Available online at website: https://journal.
id/index.
php/bcrec Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
2026, 213-225 Original Research Article Crystal Phase-Dependence of Ru@TiO2 Catalysts on the Product Selectivity in the Aqueous Phase Hydrogenolysis of Furfuryl Alcohol Thea Seventina Desiani Bodoi1.
Shauqi Aulia Rifwanda1.
Rodiansono Rodiansono1,2*.
Atina Sabila Azzahra2,3.
Utami Irawati1.
Ferensa Oemry4.
Gagus Ketut Sunnardianto4.
Indri Badria Adilina5,6.
Takayoshi Hara7* 1Department of Chemistry.
Lambung Mangkurat University.
Jl.
Yani Km 36,0 Banjarbaru 70714.
Banjarbaru.
Indonesia.
2Catalysis for Sustainable Energy and Environment (CATSuR.
Inorganic Materials and Catalysis (IMCa.
Laboratory.
Lambung Mangkurat University.
Jl.
Yani Km 36,0 Banjarbaru 70714.
Indonesia.
3School of Chemistry.
Joseph Black Building.
University of Glasgow.
Glasgow.
G12 SQQ.
United Kingdom.
4Research Centre for Quantum Physics.
BRIN.
KST BJ Habibie.
Serpong.
Tangerang Selatan.
Indonesia.
5Research Centre for Chemistry.
BRIN.
KST BJ Habibie.
Serpong.
Tangerang.
Indonesia.
6Research Centre for Catalysis.
BRIN.
KST BJ Habibie.
Serpong.
Tangerang.
Indonesia.
7Department of Applied Chemistry and Biotechnology.
Graduate School of Engineering.
Chiba University, 1-33.
Yayoi-cho.
Inage-ku.
Chiba 263-8522.
Japan.
Received: 28th November 2025.
Revised: 14th January 2026.
Accepted: 15th January 2026 Available online: 17th January 2026.
Published regularly: April 2026 Abstract The crystal phase-dependence of ruthenium supported on titania (Ru@TiO.
catalysts on the product selectivity in the aqueous phase hydrogenolysis of furfuryl alcohol (FFal.
was investigated.
The supported ruthenium nanoparticles (RuNP.
catalysts on TiO2 with different phases, c.
rutile (R), anatase (A), and brookite (B) were employed.
The Ru@TiO2(R) catalysed the hydrogenation-rearrangement reaction of furan ring to afford cyclopentanone/cyclopentanol (CPO/CPL) as the main product.
The presence of high surface acidity in Ru@TiO2(R) catalyst promoted the hydrogenation-rearrangement of furan ring leading to CPO/CPL as the main product as indicated by NH3-TPD and pyridine-ATR-IR results.
In contrast, the Ru@TiO2(A) catalyst selectively hydrogenolysed the furan ring to produce 1,5-pentanediol .
,5-PeD).
This high selectivity of 1,5-PeD over Ru@TiO2(A) catalyst may be affected by the high dispersion of Ru NPs on TiO2 facets as depicted by the high H2-uptake and small particle sizes.
Copyright A 2026 by Authors.
Published by BCREC Publishing Group.
This is an open access article under the CC BY-SA License .
ttps://creativecommons.
org/licenses/by-sa/4.
Keywords: TiO2 facets.
furfuryl alcohol.
hydrogenolysis-rearrangement.
CPO/CPL, 1,5-pentanediol How to Cite: Bodoi.
Rifwanda.
Rodiansono.
Azzahra.
Irawati.
Oemry.
Sunnardianto, .
Adilina.
Hara.
Crystal Phase-Dependence of Ru@TiO2 Catalysts on the Product Selectivity in the Aqueous Phase Hydrogenolysis of Furfuryl Alcohol.
Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 213-225.
(DOI: 10.
9767/bcrec.
Permalink/DOI: https://doi.
org/10.
9767/bcrec.
Introduction The development of selective heterogeneous catalysts for the efficient transformation of biomass-derived platform furfural (FFal.
into ,O-diols such as 1,2-, 1,4- or 1,5-pentanediol (PeD) has been attracted great attentions .
* Corresponding Author.
Email: rodiansono@ulm.
id (R.
Rodianson.
t_hara@faculty.
chiba-u.
jp (Takayoshi Har.
Since the furan ring consisted of two reactive functional groups.
terminal aldehyde (C=O) and unsaturated furang ring (C=C), the reaction of FFald involved two-competitive reaction steps: .
hydrogenation of C=O and C=C bonds to THFalc or .
direct furan ring cleavage to 1,2-PeD and 1,5-PeD, which the major product of diols was depended on the catalyst properties and reaction conditions (Scheme .
Ae.
However, achieving bcrec_20547_2025 Copyright A 2026.
ISSN 1978-2993.
CODEN: BCRECO
Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 214 a high selectivity and high yield synthesis of 1,2PeD or 1,5-PeD from either FFald or FFalc is challenging due to the reactivity of the furan ring and the complex interactions on the catalyst surface .
Ae.
Tomishige et al.
highlighted that increasing the selectivity of 1,5-PeD or 1,2-PeD from FFalc is associated with metal active sites and the basicity or acidity of catalysts used .
Nevertheless, metal-acid catalysts with strong acidity led to the hydrogenative rearrangement of furan ring to C5-cyclic compounds such as cyclopentanone (CPO) or cyclopentanol (CPL) .
