Fuel Processing Technology 265 (2024) 108143

Contents lists available at ScienceDirect

Fuel Processing Technology

journal homepage: www.elsevier.com/locate/fuproc

Competitive effects of compounding aromatic hydrogen storage carriers in

low-pressure hydrogenation reactions
Xiaopeng Mei a,b, Zixuan Ma a,b,*, Yingjie Yang a,b, Xiaofeng Gao a,b,* , Hantao Gong a ,
Ziyu Song a,b , Siyu Yao a,b,*
a
Key Laboratory of Biomass Chemical Engineering of Ministry of Education, College of Chemical and Biological Engineering, Zhejiang University, Hangzhou 310027,
China
b
Institute of Zhejiang University-Quzhou, Quzhou 32400, China

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

Keywords: The reaction activity of various liquid organic hydrogen carriers (LOHCs) over 5 wt% Rh/C (BET surface area
LOHCs 933.5 g/cm3, pore size 4.6 nm, metal dispersion 10.5 %) and 5 wt% Ru/C (BET surface area 888.4 g/cm3, pore
Low-pressure hydrogenation size 6.1 nm, metal dispersion 8.9 %) catalysts is evaluated. The results show that monocyclic aromatic hydro­
Competitive effect
carbons have the highest reactivity, followed by monocyclic aromatic rings, while polycyclic and fused cyclic
aromatic hydrocarbons have relatively low activity. It is also found that mixing different LOHCs leads to a
competitive effect, resulting in lower reactivities for all LOHCs. As the degree of LOHC hydrogenation increases,
the adsorption of multi-step hydrogenation intermediates becomes more difficult, resulting in lower yields of
fully hydrogenated products. It is important to understand the behavior of LOHCs in hydrogenation reactions and
to optimize the performance of LOHCs compound systems.

1. Introduction upstream process of LOHCs is generally maintained at low-pressure

level. If the hydrogenation reaction is unable to escape the constraints
Hydrogenation reaction of aromatics is currently a significant route of the high pressure, additional gas compression is required, thus
for the preparation of industrial organic products with higher value increasing the operating costs of the process. Guo et al. [13] investigated
[1–3]. Commercial noble metal catalysts, such as Ru- and Rh-based H2 compression to 98 MPa using different methods and found that a
catalysts, have been widely and maturely used in the hydrogenation of conventional gas compression storage method needs an additional en­
various organic compounds for the production of higher value-added ergy input about 90 kJ/kg for the compressor. After calculating the
products, and high reaction yields can be achieved [4–12]. However, additional energy consumption of pre-cooling, the power required by
industrial hydrogenation processes for the preparation of downstream the entire system is 116 kJ/kg, with a maximum instantaneous
products cannot be directly migrated to other systems that require hy­ compression power of 9.3 kW. This implies a large amount of additional
drogenation reaction, due to the fact that traditional hydrogenation energy consumption, and the high instantaneous power requires a high
processes generally require extremely high reaction temperatures and power of the compressor. Although the more advanced supercritical
ultra-high hydrogen pressures (several MPa to tens of MPa) in order to compression technology can control the unit energy consumption of
ensure high yields and fast reactions. Such stringent conditions are hydrogen compression to 30 MPa at 5.83 kWh/kg [14], which achieves
particularly difficult to provide in processes such as LOHCs since the lower energy consumption per unit hydrogen storage, this technology
reaction condition changed, mainly the dramatically drop of initial H2 has a large number of equipment and a complicated process of obtaining
pressure into the reactor. supercritical media, which increases the burden of fixed costs of the
Conventional Ru/C or Rh/C commercial catalyst have many draw­ compression process, and therefore its application has been limited. In
backs using in the low-pressure hydrogenation process, such as LOHCs, addition, the hydrogenation reaction is an exothermic reaction, and an
in spite of the satisfying performance in the industrial production with exothermic reaction in a high-pressure reactor will additionally raise the
high H2 pressure. First, the hydrogen produced for storage in the air pressure in the reactor, and the safety of the reactor will become a

* Corresponding authors.
E-mail addresses: [email protected] (Z. Ma), [email protected] (X. Gao), [email protected] (S. Yao).

https://doi.org/10.1016/j.fuproc.2024.108143
Received 6 August 2024; Received in revised form 24 September 2024; Accepted 8 October 2024
Available online 15 October 2024
0378-3820/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

