausible reaction mechanism for butyl acetate syn-

thesis in the presence of PHS/Al-NA and PHS/RHA catalysts is illustrated in Fig. 9.

4. Conclusions

Electronic, thermal and catalytic behavior of rice husk ash (RHA) and corundum (Al-

NA) supported pyridinium hydrogen sulfate (PHS) as a function of the interphase interac-

tions was studied. Characterization methods such as N

2

adsorption-desorption

measurements, XRD, FT–IR, XPS, UV-vis, and TG/DSC were applied for investigation the

surface features in the synthesized catalysts (PHS/RHA and PHS/Al-NA). The catalytic ac-

tivity of PHS/RHA and PHS/Al-NA was considered in a process of butyl acetate synthesis.

The N adsorption-desorption measurements showed that the obtained catalysts are

2

mesoporous materials with a pore diameter between 10 and 30 nm. However, a contribu-

tion of micropores was observed as well. XRD analysis established that a PHS particles with

an average size of 31.5 nm present on the Al-NA support surface when PHS/Al-NA is ob-

tained.

In contrast, diffraction peaks due to PHS particles on the RHA surface were not detect-

ed, which allowed us to state that these are highly dispersed. We attributed these effects to

a stronger surface PHS–RHA interaction than that between PHS and Al-NA. The latter was

confirmed by UV-vis study, where the electronic transition in PHS/RHA are observed to be

more hindered than those in the case of PHS/Al-NA. Based on the FT–IR and XPS analyses,

we proposed that the aforementioned surface interactions are in the form of Al(Si)–O···H–

3+

4+

N and Al(Si)–O–S bonds between the metal cations (Al and Si ) in the carriers and the

+

]^–) in the PHS. Stronger IL-RHA interaction caused the PHS

fragments ([H–Pyr] and [HSO

4

phase deposited on RHA to decompose at higher temperature (180°C) than that (120°C)

required for PHS degradation on Al-NA. However, agglomerated PHS species on Al-NA sur-

face were decomposed in a lesser extent (mass loss of 3.75%) in comparison with more

dispersed ones on RHA, where a mass loss of 29.96% was recorded. The catalytic test es-

tablished a more prominent catalytic activity of PHS/RHA with respect to that of PHS/Al-

NA in the process of butyl acetate synthesis. It was related to a great number of active sites

present on the RHA surface (due to a stronger PHS–RHA interaction), available for sub-

strate activation. Larger PHS particles on the Al-NA surface implied that a great part of a

given ionic liquid droplet will not be in a contact with the substrate, which limits the over-

2

8

Journal Pre-proof

all reaction rate. A plausible reaction mechanism for butyl acetate synthesis illustrated in

Fig. 9.

References

[

1] H. Li, P. S. Bhadury, B. Song, S. Yang, Immobilized functional ionic liquids: Efficient,

green, and reusable catalysts, RSC Adv. 2 (2012) 12525-12551.

2] I. Tankov, R. Yankova, DFT analysis, reaction kinetics and mechanism of esterifica-

tion us-ing pyridinium nitrate as a green catalyst, J. Mol. Liq. 277 (2019) 241–253.

3] D. J. Tao, X. M. Lu, J. F. Lu, K. Huang, Z. Zhou, Y. T. Wu, Noncorrosive ionic liquids

com-posed of [HSO4] as esterification catalysts, Chem. Eng. J. 171 (3) (2011) 1333-1339.

4] Y. Tao, R. Dong, I. V. Pavlidis, B. Chen, T. Tan, Using imidazolium-based ionic liquids

[

as dual solvent-catalysts for sustainable synthesis of vitamin esters: Inspiration from bio-

and organo-catalysis, Green Chem. 18 (2016) 1240–1248.

[

5] H. Amirfirouzkouhi, A.N. Kharat, Application of ionic liquids as recyclable green cata-

lysts for selective alkylation of phenol, Sep. Purif. Technol. 196 (2018) 132-139.

6] B. V. Romanovsky, I.G. Tarkhanova, Supported ionic liquids in catalysis, Russ. Chem.

Rev. 86 (2017) 444–458.

7] A. Riisager, B. Jørgensen, P. Wasserscheid, R. Fehrmann, First application of sup-

[

ported ionic liquid phase (SILP) catalysis for continuous methanol carbonylation, Chem.

Commun. 9 (2006) 994-996.

[

8] M. H. Valkenberg, C. de Castro, W. F. Hölderich, Immobilisation of chloroaluminate

ionic liquids on silica materials, Top. Catal. 14 (2001) 139–144.

9] J. M. Andanson, A. Baiker, Interactions of 1-ethyl-3-methylimidazolium trifluoro-

[

methanesulfonate ionic liquid with alumina nanoparticles and organic solvents studied by

infrared spectroscopy, J. Phys. Chem. C 117 (2013) 12210–12217.

[10] Y. Wang, W. Lou, X. Wang, A green way to synthesize lubricating ester oils: The es-

terifica-tion of pentaerythrotol with fatty acids at stoichiometric ratio over

+

]⁻/SiO

, Catal. Commun. 111 (2018) 21–25.

[

BHSO

3

MMIm] [HSO

4

2

[

11] Y. He, Q. Zhang, X. Zhan, D. guo Cheng, F. Chen, Synthesis of efficient SBA-15 immo-

bilized ionic liquid catalyst and its performance for Friedel-Crafts reaction, Catal. Today.

76 (2016) 112–120.

2

9

Article Doi

Effect of the active phase-support interaction on the electronic, thermal and catalytic properties of [H–Pyr]+[HSO4]?/support (support = rice husk ash; corundum)

DOI: 10.1016/j.molliq.2020.113725

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.molliq.2020.113725

Full text of DOI:10.1016/j.molliq.2020.113725

Products guided by the article

R&D Labs maybe for 543-54-4

Relevant to this article

Hot Product