Accepted Manuscript

Magnetically recoverable catalysts for the conversion of inulin to mannitol

Oleg V. Manaenkov, Ekaterina A. Ratkevich, Olga V. Kislitsa, Bret Lawson, David

Gene Morgan, Antonina A. Stepacheva, Valentina G. Matveeva, Mikhail G.
Sulman, Esther M. Sulman, Lyudmila M. Bronstein

PII: S0360-5442(18)30716-3

DOI: 10.1016/j.energy.2018.04.103

Reference: EGY 12742

To appear in: Energy

Received Date: 19 December 2017

Revised Date: 04 April 2018

Accepted Date: 17 April 2018

Please cite this article as: Oleg V. Manaenkov, Ekaterina A. Ratkevich, Olga V. Kislitsa, Bret
Lawson, David Gene Morgan, Antonina A. Stepacheva, Valentina G. Matveeva, Mikhail G. Sulman,
Esther M. Sulman, Lyudmila M. Bronstein, Magnetically recoverable catalysts for the conversion of
inulin to mannitol, Energy (2018), doi: 10.1016/j.energy.2018.04.103

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1 Magnetically recoverable catalysts for the conversion of

2 inulin to mannitol

4 Oleg V. Manaenkova*, Ekaterina A. Ratkevicha, Olga V. Kislitsaa, Bret Lawsonb, David Gene

5 Morganb, Antonina A. Stepachevaa, Valentina G. Matveevaa, Mikhail G. Sulmana, Esther M.

6 Sulmana, Lyudmila M. Bronsteinb,c,d

7 aTver State Technical University, A. Nikitin str. 22, 170026, Tver, Russia

8 b Indiana University, Department of Chemistry, Bloomington, IN 47405, USA

9 c King Abdulaziz University, Faculty of Science, Department of Physics, Jeddah, Saudi Arabia

10 dA.N. Nesmeyanov Institute of Organoelement Compounds of Russian Academy of Sciences,

11 Vavilova St. 28, Moscow, Russia

12

13 *Corresponding author E-mail: ovman@yandex.ru, Tel.: +7 (4822) 78-93-17

14

15 Abstract

16 Inulin is a naturally occurring polysaccharide, widely available as plant biomass. Here, we report

17 utilization of a magnetically separable Ru-containing catalyst based on magnetic silica (Fe3O4-

18 SiO2) in the inulin hydrolytic hydrogenation to mannitol (a sweetener used in diabetic foods). The

19 influence of the reaction parameters on the selectivity to mannitol has been studied. Under the

20 optimal conditions the maximum selectivity to mannitol reached 44.3% at 100% conversion of the

21 initial polysaccharide, exceeding that obtained with conventional Ru/C. The catalyst used in this

22 work is stable under hydrothermal conditions of the process. It can be easily magnetically

23 separated from the reaction mixture and reused without any loss of selectivity and activity, making

24 this catalyst promising for practical applications in biomass conversion.

25 Key words
 ACCEPTED MANUSCRIPT
1 Inulin, magnetically recoverable catalysts, hydrolytic hydrogenation, mannitol, ruthenium

2 1. Introduction

3 Polyols are important materials which are widely used in various fields. Mannitol, C6H14O6,

4 is employed in pharmaceutical, chemical, and food industry as well as in biotechnology. It is used

5 to treat brain diseases [1 - 3], as a food supplement (E421) and a sweetener in diabetic foods [4,

6 5], in the production of resins, linseed oil, coatings, surfactants, explosives, and cosmetics [6, 7].

7 In addition, nontoxic and inexpensive mannitol can be utilized as a phase change material for latent

8 heat storage [8]. These applications determine the high demand for mannitol.

9 There are several methods for mannitol syntheses. It can be obtained by electrolytic reduction

10 of glucose or by hydrogenation of invert sugars, monosaccharides, or sucrose [9]. The shortcoming

11 of the above methods is it uses food sugars, thus interfering with the food supply. This makes the

12 development of novel, effective methods for mannitol syntheses using non-food polysaccharides

13 a high priority. One of such methods is the hydrolytic hydrogenation of inulin [10, 11], a

14 polysaccharide which is not digested by humans and which is present in significant amounts in

15 such plants as Heliánthus tuberósus (16 - 18%) and Cichorium intybus (up to 20 %). Due to high

16 inulin content Heliánthus tuberósus (also known as Jerusalem artichoke, JA) is considered as

17 preferred source of bioethanol [12]. Inulin is also a promising renewable raw material to obtain

18 value-added chemicals such as 5-hydroxymethylfurfural and 2,5-furancarboxaldehyde [13 - 15].