severe problem, especially for the LOHCs process where continuous 2.2. Hydrogenation test of single LOHCs
reactor such as fixed bed reactor is mandatory since the potential haz­
ards in the fixed bed are more difficult to control compared to the batch The performance of hydrogenation for single LOHCs was evaluated
reactors, while the gas expansion effect caused by the exotherm of re­ using an E25 autoclave purchased from Beijing Century Senlong
action will be minimized at low H2 pressure. Thus, the research of low- Experimental Device Co. A total of 250 mg of Ru/C or Rh/C catalyst was
pressure hydrogen storage processes for LOHCs is of serious importance activated at 300 ◦ C for 3 h in a stream of hydrogen at 25 mL/min and
for the engineering and marketing of LOHC technology. Furthermore, then added to the reactor along with 20 mL of liquid-phase hydrogen
the most significant advantage of LOHCs technology over traditional storage carrier. The solid reactants NEC and indole were weighed at 2.5
hydrogen storage technology is that it is not constrained by the harsh g and dissolved in decahydronaphthalene and ethanol solvents,
temperature and pressure conditions typically associated with tradi­ respectively, to form a 20 mL substrate solution. The hydrogenation
tional hydrogen storage method such as pressurized and low- reaction temperatures were set at 120 and 160 ◦ C for aromatics and
temperature hydrogen storage technology. Consequently, research on nitrogen-containing heterocycles, respectively. The reaction was me­
low-pressure hydrogen storage is of paramount importance in expand­ chanically stirred at 800 rpm for 2 h in hydrogen gas at 0.8 MPa. The
ing the advantages of this technology and facilitating the commerciali­ reaction was carried out in an air-free environment. After the reaction,
zation and marketability of LOHCs hydrogen storage technology. the liquid phase samples were injected into an Agilent 8860 gas chro­
Despite the significance of the low-pressure hydrogenation reactions matograph. The samples were analyzed by gas chromatography using an
in the industrial application, there is a most significant challenge asso­ HP-5 column.
ciated with low-pressure hydrogen storage technology: the loss of re­ Reactant conversion and turnover frequency (TOF) are used to esti­
action activity and poor conversion rate resulting from a decline, which mate the reaction activity of LOHC and are calculated as follows:
still need further research to solve [15–19]. Wang et al. [20] examined ∑
(ni × yi )
the hydrogenation reaction of furan on the commercial Raney-Ni cata­
Reactant conversion = i × 100%
lyst and concluded that in hydrogen pressure range of 2–10 MPa, the n*y
reaction order with respect to hydrogen is approximately one, which ∑
show a strong influence of reaction rate when the pressure dropped. (ni × yi )
i
Heublein et al. [21] focused on the hydrogenation of NEC in different TOF =
nmetal *t
pressure and temperature and found that when the pressure dropped
from 4 MPa to 1 MPa or lower, the conversion of NEC will decrease from The equation above shows ni and yi as the molar amount and molar
90 % towards 30 % or lower respectively. Furthermore, the detrimental hydrogen storage capacity of each hydrogenation product, respectively.
impact of the diminished hydrogen partial pressure on the reaction is Meanwhile, n and y represent the initial feed amount and the total
also evident in the mass transfer step, Wan et al. [22] estimated the hydrogen storage capacity of the hydrogenation, respectively. nmetal
hydrogenation of NEC at different H2 pressure and concluded that for represents the molar amount of metal in the catalyst, and t represents the
the hydrogenation of NEC, the apparent reaction rate was found to be reaction time. The TOF is not really associated with the active reaction
proportional to the apparent concentration of hydrogen, attributing the site, but all metal atoms are considered to be the active reaction sites.
loss of reaction rate to the drop of adsorbed H2 on the surface of cata­
lysts. However, recently the scope of relevant reaction activity and ki­ 2.3. Hydrogenation test of compounding LOHCs
netic analysis is largely confined to a single carrier, and a comprehensive
comparative study of multiple carriers and compounding systems re­ To evaluate compound hydrogen storage carriers, follow the same
mains limited. methods of product and chromatographic analysis and calculation of
At present, there has been considerable research conducted into the reactant conversion and TOF as for single LOHCs system. The only dif­
hydrogenation of organics under low-pressure and mild conditions from ference is the use of different reactants in the compounded system. For
the perspective of the catalyst itself. For example, studies have explored aromatic quinoline/pyridine mixtures forming liquid phase carriers, 10
the use of improved catalysts obtained by nanoparticles (NPs) method mL of liquid phase nitrogen compounds and an equal mass of aromatic
[23–28] to control the size of catalyst, the addition of other metals as hydrocarbons are used for compounding. In the aromatic mixture, 10 mL
auxiliaries [29,30], bimetallic catalysts [26,31–37] and other related of toluene is mixed with an equal mass of benzene/diphenylmethane,
areas of research [38,39]. Nevertheless, the common limitations of these using the most reactive toluene as the precursor. Hydrogenation eval­
technologies include their complex production process, high production uation of aromatic hydrocarbons-NEC/indole is carried out in dilute
costs, and limited applicability to laboratory-scale preparations. solution by mixing 2.5 g of NEC/indole with an equal mass of the
Furthermore, it is a challenging task to convert the researches into en­ complex and diluting to 20 mL with a suitable solvent to feed the
gineering guidance, which can take a considerable period of time. This reaction.
paper therefore attempts to explore the performance of commercial
catalysts under low-pressure hydrogenation conditions based on com­ 2.4. Characterization of catalyst
mercial catalysts, and to analyze and identify patterns that can be used
to guide the application of commercial catalysts in low-pressure The catalysts are characterized by transmission electron microscopy
hydrogenation. (TEM, JEM-2100F, JEOL, Japan, 200 kV), scanning electron microscope
(SEM, Regulus8100, HITACHI, Japan, 200 kV), powder X-ray diffraction
2. Experiment (XRD, Rigaku SmartLab, Cu Kα radiation), N2 sorption isotherm
(QUADRASORB evo, Anton Paar), inductively coupled plasma (ICP,
2.1. Materials Varian ICP-OES 720) and X-ray photoelectron microscopy (XPS, Thermo
ESCALAB 250XI, Al Kα radiation). For SEM and TEM observation,
Sinopharm Chemical Reagents supplies toluene, benzene, quinoline, samples are dispersed in ethanol then dropped and dried on ultrathin
pyridine, and naphthalene (analytically pure). TCI supplies diphenyl­ carbon membranes for observation. Specific surface area and pore size
methane (analytically pure). Macklin provides N-ethylcarbazole (99 %) are calculated from N2 sorption isotherm using the BET model. For ICP
and Rh/C catalyst (water content 55–60 %). 5 % (dry weight) Ru/C test, the catalysts are dissolved in a mixture of 5 mL HNO3, 3 mL HCl, 1
catalyst and a Rh/C catalyst (with a water content of 55–60 %), both mL HF and 2 mL H2O2, the mixture is heated to 180 ◦ C and kept for 40
purchased from Macklin. min, then the mixture is cooled for the ICP test by using Varian ICP-OES
720. In analyzing the XPS spectra, all binding energies are calibrated by