Consequently, designing and controlling the selectivity of catalyst towards hydrogenative rearrangement to CPO/CPL or hydrogenolysis of furan ring to 1,2- or 1,5-PeD still remains a significant challenge.
Metal oxides, such as CeO2.
ZrO2, or TiO2 supported-PGM catalysts, played important role in the selective hydrogenation/hydrogenolysis FFald.
FFalc, or THFalc to 1,2-PeD or 1,5-PeD .
Ae.
For instance.
Pt/CeO2-nanocubes with the ceria-terminal facets exposed .
afforded higher 1,2-PeD yield .
%) than that of Pt/CeOnanorods and Pt/CeO-octahedron catalysts .
The dispersion of Ru nanoparticles (RuNPS) on rutile.
-TiO2 appeared to be higher dispersion and narrow sizes distribution due to strong interaction between RuO2 and r-TiO2 during thermal treatment.
As the results.
Ru/rTiO2 exhibited higher activity and thermal stability than that of Ru/anatase.
-TiO2 catalyst in CO2 methanation .
Similarly.
Mageed et al.
reported that the selectivity product of CO2 methanation over Ru/r-TiO2 catalysts are driven by the temperature reduction.
Ru metal loading amount as well as the dispersion on TiO2.
Therefore, controlling the catalytic properties by varying the support particle size, is of general relevance for a larger number of catalysts and reactions .
Furthermore, crystal-phasedepended strong metal-interaction of TiO2 supported-metal hydrodeoxygenation of m-cresol as a function of temperature reduction during the catalyst preparation as reported by Cui et al.
Most recently Tan et al.
reported that the aqueous phase hydrogenative-rearrangement of FFalc to CPO was crucially dictated by the Ni-O-Ti interface structure, governing nickel speciation and reactivity.
Obviously.
Ni on mixed-phase P25 exhibited highest FFalc conversion .
%) affording 89% CPO with specific rate of 72 h-1, which much higher than that of Ni on pure anatase catalyst .
Most recently.
Amin et al.
highlighted that the selectivity of Ru/TiO2 catalysts on the 5-hydroxymethylfurfural (HMF) reaction was strongly depended on morfological of TiO2, dispersion of active metal (RuNP.
, and exposure of active metal on surface of TiO2.
Catalyst of Ru/TiO2 with larger surface area resulted higher dispersion of RuNPs and exhibited the hydrogenolysis of hydroxyl group to 5-methylfurfural .
-MF).
On the other hand.
Ru/TiO2 catalyst with lower surface area and high
particle sizes of RuNPs hydrogenated C=O as well
C=C
2,5bis.
furan (BHMF) and 2,5bis.
tetrahydofuran (BHMTHF) .
Therefore, controlling the dispersion and metal exposore on oxide supports is crucial for switching the activity and selectivity of furan In the present report, the extended investigation on the effect of TiO2 phase as the support of RuNPs catalysts for hydrogenolysis or hydrogenation-rearrangement of FFalc to the target product of diols .
,2-, 1,4-, or 1,5-PeD) or CPO/CPL.
Motivated by the recent reports in our group, the modification of g-Al2O3 with TiO2(A) Tetrahydrofurfuryl alcohol (THFal.
Furfural (FFal.
Chemoselective Furfuryl alcohol (FFal.
H ,H O
1/2 H2
4-Hydroxy-2-cyclopentenone Cyclopentanone (CPO) (HCP) Cyclopentanol
(CPL)
H /H2O
4,5-dihydrofuranmethanol .
,5-DHFM) Pent-2-ene-1,5-diol 1,5-Pentanediol .
,5-PeD) Scheme 1.
Possible reaction pathways for the transformation of FFalc into high-added-value chemicals .
CPO/CPL and 1,5-PeD) using heterogeneous Ru-based catalysts .
Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 215 and TiO2(R) as the support of Ru-Sn catalysts demonstrated a higher yield of 1,5-PeD with CPO/CPL as the minor product, depending on the type of TiO2.
anatase(A) or rutile(R) .
Herein, the crystal-phase dependence of ruthenium supported on titania c.
Ru@TiO2(R) and Ru@TiO2(A) catalysts on the product selectivity in the aqueous phase hydrogenolysis of of has been investigated.
The Ru@TiO2(R) catalyst allowed to produce CPO/CPL .
p to 82% yiel.
, while the Ru@TiO2(A) afforded 1,5-PeD .
p to 80% yiel.
as the main product at 180 oC, initial H2 1030 bar for 3-5 h.
A series characterisation techniques, such as XRD.
H2-TPR.
NH3-TPD, and pyridine-ATR-IR, were employed to gain the catalytic results and catalyst structure-activity relationship and discussed systematically.
Materials and Methods 1 Materials Ruthenium .
chloride x hydrate (RuCl3A xH2O), furfuryl alcohol .
% GC), dodecane .
%), 2-methoxyethanol.
ZrO2.
Nb2O5, and ZnO were purchased from Sigma Aldrich Co.
The active charcoal (AC) (SBET = 600 m2.
and OeAl2O3 (SBET = 129 m2.
were purchased from Merck Millipore Co.
Sodium hydroxide (NaOH pellet, 99,0%) and ethanol (C2H5OH, 96%) were purchased from Sigma-Millipore.