19 The polysaccharide hydrolytic hydrogenation can be carried out as a one-pot procedure in the

20 presence of heterogeneous catalysts containing various precious metals. The reaction is performed

21 in subcritical water, which is both a hydrolysis catalyst [16] and a reagent, allowing one to carry

22 out biomass conversion without additional energy consumption for biomass drying as is needed in

23 the case of pyrolysis [17]. From this point of view, the direct conversion of inulin or inulin-

24 containing raw materials to mannitol in subcritical water is an energetically favorable process

25 similar to many other processes of biomass liquefaction using subcritical and supercritical fluids

26 [18, 19]. Because polysaccharide (including inulin) hydrolysis in subcritical water is fast, the
 ACCEPTED MANUSCRIPT
1 hydrolytic hydrogenation efficiency is mainly determined by the activity of the hydrogenation

2 catalyst. Ru-containing catalysts are considered most active in hydrolytic hydrogenation and

3 hydrogenolysis [11, 16, 20, 21].

4 Magnetically separable catalysts received considerable attention due to easy magnetic

5 separation from reaction mixtures, facilitating the catalyst repeated use and allowing one to save

6 energy and materials, thus decreasing the target product costs [22]. To date, Ru-containing

7 magnetically separable catalysts have been used in olefin metathesis [23], azide-alkyne

8 cycloaddition [24], hydrogenation [25, 26], oxidation [27], nitrile hydration [28], etc. Among other

9 reactions where such catalysts were employed, the biomass conversion is one of the most

10 promising for utilizing renewable resources. Magnetically separable catalysts have shown good

11 results in cellulose conversion [29 - 31]. It is noteworthy that when the biomass conversion is

12 incomplete, the magnetic separation is especially valuable for the catalyst recovery and the process

13 optimization.

14 In our preceding work, we developed Ru-containing magnetically separable catalyst based on

15 magnetic silica (Fe3O4-SiO2), where magnetite and Ru nanoparticles are formed in the silica pores

16 [32]. This catalyst allowed 100% of cellulose conversion and the formation of ethylene glycol and

17 propylene glycol as target molecules. Here, we explore a similar Ru-containing magnetically

18 separable catalyst in the inulin hydrolytic hydrogenation to mannitol. We demonstrate the catalyst

19 high activity and stability in repeated uses with the 100% inulin conversion and the 44.3%

20 selectivity to mannitol.

21 2. Materials and methods

22 2.1. Materials

23 Iron (III) nitrate, mesoporous silica gel (6 nm porosity, 200-425 mesh particle size) and

24 ruthenium (III) acetylacetonate (Ru(acac)3, 97%) were purchased from Sigma-Aldrich and used

25 without purification. Ethylene glycol (99.0 %) and tetrahydrofuran (THF) were purchased from
 ACCEPTED MANUSCRIPT
1 Macron Fine Chemicals and used as received. Inulin (99 %) was purchased from TCI (Japan) and

2 used without purification. D-fructose (≥ 99 %) was purchased from Merck and used as received.

3 2.2. Catalyst synthesis

4 The catalyst synthesis was carried out according to the procedure published elsewhere [32].

5 In a typical synthesis, to the solution of 2 g of Fe(NO3)3 dissolved in 10 mL of ethanol, 2.5 g of

6 silica gel was added. The mixture was allowed to stir overnight in air for ethanol evaporation. The

7 sample was then dried in a vacuum oven at room temperature for a minimum of 2 hours. The

8 powder was then stirred with a spatula, adding 25 drops of ethylene glycol. This sample was then

9 loaded into two porcelain boats and heated in a quartz tube under argon to 250 °C with a heating

10 rate of 2 °C/min. The heating at 250 °C was held for 5 h and then the sample was cooled to room

11 temperature.

12 Ru(acac)3 (0.495 g) was dissolved in 10 mL of THF and mixed with Fe3O4-SiO2 prepared in

13 the previous step. The suspension was stirred overnight in air to allow THF evaporation. The

14 sample was then dried in vacuum at room temperature until it was entirely dry. The powdered

15 product was stirred with 25 drops of ethylene glycol. The sample was then loaded into two

16 porcelain boats and heated in a quartz tube in a tube furnace under argon to 300 °C with a heating

17 rate of 2 °C/min followed by 3 h heating at this temperature. After that the product was allowed to

18 cool to room temperature.

19 Reduction of this sample was carried out before catalytic experiments by hydrogen at 300 C

20 for 2 h. After cooling the catalyst was stored in a sealed container at room temperature.

21 2.3. Characterization

22 The transmission electron microscopy (TEM) images were acquired on a JEOL JEM1010

23 transmission electron microscope. Images were analyzed with image-processing package ImageJ

24 to estimate nanoparticle diameters. High resolution TEM (HRTEM) images and scanning TEM

25 (STEM) energy dispersive X-ray spectra (EDS) were acquired at an accelerating voltage of 300

26 kV on a JEOL 3200FS transmission electron microscope equipped with an Oxford Instruments

 ACCEPTED MANUSCRIPT
1 INCA EDS system. The same TEM grids were used for all analyses. X-ray fluorescence (XRF)

2 measurements to determine the Ru content were performed with a Zeiss Jena VRA-30

3 spectrometer. X-ray powder diffraction (XRD) patterns were collected on an Empyrean from

4 PANalytical. X-rays were generated from a copper target with a scattering wavelength of 1.54 Å.

5 The step-size of the experiment was 0.02. Atomic adsorption spectrometry (AAS) to determine Ru

6 and Fe contents in the liquid phase after the catalytic reaction was carried out with MGA-915

7 (Lumex, Russia), equipped with hollow cathode lamps for Ru (349.9 nm) and Fe (248.3 nm).