2
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

setting the C (1 s) peak at 284.4 eV. Table 1

The textural properties of Ru/C and Rh/C catalysts.
Sample S aBET VbP DcP Dispersion Metal Surface
2.5. Adsorption modelling and adsorption energy calculation (m2⋅g− 1) (ml⋅g− 1) (nm) (%) contente metal
percentf
Adsorption modelling of LOHCs with catalyst atoms is carried out
Ru/C 888.4 0.64 6.1 10.5 3.3 % 1.21 %
based on density functional theory (DFT) for each LOHCs on the most Rh/C 933.5 0.57 4.6 8.9 3.4 % 4.51 %
accessible surfaces of the Ru and Rh catalysts, that is the Ru (0001) and
In Table 1, (a) The surface area Measured by BET multi-point method, (b) total
Rh (111) surfaces, and the adsorption energies are also calculated.
pore volume, (c) average diameter of the macropore by BJH method, (d) catalyst
size measured by SEM image, dispersion and mass percentage of the metal
3. Results and analysis element (e) measured by ICP, dry weight content, (f) surface metal percent
measured by XPS are shown.
3.1. The textural properties of the catalysts
pore size distribution of the catalysts is analyzed by BJH method, as
The TEM elemental mapping shown in Fig. 1 reveals the distribution shown in Fig. 3b Around the pore size range of 3–5 nm, there are similar
of metallic elements (Ru and Rh) on the carbon support particles. It can and abundant micropores and mesopores. The pore sizes of Ru/C and
be observed that Rh and Ru atoms on the catalyst surface exhibit a Rh/C catalysts are derived from the active carbon support, and the
particle-like structure on the activated carbon with uniform distribution. similarity in pore size distribution is attributed to the fact that the
The dispersion percentages of noble metals in Rh/C and Ru/C are 8.9 % loading of metal atoms does not block or damage the pores of the
and 10.5 %, respectively (Table 1). The XRD patterns are shown in Fig. 2 catalyst support. However, there is an obvious peak before 2 nm which
The XRD patterns clearly show the characteristic peaks of activated indicates the existence of micropore in the two commercial catalysts. To
carbon [40], while corresponding peaks of metal Ru and metal Rh get more information about the micropore, the D-A micropore analysis
disappear, indicating that the noble metals are uniformly dispersed [35]. results are shown in Fig. 3c. Here, according to the D-A analysis, the
After loading noble metals, the crystallinity of activated carbon is not most accessible micropore diameter of commercial Rh/C and Ru/C
affected. Fig. 3a shows the N2 adsorption-desorption isotherms of the catalysts are 1.54 nm and 1.66 nm respectively. And the pore volume are
catalysts. It can be observed that both Ru/C and Rh/C show type IV 0.386 mL/g and 0.348 mL/g respectively.
isotherms, indicating the presence of a significant amount of mesopores In Table 1, the textural properties of the catalyst used in the
in the catalysts [41]. To further understand the pore information, the

Fig. 1. (a, g) TEM image, (b, h) HAADF-STEM image. EDS element scanning of element (c, i) Ru (Rh) and (d, j) C, (e, k) SEM image, (f, l) EDS area scanning of Ru
(Rh) of the catalyst 5 wt% Ru/C and Rh/C.

3
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

dispersion of the metal atom is both low by the CO dispersion test, we

can infer that the catalyst particles show significant aggregation.
Furthermore, the metal percentage on the surface measured by XPS
method confirmed aggregation on the surface of catalyst. The measured
metal element content by ICP is significantly lower than the nominal
value of the product. It is speculated that the main reason for the devi­
ation in the determination of Ru and Rh is attributed to the inherent
limitations of the ICP method and the difficulty in completely dissolving
the metal elements loaded on the activated carbon.

3.2. XPS spectra of the catalysts

The Ru 3d XPS results of Ru/C catalyst reduction at 400 ◦ C H2/N2 (v/

v = 1:9) for 2 h are shown in Fig. 4. The binding energy values at 308.8
and 313.2 eV in XPS assign to the Rh 3d5/2 and Rh3d3/2 peaks, which are
connected to the oxidation state of Rh, while the value at 307.6 and
312.0 eV assign to the Rh 3d5/2 and Rh3d3/2 peaks, concluding to the
existence of Rh0 [24]. For the Ru 3d XPS result, the existence of C 1 s
peaks result in the superposition of C1s peaks and Ru 3d3/2 peaks, which
are all the value at 283–288 eV, moreover, the value at 283–280 eV
assign to the Ru 3d5/2 peaks [42]. It can be seen that Ru0 and Ru4+ are
mainly present in the reduced Ru/C catalyst, while high oxidation states
Fig. 2. XRD pattern of active carbon (AC, bought from SCR), 5 wt% Rh/C and
5 wt% Ru/C. such as Ru5+ and Ru6+ are hardly visible. Similarly for the Rh, the
mainly state of Rh are Rh0 and Rh4+, no other oxide state can be seen in
the Rh 3d XPS spectra.
experiment are shown. Compared with the Ru/C catalyst, Rh/C catalyst
has a bigger BET surface area, while both of them have a similar pore
diameter. The pore is mainly mesopore on both of the two catalysts, the

Fig. 3. The isotherm linear plot of Ru/C and Rh/C catalyst by BET multi-point method (a), the pore size curve of Ru/C and Rh/C catalyst by BJH desorption method
(b) and micropore analysis by D-A method (c).