TiO2 anatase (TiO2(A)) and TiO2 rutile (TiO2(R)) were purchased and used as received from Hongwunewmaterial (HWNANO) Ltd Co.
Methods Catalyst Preparation Preparation Ru@TiO2(A):
Typical procedure for the synthesis of Ru@TiO2(A) (Ru = 4 wt%) is described as follows .
A 0.
3702 mmo.
of ruthenium.
chloride x hydrate (RuCl3 A xH2O) was dissolved in deionised water at room temperature under gentle stirring.
A one-gram TiO2(A) and RuCl3 solution were mixed at the room temperature, then the temperature was raised to 50 oC and kept in stirring for 12 h.
The pH was adjusted to 9-10 by addition dropwise of an aqueous solution NaOH .
1 M).
The mixture was transferred to the sealed-Teflon autoclave reactor for hydrothermal processes at 150 oC for 24 h.
The obtaining black .
solid precipitate was washed with distilled water and acetone, and then dried in vacuo at room temperature for overnight.
Prior to characterisation and catalytic reaction, the black solid of Ru@TiO2(A) was reduced with hydrogen at 400 oC for 2 h.
For comparison, the TiO2 (B) support was prepared using a simple hydrolysis of TiCl4 solution with concentrate HCl at room temperature to obtain a clear solution of TiCl4.
Around 20% of TiCl4 was mixture with NaOH 6 M then transferred into the sealed-Teflon autoclave reactor for hydrothermal processes at 150 oC for 24 h.
The obtained white precipitate was washed with distilled water and acetone and then dried in vacuo for overnight.
Prior to use as support, the obtained white solid TiO2 brookite was dried at 110 oC for 12 h, then followed by calcination 300 oC under N2 for 2 h .
,23,.
Catalyst Characterisations The X-ray diffraction (XRD) analysis was performed on a Miniflex 600 Rigaku instrument with Cu as monochromatic source of Cu-K radiation ( = 0.
1544 n.
The XRD was operated at 40 kV and 15 mA with a step width of 0.
02o, a scan speed of 4o min-1 .
= 0.
1540 nm, 2 = 0.
, solar slit 1.
25, and using a Ni K filter.
Inductively-coupled plasma atomic emission spectroscopy (ICP-AES) measurements were performed on an SPS 1800H plasma spectrometer by Seiko Instruments Inc.
Japan (Ru: 267.
87 nm and Ti: 337.
280 n.
at Chiba University.
Scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDS) samples were taken on the Phenom Pro G6 Desktop SEM with acceleration voltage of 15 kV at Advanced Laboratory of Mathematics and Natural Sciences Faculty.
Lambung Mangkurat University.
The Brunauer-Emmett-Teller (BET) surface area (SBET) and pore volume (V.
were measured using N2 physisorption at -196 oC on a Belsorp Max (BEL Japa.
The samples were degassed at 200 oC for 2 h to remove physisorbed gases prior to the measurement.
The amount of nitrogen adsorbed onto the samples was used to calculate the SBET via the BET equation.
The pore sizes distribution and pore volume was estimated to be the liquid volume of nitrogen at a relative pressure of approximately 0.
995 according to Horvath-Kawazoe (HK) approach based on desorption data .
The NH3-TPD was conducted on a Belsorp Max (BEL Japa.
The samples were degassed at elevated temperature of 10Ae200 oC for 2 h to remove physisorbed gases prior to the The temperature was then kept at 200 oC for 2 h, while flushed with helium gas.
NH3
alanced NH3, 80% and He, 20%) was introduced at 100 oC for 30 min, then evacuated by helium gas to remove the physisorbed NH3 also for 30 min.
Finally, temperature programmed desorption was conducted at temperature of 100 Ae 800 oC and the desorbed NH3 was monitored by TCD.
The H2-TPR was performed on a Chemisorb 2750.
Micromeritics.
The samples were heated at 110 oC for 2 h under N2 stream with flow rate of 40 ml/min, then cooled to room temperature.
Before reduction processes, the line was purged Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 216 with H2 .
% Ar gas v/.
for 30 min, then reduced with the same gas (H2 .
% Ar v/.
) at elevated temperature of 30-700 oC with ramping 10 oC/min.
The H2 uptake was calculated by using calibration curve (H2 gas.
5% Ar gas v/v, and flow rate of 40 ml/mi.
The calculation of the mean metal particle size .
VA) from H2 can be achieved based on the volumeAearea mean diameter equation.
Eq.
yccyccVA = .
ayaycaycam ) ycoyco D is metal dispersion .
he ratio of the total number of metal atoms on the support surface to the total number of metal atoms in the bulk sampl.
Im is the volume occupied by a metal atom .
in the bulk .
65 y 10-3 nm3 for R.
and am is the surface area occupied by an exposed surface metal atom .
The unit cell of hcp Ru has only two atoms and when the .
plane is exposed, the surface will have one Ru atom available for H2 adsorption.
When the .
plane is exposed, two Ru atoms will be exposed including the atom at the .
/3, 2/3, 1/.
In addition, reported results have shown that two other lowindex planes, the .
planes, have similar levels of surface energy compared to the .
plane of Ru .
Therefore, the .
, .
, and .
low-index planes may contribute equally to H2 adsorption.
Similar to the treatment by Masthan et al.
, .
, and .
planes are used to calculate exposed crystal-plane area .
instead of using only the .