8 2.4 Catalytic experiment

9 The experiments were performed in a steel reactor (50 mL, Parr Instruments, USA) equipped

10 with a controller PARR4843. Inulin (0.3 g), a catalyst (0.07 g), and 30 mL of distilled water were

11 loaded into the reactor. Then the reactor was purged with hydrogen three times under 60 bar

12 pressure. The mixture was heated and stirred (~100 rpm) to prevent the formation of local hot spots

13 and the catalyst surface was saturated with hydrogen. After reaching the operating temperature the

14 stirrer speed was increased to 600 rpm. This moment was chosen as the reaction starting time. At

15 the end of the experiment the catalyst was separated by a neodymium magnet. The non-hydrolyzed

16 polysaccharide, if any, was separated by filtration. The content of the main products was

17 determined by liquid chromatography using UltiMate 3000 (Dionex, USA).

18 Inulin conversion was calculated as X = (mi0 – mi)/mi0×100 %, where mi is the weight of the

19 recovered inulin and mi0 is the initial inulin weight. A selectivity was calculated as S = mp/(mi0 –

20 mi)×100 %, where mp is the reaction product weight.

21 3. Results and Discussion

22 3.1. Catalyst characterization

23 Fig. 1 displays the TEM and HRTEM images, EDS spectrum and XRD pattern of 5% Ru-

24 Fe3O4-SiO2 as well as the XRD pattern of magnetic silica. The EDS spectrum shows Ru, Fe, Si,

25 and O from the catalyst, Cu from the TEM grid and Cr from the grid holder. TEM and HRTEM

26 images confirm the presence of nanoparticles. While the iron oxide NP diameter is difficult to
 ACCEPTED MANUSCRIPT
1 determine from TEM because of the low electron contrast difference between iron oxide and silica,

2 the Ru-containing nanoparticles are well visible and have a diameter of 2.0 nm [32]. The XRD

3 pattern of this sample displays a broad signal at approximately 22 degrees two theta which is due

4 to amorphous silica and the set of reflections that is typical of that of magnetite [33, 34]. Analogous

5 XRD pattern is observed for Fe3O4-SiO2 (Fig. 1c). The magnetite crystallite size calculated using

6 Scherrer’s formula for the (311) reflection is 12.3 nm. Considering that the magnetite NPs are

7 located in the 6 nm silica pores (Fig. 1a), it is surprising that the crystallite size is larger by a factor

8 of 2. We assume that two or more magnetite NPs are joined together due to oriented attachment

9 [35], forming larger single crystals.

10 Fig. 1. TEM image (a), EDS spectrum (b), and XRD pattern (d) of 5 % Ru-Fe3O4-SiO2. XRD

11 pattern of Fe3O4-SiO2 is shown in (c) for comparison. Inset in (a) shows the HRTEM image.

12 It is worth noting that the (400) reflection of magnetite is overlapped with the (101) reflection

13 of the Ru (0) phase. The X-ray photoelectron spectroscopy (XPS) data presented in our preceding

14 paper [32] demonstrated the presence of Ru (IV) and Ru (0) species at the 1/1 atomic ratio,

15 indicating that the RuO2 phase is amorphous. It is worth noting that the XPS is a surface method,

16 allowing one to conclude that RuO2 is most likely located on the nanoparticle surface.

17 Liquid nitrogen adsorption measurements show that for Fe3O4-SiO2 and 5% Ru-Fe3O4-SiO2,

18 the BET surface areas are 304 m2/g and 280 m2/g [32], respectively, indicating that the

19 incorporation of Ru-containing NPs in the magnetic silica pores decreases the surface area by ~8%.

20 3.2 Inulin hydrolytic hydrogenation to mannitol

21 The 5% Ru-Fe3O4-SiO2 catalyst was studied in the hydrolytic hydrogenation of inulin to

22 mannitol (Fig. 2).

23 Fig. 2. Scheme of the hydrolytic hydrogenation of inulin to mannitol [10].

24 After the catalytic reaction, the liquid phase contained mannitol as well as sorbitol, glycerol,

25 and propylene glycol along with other trace polyols, whose amounts depend on the experimental

26 conditions. To achieve the highest possible selectivity and the complete inulin conversion as well
 ACCEPTED MANUSCRIPT
1 as high activity, a number of reaction parameters such as temperature, reaction duration, hydrogen

2 partial pressure, and the Ru/inulin ratio has been varied.

3 Fig. 3. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over temperature

4 (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; P(H2) 60 bar,

5 45 min).