Fig. 4. The XPS pattern of Rh/C (a) and Ru/C (b) catalyst after reduction.

4
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

3.3. Hydrogen storage capacity of single LOHCs promote the stability of this catalyst. To get near 100 % conversion, the
ratio of catalyst and reactants is doubled, and the results are shown in
The hydrogenation reaction performance of single LOHCs is shown Fig. 6b. As shown in Fig. 6b, during the test, it can be concluded that the
in the Fig. 5, where the TOF values are the reaction TOF normalized to catalyst can maintain the reaction activity for 400 h despite some small
the number of moles of Ru atoms, rather than the Ru atoms at the active concussions at the beginning.
site, i.e., the apparent TOF values. Selectivity is defined for fully hy­
drogenated products, defined as the molar fraction of fully hydroge­ 3.5. Hydrogen storage capacity of compounding LOHCs
nated products within all hydrogenation products, and for LOHCs (e.g.,
benzene, toluene, and pyridine) where the product is unique and the For the reaction of the compound system, a ratio method is used to
hydrogenation is complete, the selectivity is 100 %. compare the change in the reaction performance of the original sub­
As can be seen from Fig. 5, the hydrogen storage of monocyclic ar­ stance after the addition of the compounded substance, i.e. the ratio of
omatic compounds has a higher hydrogen storage selectivity, which is the hydrogenation TOF after the addition of the compounded substance
related to the fact that there is no stable intermediate in the hydroge­ to that of the pure substance, and the results are shown in Fig. 7. It can
nation path of monocyclic hydrogen storage carriers, but the large be seen that in most of the compound systems the reaction rate of both
π-bonds of the aromatic ring are directly broken to form 6H products. substances is inhibited, while only one of the substances can be inhibited
For thick and polycyclic LOHCs, as a result of the long reaction path in the TOL-DPM system. Therefore, activity reduction is an issue to be
during hydrogenation [43–45], plenty of intermediates are discovered considered when using complexed systems for hydrogen storage.
in the solution after reaction. Therefore, the selectivity of thick and To investigate the dynamics of the compound systems during the
polycyclic LOHCs under low-pressure hydrogenation conditions is hydrogenation reaction, reaction rates of the TOL-DPM and TOL-QUI
apparently poor, and most of the products are partially hydrogenated. systems are examined. Hydrogenation mixtures of the compound sys­
Toluene, benzene and pyridine as monocyclic aromatic groups have tems at different conversion rates are obtained by varying the duration
higher reaction selectivity and are more suitable for low pressure hy­ of the reaction. The hydrogenation rate ratio (normalized to hydrogen
drogenation. In low-pressure hydrogenation reactions, the TOF of LOHC consumption) is used as an indication of the selectivity. The results are
follows the order of monocyclic aromatic hydrocarbons> monocyclic shown in Fig. 8, where a larger ratio means that the reaction has a
nitrogen-containing compounds> polycyclic aromatic hydrocarbons> greater selectivity advantage for the substance represented by the x-axis.
polycyclic aromatic hydrocarbons. Rh/C catalyst and Ru/C catalyst In these examples the conversion of all reactants is low (<30 %).
exhibit different activities in hydrogenation reactions on different sup­ It can be seen that for the TOL-DPM system, the weaker competi­
ports. Rh/C catalysts have stronger catalytic performance for mono­ tiveness keeps the reaction rate ratio of the system at a low level,
cyclic aromatic hydrocarbons, while Ru/C catalyst is more favorable for however, for the highly competitive system where one of the parties has
hydrogenation of polycyclic and fused ring aromatics. It is speculated a clear advantage (e.g., TOL-QUI), as the reaction proceeds, the rate
that the difference in adsorption capacity between reactants and cata­ ratio will not show obviously decrease as the reaction progresses, indi­
lysts is caused by changes in the number of rings and the presence of N cating that quinoline dominates the hydrogenation of the compound
atoms in aromatic compounds, which is related to the type of LOHC and system, and it does not change with the decrease of quinoline concen­
the exposed crystal planes of active metals. tration under lower conversion.