Therefore, the surface area occupied by one Ru atom .
verage from areas of .
, .
, and .
09 y 10-2 nm2 instead of 35 x 10-2 nm2 when assuming that only the .
plane is exposed.
D .
etal dispersion, %) = (Vmon/22.
y S y M y 100 / .
etal weight%) .
Vmon = monolayer coverage of H2 on Ru, cm3/g STP.
M = atomic mass of metal.
g/mol, and S = stoichiometric factor of H2 to Ru H2 is assumed to be dissociatively adsorbed on Ru metal surfaces, which is S = 2, as most researchers use it .
Catalytic Reactions Typical catalytic reaction procedure is described as the follows.
Catalyst .
FFalc .
, dodecane .
02 mmo.
, and H2O .
mL) as solvent were placed into a glass reaction tube, which fitted inside a stainless-steel reactor.
The reactor was flushed with H2 for 30 times, after an initial H2 pressure of 30 bar was introduced at room temperature, the reactor was heated to 180 After 3 h, the reaction mixtures were transferred into sample bin, centrifuged .
0 rpm for 10 mi.
and analysed by using GC-FID.
The used Ru@TiO2(R) catalyst was separated using either simple centrifugation or filtration, dried overnight under vacuum at room temperature, and re-activated with H2 at 400 oC for 2 h prior to reusability testing.
Product Analysis GC analyses of the reactant (FFal.
and products .
,5-PeD.
THFalc, 4,5-DHFM) were performed on a Perkin Elmer Auto System XL equipped with a flame ionization detector and Restek RtxA BAC Plus 1 capillary column.
analysis was operated at detector and injector temperatures of 250 C and 240 oC, respectively.
N2 as a carrier gas .
mL/mi.
, rates of air and H2 were 450 ml/min and H2 45 ml/min.
Products were confirmed by the comparison of their GC and GC-MS retention time and mass spectra with the literatures, except for 4,5-DHFM due to the limitation of commercial availability.
The calibration curve was performed using known concentrations of internal standard .
, reactant and products to determine the correct response factors.
The conversion of FFalc and the yield of the products were calculated according to the following equations:
yaya Oe yayat yayayayayayayayayayayayayayayayayayayaya (%) = 0 ycUycUycUycUycUycUycUycUycUycU (%) = yaya0 mol of product yayao y 100% y Conversion (%) .
F0 is the introduced mol reactant (FFal.
Ft is the remained mol reactant, which are all obtained from GC analysis using an internal standard technique.
Results and Discussion Screening of Catalyst for Reaction of FFalc to 1,5-PeD In attempting to find the most suitable heterogeneous supported ruthenium catalyst for aqueous phase hydrogenolysis of furfuryl alcohol to diols .
,2-PeD or 1,5-PeD), we then investigated the effect of various supported Ru catalysts at 180 oC.
H2 30 bar, in H2O for 3 h.
The results of catalytic reaction of FFalc over various supported Ru are summarised in Table 1.
At the first experiments, the catalytic reaction of FFalc using a commercially available Ru/C .
wt% R.
was conducted and the products were mixture of 2-MeF and 2-MeTHF .
%) and CPO/CPL 21% at 100% conversion of FFalc .
By using Ru@TiO2(R) catalyst, 84% of FFalc was converted into 82% CPO/CPL and small amount of side product .
%) .
Interestingly, a remarkable different of reaction product was Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 217 obtained over Ru@TiO2(A) catalyst under the same reaction conditions.
Over this catalyst, the main product was 1,5-PeD .
%) with small amount of 4,5-DHFM .
%) and CPO/CPL .
%) at 87% FFalc conversion .
The Ru@TiO2(B) catalyst with structure of brookite, rutile, and anatase phases (TiO2(B)) for comparison was also synthesised and tested for the reaction of FFalc to 1,5-PeD under the same conditions.
An 89% FFalc was converted and the products were distributed to 45% 1,5-PeD, 3% 4,5-DHFM, 15% THFalc, and 26% CPO/CPL .
Furthermore, ruthenium supported on carbon-doped TiO2(A) (Ru@CTiO2(A)) and carbon-doped TiO2(R) (Ru@CTiO2(R)) catalysts afforded diols .
,5-PeD and 1,2PeD mixtur.
and CPO/CPL with total yield of 2031% and 4-8%, respectively .
ntries 5-.
These results indicate that the presence of carbon doping affected to the product distribution of FFalc reaction .
Catalysts ruthenium-based supported on various metal oxides .
OeAl2O3.
ZrO2.
ZnO, and Nb2O.
gave THFalc as the main product .
-81%) with 18-22% yield of 1,5-PeD at completed reaction under the same conditions .
ntries 7-.
Based on these results, it can be presumably concluded that Ru@TiO2(R) selectively catalysed the furan ring rearrangement to CPO/CPL while Ru@TiO2(A) promoted the hydrogenolysis of C-O bond of FFalc towards 1,5-PeD or 1,2-PeD.
complete this conclusion, a series investigation on .
emperature, initial H2 pressure, and time profile.
both Ru@TiO2(R) and Ru@TiO2(A) 2 Evaluation of Reaction Ru@TiO2(R) vs Ru@TiO2(A) Parameters:
Effect of reaction temperature The influence of reaction temperature on the FFalc conversion and product distribution was investigated and the results are shown in Figure Figure 1.