6 Fig. 3 displays temperature dependencies of selectivity values in the temperature range of 140-

7 180 °C. The selectivity to mannitol reaches its maximum at 150 C. With the temperature increase

8 the selectivity considerably decreases due to the acceleration of the hydrogenolysis reaction of

9 fructose formed due to inulin hydrolysis. It is confirmed by the increase of the lower polyol

10 (hydrogenolysis product) concentration in the target product. It is noteworthy that the maximum

11 selectivity to sorbitol (14.7 %) was achieved at the higher temperature (160 °С) which is most

12 likely due to the acceleration of the fructose-to-glucose isomerization (according to the Lobry de

13 Bruyn-van Ekenstein reaction [36]). Glucose is then hydrated, forming sorbitol.

14 Fig. 4. Dependencies of the mannitol, sorbitol, glycerol, and propylene glycol selectivities on

15 time (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; P(H2)

16 60 bar, 150 °С).

17 Fig. 4 displays dependences of selectivities to mannitol, sorbitol, glycerol, and propylene

18 glycol depending on the reaction time. A typical chromatogram of the liquid phase after the

19 experiment is presented in Fig. 5a, showing mannitol along with other products. At the starting

20 reaction time (the moment when the temperature reached 150 °С), the inulin conversion was

21 already 100% (Fig. 5b) and the liquid phase contained inulin oligomers, fructose (6.3%) and

22 mannitol (16.7%). The highest selectivity to mannitol was observed after 45 min of the reaction.

23 Longer reaction times resulted in the decrease of the selectivity to hexitols due to their

24 hydrogenolysis to lower polyols, the noticeable amount of which began to form only after 30 min.

25 Fig. 5. Chromatogram of the liquid phase in the experiment with 5 % Ru-Fe3O4-SiO2: (a) after

26 45 min; (b) at the starting reaction time. The products: mannitol (1), sorbitol (2), glycerol (3),
 ACCEPTED MANUSCRIPT
1 propylene glycol (4), fructose (5), and inulin oligomers (6). (0.1167 mmol of Ru per 1 g of inulin;

2 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 255 °С; P(H2) 60 bar).

3 Fig. 6. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over hydrogen pressure

4 (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150 °C, 45

5 min).

6 Fig. 6 displays the dependence of the selectivity to mannitol and other products on the

7 hydrogen pressure. The maximum selectivity to mannitol (44.3%) exceeding that of Ru/C [11]

8 was observed at the hydrogen pressure of 60 bar. Al lower pressures the selectivity to mannitol

9 significantly decreased: to 32.1% at 40 bar. This could be explained by the decrease of the

10 hydrogen concentration in the liquid phase and as a consequence, on the catalyst surface. This is

11 confirmed by the characteristic smell and the yellowish solution color at low hydrogen pressures

12 due to fructose caramelization because fructose is too slow to convert to mannitol at the hydrogen

13 shortage. In addition, in these conditions a fraction of fructose is subjected to the retroaldol

14 decomposition [10], which is confirmed by higher selectivities to glycerol and propylene glycol at

15 the 40 bar hydrogen pressure (Fig. 6). A minor decrease of the selectivity to mannitol at 70 bar

16 could be explained by oversaturation of the catalyst active centers with hydrogen, limiting the

17 access of the reacting molecules. In addition, fructose hydrogenolysis to low polyols intensifies at

18 high pressures, also leading to the lower selectivity to mannitol.

19 Fig. 7. Dependence of the selectivity to mannitol on the Ru/inulin (mmol/g) ratio (0.3 g of

20 inulin; 30 mL of H2O; 150 °С; P(H2) 60 bar; 45 min).

21 Fig.7 depicts the dependence of the selectivity to mannitol on the Ru/inulin (mmol/g) ratio.

22 The maximum selectivity to mannitol is achieved at 0.1167 mmol of Ru per 1 g of inulin. When

23 the catalyst amount is insufficient, again caramelization of fructose and glucose is observed,

24 decreasing the selectivity to mannitol to 21.6 %. The increase of the Ru/inulin ratio also leads to

25 some decrease of the selectivity to mannitol, most likely due to the acceleration of hydrogenolysis

26 of fructose and glucose to ethylene glycol and propylene glycol.

 ACCEPTED MANUSCRIPT
1 To evaluate the catalyst stability, the catalyst was separated from the reaction medium using

2 a neodymium magnet (Fig. 8) and used in the successive experiments. The results presented in

3 Table 1 show that the selectivity to mannitol and the catalyst activity do not change after three

4 consecutive reactions, demonstrating excellent stability in the inulin-to-mannitol conversion. The

5 analysis of the liquid phase by AAS showed no Ru leaching and only traces of Fe after the first

6 catalytic reaction. After the following cycles no Fe or Ru was detected.

7 Fig. 8. 5% Ru-Fe3O4-SiO2 before (left) and after (right) magnetic separation (0.1167 mmol of

8 Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150 °С; P(H2) 60 bar; 45

9 min).

10 Table 1. The catalytic performance of 5% Ru-Fe3O4-SiO2 in the repeated use.

11 3.3 Kinetic and thermodynamic studies

12 The presence of fructose in the reaction solution (Fig. 5) indicates that among two processes

13 - inulin hydrolysis and fructose hydrogenation (Fig. 2) - the latter is the rate determining step.