3.4. Stability test towards the Rh/C and Ru/C catalysts 3.6. Analysis and discussions

In order to ensure the conclusion acceptable for the continuous At part 3.3–3.5, plenty of reaction tests are operated to get some
reactor such as fixed bed and fluid bed reactors, the long-time stability principles for LOHCs at the hydrogenation step. And it is concluded that
test towards the catalyst is operated. Toluene is chosen to be the model the compounding LOHCs tend to have a strong competitive effect and
reactant because of its high reaction activity and selectivity. To describe finally resulted in the activity loss of all LOHCs thus the total TOF de­
the activity loss, reaction TOF of toluene in each reaction cycle are creases obviously. Combining the BET results which concentrated on the
drawn in Fig. 6a. surface structure of catalyst. The catalyst with mainly micropore struc­
In Fig. 6, it can be seen that the catalyst conversion can keep at the tures will enhance the energy of catalyst to catch the LOHCs, which
initial level, furthermore, we can conclude that the Rh/C and Ru/C finally caused the adsorption limitation rather than the reaction limit.
catalysts have satisfying stability. Besides, a 200-cycle-long test, where Thus, the strong activity loss and competitive effect can be associated
the reaction keep working for 400 h. to Ru/C test is also conducted to towards the adsorption step of LOHCs, which have been reported by

Fig. 5. Activity and selectivity of different reactants under Rh/C and Ru/C catalytic hydrogenation conditions.
BEN = benzene, TOL = toluene, NAP = naphthalene, DPM = dibenzyltoluene.
PYR = pyridine, QUI = quinoline, NEC = N-ethylcarbazole, IND = indole.

5
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

Fig. 6. The hydrogenation reaction by Ru/C and Rh/C catalysts in 5 reaction cycles (a) and 200-cycle-long reaction stability test of Ru/C catalysts.

Fig. 7. The TOF of different compounding LOHCs under Rh/C (a) and Ru/C (b) catalytic hydrogenation conditions.

Fig. 8. The reaction rate of two LOHCs in the TOL-DPM (a) and TOL-QUI (b) compounding system.

researches focused on the kinetics of hydrogenation of LOHCs [46]. LOHCs. However, the adsorption energy is still negative, which supports
Thus, the DFT models are built to combine the reaction activity with our hypothesis that highly active molecules can spontaneously adsorb
microcosmic behaviors of the adsorption step between LOHCs and on the catalyst surface but will not be strongly bound, promoting the
catalysts. cycle of catalytic hydrogenation at the active site. For most fused and
multi-ring molecules, their adsorption energies are low, indicating that
they are difficult to detach after adsorption and thus occupy the active
3.7. The adsorption model between the reactant and catalysts sites of the catalyst. The main reason for the low reaction rates of NEC,
indole, and naphthalene is attributed to the dilution of the reaction
Using first-principles analysis based on DFT, we modeled the substrate by the solvent. In addition, as the number of carbon atoms in
adsorption of different hydrogen storage molecules on the catalyst sur­ LOHC increases, the adsorption energy from monocyclic aromatic hy­
face, the adsorption energy between catalyst and LOHCs is shown in drocarbons to polycyclic and polycyclic compounds gradually decreases,
Table 2. We can conclude that the adsorption energy of benzene, toluene which is consistent with the trend of single LOHC reaction rate. It is
and pyridine on the Rh catalyst surface is higher compared to other

6
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

Table 2 Table 3
The adsorption energy between the LOHC and commercial catalysts. The adsorption energy of reactants and all hydrogenation products.
LOHC 5 wt% Rh/C 5 wt% Ru/C LOHC 5 wt% Rh/C 5 wt% Ru/C

Adsorption energy TOF Adsorption energy TOF Adsorption energy TOF Adsorption energy TOF
(eV) (h− 1) (eV) (h− 1) (eV) (h− 1) (eV) (h− 1)