Effect of reaction temperature on the FFalc conversion and yields of 4,5-DHFM.
CPO/CPL, and 1,5-PeD over .
Ru@TiO2(R) and .
Ru@TiO2(A) catalysts.
Reaction conditions: Cat.
FFalc .
H2O .
mL).
H2 .
, 160-200 oC, 3 h.
Table 1.
Results of selective hydrogenolysis of FFalc using various supported Ru-based catalysts.
Reaction conditions: Cat.
FFalc .
H2O .
mL).
H2 .
, 180 oC, 3 h.
aConversion and yield were determined by GC (FID) using an internal standard technique.
The carbon balance was more than 95% for all the reactions.
bOthers include the condensation product of FFalc .
ccording to the GC-MS dat.
Entry Catalyst Conv.
(%)
Ru@C Ru@TiO2(R) Ru@TiO2(A) Ru@TiO2(B) Ru@C-TiO2(A) Ru@C-TiO2(R) Ru@OeAl2O3 Ru@ZrO2 Ru@ZnO Ru@Nb2O5 1,5-PeD .
,2-PeD) 26.
4,5DHFM Yielda (%) CPO
THFalc
(CPL)
Copyright A 2026.
ISSN 1978-2993 2-MeF .
MeTHF) Othersb Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 218 As the Ru@TiO2(R) catalyst selectively produced CPO/CPL, the influence of reaction temperature was firstly evaluated during FFalc reaction at 160-200 oC.
H2 30 bar for 3 h.
The conversion of FFalc smoothly increased as reaction temperature was increased to reach 100% conversion.
At 160 oC, the product was only CPO/CPL .
%) at 58% conversion of FFalc.
When reaction temperature was increased to 180 oC, the conversion of FFalc slightly increased to 72% to afford 68% CPO/CPL and 4% 4,5-DHFM.
However, temperature to 200 oC caused not only FFalc was increased to 100% but also the formation 1,5-PeD .
%) was also observed while yield of CPO/CPL slightly increased to 70% (Figure 1.
In the case of Ru@TiO2(A) catalyst, the conversion of FFalc slightly increased as the reaction temperature was elevated from 160 oC to 200 oC to reach 90% FFalc conversion.
The yield of 1,5-PeD increased gradually from 73% at 160 oC to 80% and 81% at 180 oC and 200 oC, respectively.
Similar to 1,5-PeD yield, the amount of CPO/CPL was nearly constant .
-5%) at the temperature of 160-180 oC, but remaining 2% after reaction temperature was increased to 200 oC.
In contrast, yield of 4,5-DHFM increased smoothly as the reaction temperature was increased (Figure 1.
These results suggest that catalytic reaction of FFalc using Ru@TiO2(A) did not depend on the current reaction temperature, which is consistent with the hydrogenation of LA to GVL .
or hydrogenation of functionalized carboxylic acids .
using similar Ru/TiO2 catalyst as reported Therefore, it can be concluded that the optimised reaction temperature for the synthesis of CPO/CPL from FFalc over Ru@TiO2(R) catalyst was 160-180 oC, while the synthesis of 1,5-PeD from FFalc over Ru@TiO2(A) catalyst was 180-200 However, further study on the influence of catalyst preparation .
, thermal treatment or loading amoun.
on the catalytic activity and selectivity is important to ensure the role of structure-activity Effect of reaction time The reaction profiles of FFalc conversion as a function of reaction time at 3 h and 5 h over Ru@TiO2(R) and Ru@TiO2(A) catalysts are shown in Figure 2.
Over Ru@TiO2(R) catalyst, the conversion of FFalc was 72% to produce 68% CPO/CPL and 4% 4,5-DHFM and 1,5-PeD product was not observed after reaction time of 3 h.
After reaction was extended to 5 h, the conversion of FFalc slightly increased to 84% with the main product was CPO/CPL .
%), whereas other products were 1,5-PeD .
%) and 4,5-DHFM .
%) Figure 2.
The formation small amount of 1.
5PeD and remaining 4,5-DHFM seems likely due to the side reaction of FFalc to 4,5-DHFM then followed by ring opening to 1,5-PeD occurred as the subsequent reaction after 5 h.
It has been reported that, the competitive reaction between hydrogenation-hydrogenolysis of furan ring to diols .
,5-PeD or 1,2-PeD) and furan ring rearrangement might be regulated by the surface acidity of catalyst systems, both Brynsted and Lewis acidity .
In contrast to Ru@TiO2(R) catalyst.
Ru@TiO2(A) catalyst exhibited slightly higher FFalc conversion and displayed different product distribution (Figure 2.
After 3 h.
FFalc conversion was 88% to selectively produce 1,5-PeD .
%), 4% CPO/CPL, and 4% 4,5-DHFM.
These results indicate that the reaction of FFalc over Ru@TiO2(A) Nevertheless, the transformation of FFalc may involve two reaction pathways, first, partial Figure 2.
Effect of reaction time on the FFalc conversion and yields of 4,5-DHFM.
CPO/CPL, and 1,5-PeD over .
Ru@TiO2(R) and .
Ru@TiO2(A) catalysts.
Reaction conditions: Cat.
FFalc .
H2O .
mL).
H2 .
, 180 oC, 3-5 h.
Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 219 hydrogenation of C=C furan ring to form 4,5DHFM then followed by the ring opening of C2-O of formed 4,5-DHFM to selectively produce 1,5PeD .