14 Because of this reasoning fructose hydrogenation was chosen for kinetic and thermodynamic

15 studies. To determine the order of reaction, we carried out kinetic experiments, evaluating the

16 influence of the initial substrate (fructose) concentration on the reaction rate. The data (not

17 presented) showed that in optimal conditions (60 bar H2, 150 °С) the fructose hydrogenation is the

18 first order reaction, in a good agreement with the literature data [37, 38].

19 To assess the activation energy (Ea) of the fructose hydrogenation with 5% Ru-Fe3O4-SiO2,

20 we carried out experiments varying the temperature in the range 130-160 °С in otherwise optimal

21 conditions (0.3 g of fructose; 0.07 g of catalyst; 30 mL of H2O; P(H2) 60 bar). For calculations,

22 we used the time for 30% fructose conversion (τ0.3). The dependencies obtained are displayed in

23 Fig. 9.

24 Fig. 9. Dependences of the fructose conversion on time for different reaction temperatures

25 (0.3 g of fructose; 0.07 g of 5% Ru-Fe3O4-SiO2 catalyst; 30 mL of H2O; P(H2) 60 bar).

 ACCEPTED MANUSCRIPT
1 Based on the results presented in Fig. 9, the Arrhenius dependence was plotted in the ln 1/τ0.3

2 – 1/T coordinates (Fig. 10) and the value of Ea was calculated from the following relation:
𝐸
𝑎
𝑅∙𝑇
3 𝑘 = 𝑘0 ∙ 𝑒

4 Fig. 10. Dependence of ln 1/τ0.3 on 1/T.

5 Еa was found to be equal 7.5 ± 0.5 kJ/mol which is lower than the activations energy values

6 for other catalysts [37, 38], revealing a higher efficiency of 5% Ru-Fe3O4-SiO2.

7 CONCLUSIONS

8 We studied the behavior of the Ru-containing magnetically separable catalyst in the

9 conversion of inulin to mannitol, exploring different reaction conditions. In the hydrolytic

10 hydrogenation of inulin the maximum selectivity to mannitol of 44.3% was obtained at 150°С,

11 P(H2) 60 bar, 0.1167 mmol of Ru per 1 g of inulin for 45 min with the catalytic activity of 2.53

12 h-1. Under these conditions, the inulin conversion reached 100%. It is noteworthy that the

13 selectivity reported here exceeds that of commercially available Ru/C. Moreover, the 5% Ru-

14 Fe3O4-SiO2 catalyst is more efficient than other catalysts reported in literature, judging by the

15 activation energy. These factors and the excellent catalyst stability under hydrothermal conditions

16 as well as easy magnetic separation make 5% Ru-Fe3O4-SiO2 the catalyst of choice for practical

17 applications in biomass conversion.

18 Acknowledgments

19 The research leading to these results has received funding from the Russian Foundation for

20 Basic Research (16-08-00401, 18-08-00404). M.S., O.M., O.K. thank Russian Science Foundation

21 (project 18-19-00240) for the financial support. We also thank the Indiana University Nanoscale

22 Characterization Facility for access to the instrumentation as well as NSF grant #CHE-1048613

23 which funded the Empyrean from PANalytical.

24 References

25 [1] Okoromah CA, Afolabi BB, Wall EC. Mannitol and other osmotic diuretics as adjuncts

26 for treating cerebral malaria. The Cochrane database of systematic reviews. 2011;(4):CD004615.
 ACCEPTED MANUSCRIPT
1 [2] Prabhakar H, Singh GP, Anand V, Kalaivani M. Mannitol versus hypertonic saline for

2 brain relaxation in patients undergoing craniotomy. Cochrane Database of Systematic

3 Reviews2014, Issue 7. Art. No.: CD010026.

4 [3] Bereczki D, Liu M, Fernandes do Prado G, Fekete I. Mannitol for acute stroke. Cochrane

5 Database of Systematic Reviews 2007, Issue 3. Art. No.: CD001153.

6 [4] Grenby TH. (2011) Advances in Sweeteners. Springer. ISBN 1461285224. p. 66.

7 [5] Lawson P. (2007) Mannitol. Blackwell Publishing Ltd. pp 219-225.

8 [6] Imhof PJ, Van der Waal in: Catalytic Process Development for Renewable Materials,

9 Wiley-Vch, Weinheim, Germany 2013.

10 [7] Ohrem HL, Schornick E, Kalivoda A, Ognibene R. Why is mannitol becoming more and

11 more popular as a pharmaceutical excipient in solid dosage forms? Pharm Dev Technol 2014;

12 19(3): 257-262.

13 [8] Alva G, Lin Y, Fang G. An overview of thermal energy storage systems. Energy 2018;

14 144: 341-378.