BEN − 3.621 480.2 − 3.983 384.0 No intermediate products

TOL − 3.620 639.5 − 4.215 498.5 BEN/CYH − 3.62/− 2.61 480.2 − 3.98/− 2.59 384.0
DPM − 5.074 72.3 − 4.371 135.2 TOL/MCH − 3.62/− 2.67 639.5 − 4.21/− 3.01 498.5
PYR − 3.343 265.4 − 3.853 304.1 PYR/6H-PYR − 3.34/− 2.36 265.4 − 3.85/− 2.89 304.1
QUI − 5.322 173.4 − 6.167 235.7
IND − 4.613 29.3 − 5.180 24.3
One intermediate products
NAP − 5.582 8.8 − 5.623 4.4
DPM (0H-6H- − 5.07/− 4.62/ 72.3 − 4.37/− 5.82/ 135.2
NEC − 8.379 45.0 − 9.846 7.1
12H) − 3.92 − 4.82
QUI (0H-4H- − 5.72/− 4.73/ 173.4 − 6.16/− 5.12/ 235.7
10H) − 3.60- 3.97
concluded that a LOHC with more carbon atoms implies a larger mo­
−
NAP (0H-4H- − 5.58/− 4.38/ 8.8 − 5.62/− 5.37/ 4.4
lecular size and inefficient use of spatial site resistance, intermolecular 10H) − 3.11 − 3.18
repulsion and active sites on the catalyst surface. The low adsorption
energy leads to difficulty in desorption of 0H and intermediates with low More than 1 intermediate products
hydrogen storage capacity, resulting in inhibition of the reaction rate. NEC (0H-4H- − 8.38/− 5.81/ 45.0 − 9.84/− 6.85/ 7.1
The adsorption on Rh/C and Ru/C catalysts is complex, and it is unre­ 8H-12H) − 5.04/− 4.37 − 5.57/− 5.01
liable to directly correlate the adsorption energy on LOHC with the
corresponding reaction rates on different catalysts.
reaction activity of all LOHCs. The polycyclic and fused-ring aromatic
The adsorption models are all shown in Fig. S1, which can explain
hydrocarbons tend to have a lower adsorption energy in the first step
the relation between the adsorption and reaction more directly. For
reaction, therefore, the LOHC with higher reaction activity fail in the
example, in the adsorption model, only one benzene ring of dibenzyl
competition of adsorption step and the reaction activity shows an
methane exhibits strong adsorption to the catalyst. As a result, most of
enormous decrease. In addition, as the degree of LOHC hydrogenation
the hydrogenation product of dibenzyl methane is 6H-dibenzylmethane,
increases, the multi-step hydrogenation intermediates become more
which leads to a low yield of completely hydrogenated product.
difficult to adsorb on the catalyst, resulting in low yields of fully hy­
For the compound LOHCs, we can attribute to the reaction activity
drogenated products.
loss to the contradiction between the high adsorption energy of the
initial reaction and the overall reaction activity. The LOHC with higher
CRediT authorship contribution statement
reaction activity also have a higher adsorption energy in the first reac­
tion step, when the LOHC with high activity compounds and the lower
Xiaopeng Mei: Writing – review & editing, Writing – original draft,
one, it will be more difficult to be adsorbed by catalyst so that be hy­
Formal analysis, Data curation, Conceptualization. Zixuan Ma: Data
drogenated. Besides, the bigger molecular size of some LOHCs will in­
curation, Writing – review & editing, Methodology. Yingjie Yang:
crease the steric hindrance causing the reaction more difficult to
Writing – review & editing. Xiaofeng Gao: Formal analysis, Data
continue.
curation, Conceptualization. Hantao Gong: Data curation, Writing –
To furthermore acquire the adsorption behavior of the LOHCs above
review & editing. Ziyu Song: Writing – review & editing, Data curation.
on the catalyst during the reaction, the adsorption energy of all hydro­
Siyu Yao: Writing – review & editing, Methodology, Investigation,
genation products is also calculated and the results are shown in Table 3.
Funding acquisition, Conceptualization.
As the hydrogenation reaction goes on, the adsorption energy of most
LOHCs is increasing, therefore, the hydrogenation products are more
Declaration of competing interest
easily to desorb from the catalyst, which finally result in the low
selectivity of fully hydrogenated productions, lower conversion and TOF
All authors disclosed no conflict of interests.
result. For some LOHCs such as NEC, the effect of adoption between
LOHCs and catalyst have been reported yet [47]. Specially, for the DPM
Data availability
adsorption on the Ru/C catalyst, the adsorption energy of 6H-DPM
(partially hydrogenated product) and 12H-DPM is lower, with 6H-
No data was used for the research described in the article.
DPM having the lowest energy, which means that the desorption step
of 6H-DPM is the most difficult step, resulting in preventing the
Acknowledgement
adsorption step of 0H-DPM. Therefore, the hydrogenation activity of
DPM on Rh/C catalyst is poor.
This study is financially supported by the National Key Research and
Development Program of China (2022YFB4003100), the Natural Sci­
4. Conclusion ence Foundation of China (22178302), Zhejiang Provincial Natural
Science Foundation of China (LR21B030001), and Beijing National
In this study, the reactivity of different liquid organic hydrogen Laboratory for Molecular Sciences (BNLMS202003). The authors
carriers (LOHCs) on Rh/C and Ru/C was investigated under low pressure gratefully acknowledge Siyu Yao Group of Zhejiang University for
hydrogenation. The different LOHCs exhibited varying levels of reaction providing gas phase chromatography and autoclave equipment.
activity, with monocyclic aromatic hydrocarbons demonstrating the
highest reactivity followed by monocyclic N-heterocyclic aromatic Appendix A. Supplementary data
compounds. Polycyclic and fused-ring aromatic hydrocarbons (aromatic
compounds) displayed relatively lower activity. In addition, the effect of Supplementary data to this article can be found online at https://doi.
LOHC compounding on hydrogenation reaction rate and hydrogen org/10.1016/j.fuproc.2024.108143.
production capacity was investigated by mixing different LOHCs. Due to
the competition of the adsorption step, the compounding of different
LOHCs will lead to a strong competitive effect, resulting in the loss of the