Second, furan ring rearrangement on the acidic surface of catalyst to form HCP, then followed by hydrogenation to CPO .
To confirm these suggestions, the reaction was extended to 11 h both of catalysts.
As expected.
Ru@TiO2(R) catalyst converted 100% FFalc and the products were 16% 1,5-PeD, 83% CPO/CPL, and small amount of remained 4,5-DHFM .
%).
On the other hand.
Ru@TiO2(A) afforded 87% 1,5-PeD, 10% CPO/CPL, and 3% 4,5-DHFM at 100% FFalc conversion after 11 h.
This indicates that Ru@TiO2(A) catalyst system preferentially follow the reaction pathways of 1,5-PeD via 4,5-DHFM intermediate which in accordance with the previous work using bimetallic Ru-Sn catalysts .
Since the 4,5-DHFM intermediate is not commercially available, the synthesis of 4,5DHFM using our Ru-Sn catalysts, isolation, purification, and 1H- and 13C-NMR analyses would be the important for the next investigation.
Effect of initial H2 pressure The reaction of furan ring of FFalc in the presence of transition metal catalyst strongly affected by hydrogen concentration.
At high concentration of H2 over Ru or Ni-based catalysts proceeded C=C hydrogenation to produce tetrahydrofurfuryl alcohol (THFal.
The affinity of Ru or Ni-based catalysts towards C=C hydrogenation can be regulated by two simple modifying with second electropositive metal to form bimetallic alloy .
or constructing on the metal oxide supports that enable strong metal support interaction (SMSI) phenomena .
The effect of initial H2 pressure on the FFalc conversion and yields of 4,5-DHFM.
CPO/CPL, and 1,5-PeD over .
Ru@TiO2(R) and .
Ru@TiO2(A) catalysts is shown in Figure 3.
Both Ru@TiO2(R) Ru@TiO2(A) demonstrated low affinity towards C=C hydrogenation as indicated by the total hindrance of THFalc formation alongside the increase of initial H2 pressure.
By using Ru@TiO2(R) catalyst, the increase in the initial H2 pressure to 30 bars resulted the low conversion .
%) but remained high selectivity of CPO/CPL .
% yiel.
A small 5-DHFM .
%) was observed without the formation of 1,5-PeD (Figure 3.
On the other hand, at lower initial H2 pressure .
-20 ba.
, the conversion of FFalc reached maximum.
However, while the amount of CPO/CPL remain high .
5%), the formation of 1,5-PeD significantly increased to 18-25% with 7-8% yield of 4,5-DHFM was remained (Figure 3.
Furthermore, similar product profiles were also obtained from hydrogenolysis of FFalc over Ru@TiO2(A) catalyst.
The completed conversion of FFalc .
%) was obtained at initial H2 pressure of 10 bar with products were distributed to 84% 1,5-PeD, 1% CPO/CPL, and 15% 4,5-DHFM.
The amount of 4,5-DHFM increased as the initial H2 pressure was decreased, whereas CPO/CPL decreased oppositely (Figure 3.
These results suggest that the cleavage of C2-O bond of furan ring over both Ru@TiO2(R) and Ru@TiO2(A) catalysts preferentially occurred under low concentration of H2, which consistent with the previous reports .
,20,.
Therefore, it can be presumably concluded that the optimised reaction parameters for FFalc conversion towards CPO/CPL using Ru@TiO2(R) was 180 oC.
H2 30 bar for 3 h, while for the synthesis of 1,5-PeD over Ru@TiO2(A) was 180 oC, 10-20 bar H2 after 3 h.
Catalyst Structure-Activity Relationship Figure 4 shows the XRD patterns of Ru@TiO2(A) (Fig.
) and Ru@TiO2(R) (Figure 4.
) catalysts which were synthesised by using Figure 3.
Effect of initial H2 pressure on the FFalc conversion and yields of 4,5-DHFM.
CPO/CPL, and 1,5-PeD over .
Ru@TiO2(R) and .
Ru@TiO2(A) catalysts.
Reaction conditions: Cat.
FFalc .
H2O .
mL).
H2 .
-30 ba.
, 180 oC, 3 h.
Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 220 coprecipitation-hydrothermal from commercially available of TiO2 rutile (TiO2(R)) and TiO2 anatase (TiO2(A)) at 150 oC for 24 h, followed by reduction with H2 at 400 oC for 2 h.
The sole structure of anatase in Ru@TiO2(A) catalyst was clearly observed as indicated by series diffraction peaks of TiO2(A) phases at 2 = 3o, 37.
8o, 48.
0o, 54.
0o, 55.
1o, 62.
7o, 68.
8o, 70.
1o, and 82.
8o (JCPDS No.
which can be assigned as .
, .
, .
, .
, .
, .
, .
, .
, .
, and .
planes of TiO2 anatase phases, respectively (Figure 4.
The average crystallite sizes of TiO2.
anatase phase at 2 = 25.
3o was 36.
2 nm (Table .
In the case of of rutile as indicated by a series of sharp diffraction peaks at 2 = 27.
32o, 36.
06o, 41.
36o, 56.
40o, 63.
10o and 69.
38o (JCPDS No.
which can be attributed as .
, .
, .
, .
, .
, .
, and .
planes of TiO2 rutile phases, respectively (Figure 4.