15 [9] Budavari, S. (ed.). The Merck Index - An Encyclopedia of Chemicals, Drugs, and

16 Biologicals. Whitehouse Station, NJ: Merck and Co., Inc., 1996, p. 979.

17 [10] Rinaldi R. Catalytic Hydrogenation for Biomass Valorization. RSC Publishing, 2014.

18 310 p.

19 [11] Heinen AW, Peters JA, van Bekkum H. The combined hydrolysis and hydrogenation of

20 inulin catalyzed by bifunctional Ru/C. Carbohyd Res 2001; 330: 381-390.

21 [12] Paixao SM, Alves L, Pacheco R, Silva CM. Evaluation of Jerusalem artichoke as a

22 sustainable energy crop to bioethanol: energy and CO2eq emissions modeling for an industrial

23 scenario. Energy 2018; 150: 468-481.

24 [13] Hu S, Zhang Z, Zhou Y, Song J, Fan H and Han B. Direct conversion of inulin to 5-

25 hydroxymethylfurfural in biorenewable ionic liquids. Green Chem 2009; 11: 873–877.

 ACCEPTED MANUSCRIPT
1 [14] Shen X, Wang YX, Hu CW, Qian K, Ji Z and Jin M. One-Pot Conversion of Inulin to

2 Furan Derivatives Catalyzed by Sulfated TiO2/Mordenite Solid Acid. ChemCatChem 2012; 4:

3 2013-2019.

4 [15] Xie H, Zhao ZK and Wang Q. Catalytic Conversion of Inulin and Fructose into 5-

5 Hydroxymethylfurfural by Lignosulfonic Acid in Ionic Liquids. ChemSusChem 2012; 5: 901-905.

6 [16] Luo C, Wang S, Liu H. Cellulose conversion into polyols catalyzed by reversibly formed

7 acids and supported ruthenium clusters in hot water. Angew Chem Int Ed 2007; 46(40): 7636-

8 7639.

9 [17] Toor SS, Rosendahl L, Rudolf A. Hydrothermal liquefaction of biomass: A review of

10 subcritical water technologies. Energy 2011; 36: 2328-2342.

11 [18] Brand S, Hardi F, Kim J, Suh DJ. Effect of heating rate on biomass liquefaction:

12 differences between subcritical water and supercritical ethanol. Energy 2014; 68: 420-427.

13 [19] Brand S, Susanti RF, Kim SK, Lee H-S, Kim J, Sang B-I. Supercritical ethanol as an

14 enhanced medium for lignocellulosic biomass liquefaction: influence of physical process

15 parameters. Energy 2013; 59: 173-182.

16 [20] Matveeva VG, Sulman EM, Manaenkov OV, Filatova AE, Kislitza OV, Sidorov AI,

17 Doluda VYu, Sulman MG, Rebrov EV Hydrolytic hydrogenation of cellulose in subcritical water

18 with the use of the Ru-containing polymeric catalysts. Catalysis Today 2017; 280: 45-50.

19 [21] Kobayashi H, Ito Y, Komanoya T, Hosaka Y, Dhepe PL, Kasai K, Hara K, Fukuoka A.

20 Synthesis of Sugar Alcohols by Hydrolytic Hydrogenation of Cellulose over Supported Metal

21 Catalysts. Green Chem 2011; 13(2): 326-333.

22 [22] Wang D, Astruc D. Fast-Growing Field of Magnetically Recyclable Nanocatalysts. Chem

23 Rev 2014; 114(14): 6949-6985.

24 [23] Che C, Li W, Lin S, Chen J, Zheng J, Wu J-C, Zheng Q, Zhang G, Yang Z, Jiang B.

25 Magnetic nanoparticle-supported Hoveyda-Grubbs catalysts for ring-closing metathesis reactions.

26 Chem Commun 2009; 5990–5992.

 ACCEPTED MANUSCRIPT
1 [24] Wang D, Salmon L, Ruiz J, Astruc D. A recyclable ruthenium (II) complex supported on

2 magnetic nanoparticles: A regioselective catalyst for alkyne–azide cycloaddition. Chem Commun

3 2013; 49: 6956-6958.

4 [25] Baruwati B, Polshettiwar V, Varma RS. Magnetically recoverable supported ruthenium

5 catalyst for hydrogenation of alkynes and transfer hydrogenation of carbonyl compounds.

6 Tetrahedron Lett 2009; 50: 1215-1218.

7 [26] Baig RBN, Varma RS. Magnetic Silica-Supported Ruthenium Nanoparticles: An

8 Efficient Catalyst for Transfer Hydrogenation of Carbonyl Compounds. ACS Sustain Chem Eng

9 2013; 1: 805-809.

10 [27] Podolean I, Kuncser V, Gheorghe N, Macovei D, Parvulescu VI, Coman SM. Ru-based

11 magnetic nanoparticles (MNP) for succinic acid synthesis from levulinic acid. Green Chem 2013;

12 15: 3077-3082.

13 [28] Baig RBN, Varma RS. A facile one-pot synthesis of ruthenium hydroxide nanoparticles

14 on magnetic silica: Aqueous hydration of nitriles to amides. Chem Commun 2012; 48: 6220-6222.