7
X. Mei et al. Fuel Processing Technology 265 (2024) 108143

References [24] G.-Y. Fan, J. Wu, Mild hydrogenation of quinoline to decahydroquinoline over
rhodium nanoparticles entrapped in aluminum oxy-hydroxide, Catal. Commun. 31
(2013) 81–85.
[1] J. Wristers, Strong acid-catalyzed hydrogenation of aromatics, J. Am. Chem. Soc.
[25] E.G. Shubina, N.S. Filimonov, R.V. Shafigulin, et al., Effect of size of nickel
97 (15) (1975) 4312–4316.
nanoparticles on hydrogenation of benzene, Pet. Chem. 57 (5) (2017) 410–414.
[2] N. Kimbara, J.-P. Charland, M.F. Wilson, Hydrogenation of aromatics in synthetic
[26] H. Zhang, Y. Wang, Z. Wang, et al., Synthesis of alumina-supported rhsn alloy
crude distillates catalyzed by platinum supported in molecular sieves, Ind. Eng.
nanocatalysts by using Rh@Sn Core–shell nanoparticle precursors for toluene
Chem. Res. 35 (11) (1996) 3874–3883.
catalytic hydrogenation, Ind. Eng. Chem. Res. 62 (17) (2023) 6626–6633.
[3] I.P. Fisher, M.F. Wilson, Kinetics and thermodynamics of aromatics hydrogenation
[27] G. Zhou, T. Li, J. Chen, et al., Nano-Pd/CeO2 catalysts for hydrogen storage by
in distillates from Athabasca syncrudes, Energy Fuel 2 (4) (1988) 548–555.
reversible benzene hydrogenation/dehydrogenation reactions, Int. J. Hydrog.
[4] A. Bjelić, M. Grilc, M. Huš, et al., Hydrogenation and hydrodeoxygenation of
Energy 46 (27) (2021) 14540–14555.
aromatic lignin monomers over Cu/C, Ni/C, Pd/C, Pt/C, Rh/C and Ru/C catalysts:
[28] Z. Ma, Y. Yang, Z. Song, et al., Fine-tuning Pt nanoparticle and coordination for
Mechanisms, reaction micro-kinetic modelling and quantitative structure-activity
enhanced catalytic efficiency in microwave-assisted methylcyclohexane
relationships, Chem. Eng. J. 359 (2019) 305–320.
dehydrogenation over Pt/Al2O3 catalysts, Fuel 378 (2024) 132851.
[5] A.B. Jain, P.D. Vaidya, Kinetics of hydrogenation of furfuryl alcohol and
[29] P.M.J. Modisha, J. H, A. Bösmann, P. Wasserscheid, D. Bessarabov, Naphthalene
γ-valerolactone over Ru/C catalyst, Energy Fuel 34 (8) (2020) 9963–9970.
hydrogenation over Mg-doped Pt/Al2O3, Int. J. Hydrog. Energy 114 (11) (2018)
[6] I. Bergault, C. Joly-Vuillemin, P. Fouilloux, et al., Modeling of acetophenone
9720–9730.
hydrogenation over a Rh/C catalyst in a slurry airlift reactor, Catal. Today 48 (1)
[30] Y. Wu, H. Yu, Y. Guo, et al., A rare earth hydride supported ruthenium catalyst for
(1999) 161–174.
the hydrogenation of N-heterocycles: boosting the activity via a new hydrogen
[7] A.K. Manal, G.V. Shanbhag, R. Srivastava, Design of a bifunctional catalyst by
transfer path and controlling the stereoselectivity, Chem. Sci. 10 (45) (2019)
alloying Ni with Ru-supported H-beta for selective hydrodeoxygenation of
10459–10465.
bisphenol a and polycarbonate plastic waste, Appl. Catal. B Environ. 338 (2023)
[31] J. Escobar, J.A. Colín-Luna, M.C. Barrera, Thioresistant PdPt/Al/SBA-15 for
123021.
naphthalene hydrogenation, Ind. Eng. Chem. Res. 63 (3) (2024) 1248–1260.
[8] Y. Tian, L. Guo, C. Qiao, et al., Dynamics-driven tailoring of sub-nanometric Pt-Ni
[32] K.J. Li, H.L. An, P.J. Yan, et al., Hydrogenation of toluene to methyl cyclohexane
bimetals confined in hierarchical zeolite for catalytic hydrodeoxygenation, Appl.
over PtRh bimetallic nanoparticle-encaged hollow mesoporous silica catalytic
Catal. B Environ. 336 (2023) 122945.
nanoreactors, ACS Omega 6 (8) (2021) 5846–5855.
[9] Q. Pei, J. Yu, G. Qiu, et al., Fabrication of ultrafine metastable Ru-B alloy for
[33] Y. Qin, X. Bai, N-doped graphitized carbon supported Co@Ru core–shell bimetallic
catalytic hydrogenation of NEC at room temperature, Appl. Catal. B Environ. 336
catalyst for hydrogen storage of N-ethylcarbazole, Cat. Sci. Technol. 12 (9) (2022)
(2023) 122947.
2829–2836.
[10] W. Xue, H. Liu, B. Zhao, et al., Single Rh1Co catalyst enabling reversible
[34] B. Wang, P.Y. Li, Q. Dong, et al., Bimetallic NiCo/AC catalysts with a strong
hydrogenation and dehydrogenation of N-ethylcarbazole for hydrogen storage,
coupling effect for high-efficiency hydrogenation of N-ethylcarbazole, ACS Appl.
Appl. Catal. B Environ. 327 (2023) 122453.
Energy Mater. 6 (3) (2023) 1741–1752.
[11] W. Xue, B. Zhao, H. Liu, et al., Ultralow Pd bimetallic catalysts boost (de)
[35] H. Yu, X. Yang, Y. Wu, et al., Bimetallic Ru-Ni/TiO2 catalysts for hydrogenation of
hydrogenation for reversible H2 storage, Appl. Catal. B Environ. 343 (2024)
N-ethylcarbazole: role of TiO2 crystal structure, J. Energy Chem. 40 (2020)
123574.
188–195.