The average crystallite sizes of TiO2.
at 2 = 32o in Ru@TiO2(R) sample was 26.
5 nm.
small diffraction peak of metallic Ru.
was clearly observed at 2 = 44.
06o (JCPDS No.
over Ru@TiO2(R) sample, whereas the diffraction peak of metallic Ru.
was not observed over Ru@TiO2(A) sample.
This can be attributed to due to strong metal support interaction (SMSI) effect in Ru@TiO2(A) may dominate than that of in Ru@TiO2(R), resulting in the coverage of Ru nanoparticles by TiOx overlayers and leading to higher dispersion RuNPs as it has been previously observed by Lin et al.
and Azzahra .
The hydrogen-temperature programmed reduction (H2-TPR) of catalysts help to identify the surface species during the hydrogen reduction The H2-uptake, metal dispersion, and particle diameter the synthesised Ru@TiO2 catalysts are summarised in Table 2.
It is found that the specific surface area BET (SBET) derived from N2 adsorption-desorption for Ru@TiO2(A) and Ru@TiO2(R) samples was 37.
1 m2.
g-1 and 28.
g-1, respectively (Table .
In the case of Ru@TiO2(A) sample, a high intensity of reduction peak at 73 oC, which can be attributed to the reduction peak of RuO2 to Ru0 that interacting with TiO2(A) support (Figure 5.
A broad reduction peak at around 295 oC over this sample was also observed, suggesting typical reduction of RuO2 with strong interaction with the TiO2 .
The highest reduction peak was observed at 441 oC with low and broad intensity, which can be suggested as the extendedreduction peak of RuO2 with strong interaction with TiO2(A).
Zhao and co-workers have suggested that the reduction peak at around 400-450 oC due to the strong-metal support interaction (SMSI) phenomena in TiO2 supported Ru metal catalysts .
A series reduction peaks at 276 oC, 323 oC and 441 oC were clearly observed over Ru@TiO2(R) sample (Figure 5.
As similar to the Ru@TiO2(A) sample, two distinctive reduction peaks at 276 oC and 323 oC were observed, which Figure 4.
XRD patterns of .
Ru@TiO2(A) and .
Ru@TiO2(R) catalysts after reduction with H2 at 400 oC for 2 h.
Figure 5.
H2-TPR profiles of .
Ru@TiO2(A) and .
Ru@TiO2(R) catalysts.
Table 2.
Physico-chemical properties (H2-uptake, metal dispersion, and particle diameter of synthesised Ru@TiO2 catalysts.
Entry Catalyst Ru@TiO2(A) Ru@TiO2(R) SBETa .
H2 uptakeb .
Dc (%) dVAd .
TiO2e .
aSpecific surface area BET (SBET), calculated from the amount of adsorbed N2 from N2 adsorption-desorption data.
bThe H2 uptake was derived from H2-TPR data.
cMetal dispersion (%).
dThe mean metal particle size .
VA).
eAverage crystallite sizes of TiO2.
anatase at 2 = 25.
3o and TiO2.
rutile at 2 = 27.
3o using the Scherrer`s equation.
Copyright A 2026.
ISSN 1978-2993 Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 221 can be assigned as the typical reduction of RuO2 with strong interaction with the TiO2.
However, the reduction peak at 441 oC for Ru@TiO2(R) is hardly distinguished, suggesting the SMSI phenomena over this sample is less extend and it may less intimate interaction between Ru and TiO2 surface .
It is found that the H2 uptakes of Ru@TiO2(A) and Ru@TiO2(R) catalysts 52 mmol.
g-1 and 1.
36 mmol.
g-1, respectively (Table .
The dispersion and particles sizes of Ru metal were also estimated from the H2 uptake data using the proposed equation of Shen et al.
with an assumption that the surface area occupied by an exposed surface of Ru.
It is found that the dispersion (D) of Ru nanoparticles in Ru@TiO2(A) and Ru@TiO2(R) was 21.
4% and 2% with and the mean metal particle size .
VA) 2 nm and 8.
3 nm, respectively (Table .
The NH3-TPD profiles were formally divided into three desorption temperature regions to denote three types of acid sites .
: .
weak acid sites, ranging from 100 to 200 oC, .
moderate acid sites, ranging from 200 to 350 oC, and .
strong acid sites, ranging from >350 oC (Figure .
Both Ru@TiO2(A) and Ru@TiO2(R) have similar patterns of desorption peaks with different intensities.
A desorption peak at 138-168 oC is assigned as the weak acid sites, and the desorption peak at 214-282 oC is attributed as the strong acid sites.
No strong acid sites were observed upon both of samples (Figure 6.
The amount of each acid sites and total acid sites are summarised in Table 3.
Ru@TiO2(A) catalyst consisted of 87 AAmol NH3.
g-1 as the weak acid sites and 112 AAmol NH3.
g-1 as the medium acid sites .
On the other hand.
Ru@TiO2(R) comprised 201 AAmol NH3.
g-1 as the weak acid sites and 707 AAmol NH3.
g-1 as the medium acid sites .
The presence large amount of strong acid sites may affect the product selectivity as indicated by the high yield of THFalc and CPO/CPL (Table 1, entry .
Though the Ru@TiO2(A) has much lower total acidity of 199 AAmol NH3 per gram than that of Ru@TiO2(A) .
AAmol NH3 per gra.