15 [29] Podolean I, Negoi A, Candu N, Tudorache M, Parvulescu VI, Coman SM. Cellulose

16 Capitalization to Bio-chemicals in the Presence of Magnetic Nanoparticle Catalysts. Top Catal

17 2014; 57(17-20): 1463-1469.

18 [30] Zhang C, Wang H, Liu F, Wang L, He H. Magnetic core-shell Fe3O4@C-SO3H

19 nanoparticle catalyst for hydrolysis of cellulose. Cellulose 2013; 20: 127-134.

20 [31] Zhang J, Wu S, Liu Y. Direct conversion of cellulose into sorbitol over a magnetic

21 catalyst in an extremely low concentration acid system. Energy Fuels 2014; 28: 4240-4246.

22 [32] Manaenkov OV, Mann JJ, Kislitza OV, Losovyj Y, Stein BD, Morgan DG, Pink M,

23 Lependina OL, Shifrina ZB, Matveeva VG, Sulman EM, and Bronstein LM. Ru-Containing

24 Magnetically Recoverable Catalysts: A Sustainable Pathway from Cellulose to Ethylene and

25 Propylene Glycols. ACS Applied Materials & Interfaces 2016; 8(33): 21285-21293.
 ACCEPTED MANUSCRIPT
1 [33] Tian Y, Yu B, Li X, Li K. Facile Solvothermal Synthesis of Monodisperse Fe3O4

2 Nanocrystals with Precise Size Control of One Nanometre as Potential MRI Contrast Agents. J.

3 Mater. Chem. 2011; 21: 2476-2481.

4 [34] Guivar JAR, Martinez AI, Anaya AO, De Los Santos Valladares L, Felix LL, Dominguez

5 AB. Structural and Magnetic Properties of Monophasic Maghemite (γ-Fe2O3) Nanocrystalline

6 Powder. Adv. Nanopart. 2014; 3: 114-121.

7 [35] Niederberger M, Coelfen H. Oriented Attachment and Mesocrystals: Non-classical

8 Crystallization Mechanisms Based on Nanoparticle Assembly. Phys. Chem. Chem. Phys. 2006; 8:

9 3271-3287.

10 [36] Miljkovic M. Carbohydrates: Synthesis, Mechanisms, and Stereoelectronic Effects 2009;

11 543 p.

12 [37] Ahmed MJ, Kadhum AAH. Hydrogenation of D-fructose over activated charcoal

13 supported platinum catalyst. Journal of the Taiwan Institute of Chemical Engineers 2011; 42: 114-

14 119.

15 [38] Kuusisto J, Mikkola J-P, Casal´ PP, Karhu H, Vayrynen J, Salmi T. Kinetics of the

16 catalytic hydrogenation of d-fructose over a CuO-ZnO catalyst. Chemical Engineering Journal

17 2005; 115: 93-102.

 ACCEPTED MANUSCRIPT
1 Table captions

2 Table 1. The catalytic performance of 5% Ru-Fe3O4-SiO2 in the repeated use.

3
 ACCEPTED MANUSCRIPT
1 Table 1

2 The catalytic performance of 5% Ru-Fe3O4-SiO2 in the repeated use.

Catalytic activity calculated as a gram of

Catalyst use Selectivity, %
mannitol per gram of the catalyst per hour, h-1

1 44.3 2.53

2 43.7 2.50

3 43.6 2.49

0.1167 mmol of Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150

°С; P(H2) 60 bar, 45 min.

 ACCEPTED MANUSCRIPT
1 Figure captions

2 Fig. 1. TEM image (a), EDS spectrum (b), and XRD pattern (d) of 5 % Ru-Fe3O4-SiO2. XRD

3 pattern of Fe3O4-SiO2 is shown in (c) for comparison. Inset in (a) shows the HRTEM image.

4 Fig. 2. Scheme of the hydrolytic hydrogenation of inulin to mannitol.

5 Fig. 3. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over temperature (0.1167

6 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; P(H2) 60 bar, 45 min).

7 Fig. 4. Dependencies of the mannitol, sorbitol, glycerol, and propylene glycol selectivities on time

8 (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; P(H2) 60 bar,

9 150 °С).

10 Fig. 5. Chromatogram of the liquid phase in the experiment with 5 % Ru-Fe3O4-SiO2: (a) after 45

11 min; (b) at the starting reaction time. The products: mannitol (1), sorbitol (2), glycerol (3),

12 propylene glycol (4), fructose (5), and inulin oligomers (6). (0.1167 mmol of Ru per 1 g of inulin;

13 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 255 °С; P(H2) 60 bar).

14 Fig. 6. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over hydrogen pressure

15 (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150 °C, 45

16 min).

17 Fig. 7. Dependence of the selectivity to mannitol on the Ru/inulin (mmol/g) ratio (0.3 g of inulin;

18 30 mL of H2O; 150 °С; P(H2) 60 bar; 45 min).