[12] L. Ge, Y. Zhu, M. Qiu, et al., A highly active Pd clusters hosted by magnesium
[36] L. Zhu, J. Zheng, C. Yu, et al., Effect of the thermal treatment temperature of RuNi
hydroxide nanosheets promoting hydrogen storage, Appl. Catal. B: Environ. Energy
bimetallic nanocatalysts on their catalytic performance for benzene hydrogenation,
333 (2023) 122793.
RSC Adv. 6 (16) (2016) 13110–13119.
[13] J. Guo, L. Xing, Z. Hua, et al., Optimization of compressed hydrogen gas cycling
[37] G. Moretti, W.M.H. Sachtler, Characterization and catalysis of Pt-Cu clusters in
test system based on multi-stage storage and self-pressurized method, Int. J.
NaY, J. Catal. 115 (1) (1989) 205–216.
Hydrog. Energy 41 (36) (2016) 16306–16315.
[38] R. Atsumi, K. Kobayashi, C. Xieli, et al., Effects of steam on toluene hydrogenation
[14] J. Xu, F. Huang, X. Chen, et al., Design and optimization of cryogenic supercritical
over a Ni catalyst, Appl. Catal. A Gen. 590 (2020) 117374.
hydrogen storage system coupled with mixed refrigerant and gas expansion cycle,
[39] S. Furukawa, Y. Matsunami, I. Hamada, et al., Remarkable Enhancement in
Appl. Therm. Eng. 243 (2024) 122606.
Hydrogenation Ability by Phosphidation of Ruthenium: specific Surface Structure
[15] S. Park, M.M. Abdullah, K. Seong, et al., Kinetic analysis of dibenzyltoluene
having Unique Ru Ensembles, ACS Catal. 8 (9) (2018) 8177–8181.
hydrogenation on commercial Ru/Al2O3 catalyst for liquid organic hydrogen
[40] V. Mozhiarasi, T.S. Natarajan, Bael fruit shell–derived activated carbon adsorbent:
carrier, Chem. Eng. J. 474 (2023) 145743.
effect of surface charge of activated carbon and type of pollutants for improved
[16] R.V. Kazantsev, N.A. Gaidai, N.V. Nekrasov, et al., Kinetics of benzene and toluene
adsorption capacity, Biomass Convers. Biorefinery 14 (7) (2024) 8761–8774.
hydrogenation on a Pt/TiO2 catalyst, Kinet. Catal. 44 (4) (2003) 529–535.
[41] V. Gómez-Serrano, C.M. González-Garcia, M.L. González-Martin, Nitrogen
[17] A.K. Neyestanaki, H. Backman, P. Mäki-Arvela, et al., Kinetics and modeling of o-
adsorption isotherms on carbonaceous materials. Comparison of BET and Langmuir
xylene hydrogenation over Pt/γ-Al2O3 catalyst, Chem. Eng. J. 91 (2) (2003)
surface areas, Powder Technol. 116 (1) (2001) 103–108.
271–278.
[42] O. Bezkrovnyi, M. Vorokhta, M. Pawlyta, et al., In situ observation of highly
[18] X. Ye, Y. An, G. Xu, Kinetics of 9-ethylcarbazole hydrogenation over Raney-Ni
oxidized Ru species in Ru/CeO2 catalyst under propane oxidation, J. Mater. Chem.
catalyst for hydrogen storage, J. Alloys Compd. 509 (1) (2011) 152–156.
A 10 (31) (2022) 16675–16684.
[19] F. Sotoodeh, K.J. Smith, Kinetics of hydrogen uptake and release from
[43] J. Ma, L. Yin, L. Ling, et al., The formation of high energy density fuel via the
heteroaromatic compounds for hydrogen storage, Ind. Eng. Chem. Res. 49 (3)
hydrogenation of naphthalene over Ni catalyst: the combined DFT and
(2010) 1018–1026.
microkinetic analysis, Fuel 333 (2023) 126307.
[20] Z.Y. Ren Wang, Shijin Li, Research of furan: kinetics of the hydrogenation of furan,
[44] F.E. Massoth, S.C. Kim, Kinetics of the HDN of Quinoline under Vapor-phase
Chin. Sci. Bull. 9 (14) (1958) 434–435.
Conditions, Ind. Eng. Chem. Res. 42 (5) (2003) 1011–1022.
[21] N. Heublein, M. Stelzner, T. Sattelmayer, Hydrogen storage using liquid organic
[45] F. Sun, Y. An, L. Lei, et al., Identification of the starting reaction position in the
carriers: Equilibrium simulation and dehydrogenation reactor design, Int. J.
hydrogenation of (N-ethyl)carbazole over Raney-Ni, J. Energy Chem. 24 (2) (2015)
Hydrog. Energy 45 (46) (2020) 24902–24916.
219–224.
[22] C. Wan, Y. An, F. Chen, et al., Kinetics of N-ethylcarbazole hydrogenation over a
[46] K.M. Eblagon, K. Tam, S.C.E. Tsang, Comparison of catalytic performance of
supported Ru catalyst for hydrogen storage, Int. J. Hydrog. Energy 38 (17) (2013)
supported ruthenium and rhodium for hydrogenation of 9-ethylcarbazole for
7065–7069.
hydrogen storage applications, Energy Environ. Sci. 5 (9) (2012) 8621–8630.
[23] Y.A. Chudakov, V.D. Stytsenko, G.O. Zasypalov, et al., The influence of ruthenium
[47] K. Morawa Eblagon, K. Tam, K.M.K. Yu, et al., Study of Catalytic Sites on
nanoparticle size on the activity of benzene hydrogenation catalysts, Chem.
Ruthenium for Hydrogenation of N-ethylcarbazole: Implications of Hydrogen
Technol. Fuels Oils 56 (2) (2020) 136–143.
Storage via Reversible Catalytic Hydrogenation, J. Phys. Chem. C 114 (21) (2010)
9720–9730.