, whereas the types of acid sites of those catalysts were almost same (Figure Since NH3-TPD does not allow for the differentiation of Lewis and Brynsted acid sites, pyridine-ATR-IR analysis was conducted on the synthesised catalysts of Ru@TiO2(A) and Ru@TiO2(R) and the results are shown in Figure Figure 7.
Pyridine-FTIR spectra of Ru@TiO2(A) and .
Ru@TiO2(R) catalysts.
Figure 6.
NH3-TPD profiles of .
original spectra and .
deconvoluted spectra of Ru@TiO2(R) and Ru@TiO2(A) catalysts, respectively after reduction with H2 at 400 oC for 2 h.
Table 3.
Acidic properties of the synthesised Ru@TiO2 catalysts.
aAcidity was derived from NH3-TPD spectra according to the formula as reported by Azzouz et al.
Entry Catalyst Ru@TiO2(A) Ru@TiO2(R) Weaka .
-200 oC) Mediuma .
-350 oC) Copyright A 2026.
ISSN 1978-2993 Strong (>350 oC) Total aciditya (AAmol NH3.
Bulletin of Chemical Reaction Engineering & Catalysis, 21 .
, 2026, 222 According to the literatures on pyridine adsorption peaks in Ru-containing catalysts .
the bands are assigned as follows: the pyridinium ion (PyH ), which forms from the reaction of pyridine with Brynsted acid sites (B) .
coordinatively bound pyridines on Lewis acid sites (L) shows bands around 1445 .
and 1575 cm-1.
and physisorbed or hydrogen-bonded pyridine (H) exhibits bands at near 1437 and 1599 cm-1.
The band around 1490 cm-1 represents common vibrations from both PyH (B) and coordinatively bound pyridine (L) .
The Py-ATR-IR spectrum of Ru@TiO2(A) confirmed the sole presence of Lewis acid sites, indicated by bands at 1435, 1480, 1586 cm-1 .
In contrast, the Ru@TiO2(R) sample displayed the designed bands at 1637 cm1, signifying the presence of Brynsted acid sites.
However, the role distinguished acid site.
Brynsted or Lewis acid sites in the catalyst for both Ru@TiO2(A) and Ru@TiO2(R) catalysts during FFalc transformation remained unclear.
Therefore, the investigations on role of types of acid sites (Brynsted or Lewi.
and the detail structure catalyst-activity relationship system are still necessary for future study.
Conclusions We have described the crystal phasedependence of ruthenium supported on titania (Ru@TiO.
catalysts on the product selectivity in the aqueous phase hydrogenolysis of furfuryl alcohol was investigated.
Two types of TiO2 with different phases of rutile (R) dan anatase (A) were employed as the support.
The Ru@TiO2 catalysts were synthesised by using coprecipitationhydrothermal at 150 oC for 24 h, followed by reduction with H2 at 400 oC for 2 h.
Results of FFalc reaction showed that Ru@TiO2(R) catalyst produced high yield of CPO/CPL .
p to 82%), whereas Ru@TiO2(R) catalyst afforded 1,5-PeD .
% yiel.
at 180 oC, initial H2 10-30 bar for 3-5 XRD patterns of Ru@TiO2(A) revealed the sol structure of TiO2 anatase, while Ru@TiO2(R) exhibited the presence of Ru.
and pristine structure of TiO2 rutile.
These suggest that the dispersion of Ru on TiO2(A) is higher than that of on TiO2(R), which consistent with results of H2 uptake, dispersion (D), average diameter of Ru NPs, conversion of FFalc.
The presence of higher surface acidity promoted the hydrogenationrearrangement of furan ring leading to CPO/CPL as the main product as indicated by NH3-TPD and pyridine-ATR-IR results.
This high selectivity of 1,5-PeD over Ru@TiO2(A) catalyst may be affected by the high dispersion of Ru NPs on TiO2 facets as depicted by the high H2-uptake and small particle Therefore.
Further investigations on the role of crystal-phase both of TiO2(A) and TiO2(B) using advanced techniques .
XPS.
FFalc- or CO-DRIFT spectroscopies or HR-TEM) and microkinetic studies are challenging for future Acknowledgment The authors acknowledge The Indonesian Endowment Funds for Education (LPDP) through BRIN-RIIM2 scheme .
ontract number of 79/IV/KS/11/2.
DRPTM-Kemendiktisaintek through Regular Fundamental scheme .
ontract number of 056/E5/PG.
PL/2.
, and LPPMULM through Internal Fundamental scheme .
ontract number of 1878/UN8.
2/PG/2.
for financial support.
We also acknowledge the facilities, scientific and technical support from Advanced Chemical Characterization Laboratory.
National Research, and Innovation Agency through E- Layanan Sains - BRIN.
CRedit Author Statement Author Contributions: Thea Seventina Desiani Bodoi: Writing original draft.
Formal Analysis.
Investigation.
Experiment.
Shauqi Aulia Rifwanda.
Atina Sabila Azzahra: Formal Analysis.
Investigation.
Experiment.
Rodiansono:
Conceptualization.
Editing-Review.
Supervision.
Ferensa Oemry.
Gagus Ketut Sunnardianto.
Indri Badria Adilina: Advanced Characterisation and Data Interpretation.
Takayoshi Hara: EditingReview.
Supervision.
All authors have read and agreed to the published version of the manuscript.
References