19 Fig. 8. 5 % Ru-Fe3O4-SiO2 before (left) and after (right) magnetic separation (0.1167 mmol of Ru

20 per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150 °С; P(H2) 60 bar; 45 min).

21 Fig. 9. Dependences of the fructose conversion on time for different reaction temperatures (0.3 g

22 of fructose; 0.07 g of 5% Ru-Fe3O4-SiO2 catalyst; 30 mL of H2O; P(H2) 60 bar).

23 Fig. 10. Dependence of ln 1/τ0.3 on 1/T.

24
 ACCEPTED MANUSCRIPT

2 Fig. 1. TEM image (a), EDS spectrum (b), and XRD pattern (d) of 5 % Ru-Fe3O4-SiO2. XRD

3 pattern of Fe3O4-SiO2 is shown in (c) for comparison. Inset in (a) shows the HRTEM image.
 ACCEPTED MANUSCRIPT
glucose OH OH catalyst sorbitol OH OH
+H2
O OH
HO HO
OH OH OH OH
catalyst
+H2
Inulin hydrolysis isomerization
hydrogenation

fructose OH O catalyst mannitol OH OH

+H2
OH OH
HO HO
OH OH OH OH
1

2 Fig. 2. Scheme of the hydrolytic hydrogenation of inulin to mannitol [10].

 ACCEPTED MANUSCRIPT

2 Fig. 3. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over temperature

3 (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; P(H2) 60

4 bar, 45 min).

5
 ACCEPTED MANUSCRIPT
1

50 mannitol
sorbitol
glycerol
40 propylene glycol
Selectivity, %

30

20

10

0
0 15 30 45 60 75 90

2
Time, min
3 Fig. 4. Dependencies of the mannitol, sorbitol, glycerol, and propylene glycol selectivities

4 on time (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O;

5 P(H2) 60 bar, 150 °С).

6
 ACCEPTED MANUSCRIPT

1
2

3 Fig. 5. Chromatogram of the liquid phase in the experiment with 5 % Ru-Fe3O4-SiO2: (a) after

4 45 min; (b) at the starting reaction time. The products: mannitol (1), sorbitol (2), glycerol (3),

5 propylene glycol (4), fructose (5), and inulin oligomers (6). (0.1167 mmol of Ru per 1 g of

6 inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 255 °С; P(H2) 60 bar).

7
 ACCEPTED MANUSCRIPT

50

40
Selectivity, %

mannitol
30 sorbitol
glycerol
propylene glycol
20

10

0
40 50 60 70
Hydrogen pressure, bar
1

2 Fig. 6. Selectivity to mannitol, sorbitol, glycerol, and propylene glycol over hydrogen

3 pressure (0.1167 mmol Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O;

4 150 °C, 45 min).

5
 ACCEPTED MANUSCRIPT
1

50

40
Selectivity, %

30

20

10

0
0

01

01

02

03

04
0

0
00

00

00

00

00
00

00

00

00

00
Ru loading (mmol) per 1 g of inulin
05

10

15

20

25
2
0.

0.

0.

0.

3 0.
Fig. 7. Dependence of the selectivity to mannitol on the Ru/inulin (mmol/g) ratio (0.3 g of inulin;

4 30 mL of H2O; 150 °С; P(H2) 60 bar; 45 min).

5
 ACCEPTED MANUSCRIPT

2 Fig. 8. 5% Ru-Fe3O4-SiO2 before (left) and after (right) magnetic separation (0.1167 mmol of Ru

3 per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150 °С; P(H2) 60 bar; 45

4 min).

5
 ACCEPTED MANUSCRIPT
100

Fructose convertion, %
80

60

40 130 °С
140 °С
20 150 °С
160 °С
0
0 50 100
τ, min 150 200 250
1

2 Fig. 9. Dependences of the fructose conversion on time for different reaction temperatures (0.3 g

3 of fructose; 0.07 g of 5% Ru-Fe3O4-SiO2 catalyst; 30 mL of H2O; P(H2) 60 bar).

4
 ACCEPTED MANUSCRIPT

1/T
-5.4
0.0023 0.0024 0.0025
-5.6

Ln1/τ0.3
-5.8
-6
-6.2
-6.4
-6.6
-6.8
-7
-7.2
1

2 Fig. 10. Dependence of ln 1/τ0.3 on 1/T.

 ACCEPTED MANUSCRIPT
The behavior of the Ru-containing magnetically separable catalyst in the

conversion of inulin to mannitol was studied.

The maximum selectivity to mannitol of 44.3% was obtained.

The inulin conversion reached 100%.

 ACCEPTED MANUSCRIPT

Table 1

The catalytic performance of 5% Ru-Fe3O4-SiO2 in the repeated use.

Catalytic activity calculated as a gram of

Catalyst use Selectivity, %
mannitol per gram of the catalyst per hour, h-1

1 44.3 2.53

2 43.7 2.50

3 43.6 2.49

0.1167 mmol of Ru per 1 g of inulin; 0.3 g of inulin; 0.07 g of catalyst; 30 mL of H2O; 150

°С; P(H2) 60 bar, 45 min.