Supporting Informaiton

Rational Design of Indium−Palladium Intermetallic Catalysts for Selective

CO2 Hydrogenation to Methanol
Jiabao Lv†,1, Hongman Sun†,1,*, Guanying Liu† Tong Liu†, Guofeng Zhao‡, Youhe Wang†,*, Xin Tu§, Zifeng Yan†

†
State Key Laboratory of Heavy Oil Processing, College of Chemistry and Chemical Engineering, China

University of Petroleum, Qingdao 266580, China

‡
Key Laboratory of Functional Molecular Solids, Ministry of Education, College of Chemistry and Materials

Science, Anhui Normal University, Wuhu 241002, China

§Department
of Electrical Engineering and Electronics, University of Liverpool, Liverpool L69 3GJ, UK

* Corresponding authors: E-mail: hongman.sun@upc.edu.cn (H. Sun); yhewang@upc.edu.cn (Y. Wang)

1
These authors contributed equally to this work
Experimental Section

Chemicals

ZrOCl2·8H2O (>98.0%), Pd(NO3)2·xH2O (>99.0%) and SiO2 (10 μm) were purchased from

Aladdin. In(NO3)3·5H2O (>99.9%) was obtained from Macklin. NH3·H2O (25.0%~28.0%) was

obtained from Sinopharm Chemical Reagent Co., Ltd. All the chemicals did not require further

purification.

Preparation of catalysts

The m-ZrO2 support was prepared by a hydrothermal method.1 Specifically, a 0.5 mol/L solution

of ZrOCl2·8H2O was prepared using deionized water and maintained in a Teflon-lined hydrothermal

kettle at 150 °C for 8 h. The resulting emulsion was adjusted to a pH of 8 by NH3·H2O. The white

precipitate was centrifuged repeatedly until Cl- could not be detected using AgNO3/HNO3 solution

(1.0 mol/L). Next, the sample was dried at 110 °C for 12 h in an oven. Finally, m-ZrO2 was obtained

by calcination at 350 °C (5 °C/min) for 4 h.

InPd/m-ZrO2 samples with different molar ratios were prepared by wetness impregnation. Taking

a typical InPd(2:1)/m-ZrO2 mixture as an example, 0.1532 g of In(NO3)3·5H2O and 0.0555 g of

Pd(NO3)2·xH2O were dissolved in 1 mL of deionized water, followed by the addition of 0.5 g of m-

ZrO2. The sample was stirred sufficiently for 1 h and then transferred to an oven to dry at 100 °C

overnight. The dried sample was calcined in a muffle furnace at 300 °C for 3 h to obtain InPd(2:1)/m-

ZrO2. Keeping the indium loading in the sample constant at 9 wt.%, the amount of Pd was varied to

prepare catalysts with different ratios (3:1, 2:1, 1:1, 1:2), and the catalysts were named according to

the ratio of In to Pd. For catalysts with pure indium or palladium, the dosage was kept the same as

that for InPd(2:1)/m-ZrO2, and the catalysts were named In/m-ZrO2 or Pd/m-ZrO2. The same strategy
was used to prepare InPd(2:1)/SiO2, except that m-ZrO2 was replaced by SiO2. All the reduced

catalysts were obtained by maintaining them in a 10% H2/N2 mixed atmosphere (50 mL/min) at

300 °C for 1 h.

The In3Pd2/m-ZrO2 catalyst was obtained by implementing an additional acid-washing step.

Specifically, the reduced InPd(2:1)/m-ZrO2 and 2 mol/L H2SO4 were mixed according to a mass ratio

1:10 and kept at 90 °C for 90 minutes. Next, the sample was centrifuged and vacuum dried at 80 °C

for 12 hours to obtain In3Pd2/m-ZrO2. The same strategy was used for InPd(2:1)/SiO2 to prepare

In3Pd2/SiO2.

Material characterizations

Powder X-ray diffraction (XRD) patterns were obtained on an Empyrean diffractometer with Cu

Kα radiation. Continuous scans were collected between 2θ = 5 and 75°. The specific surface area of

the catalysts was obtained from nitrogen adsorption-desorption isotherms using a Tristar-3020 at -

196 °C (Micromeritics, USA). The morphologies of the catalysts were examined using a JEOL JSM-

7900F scanning electron microscope (SEM) at 3 kV. High-resolution transmission electron

microscopy (HRTEM) and high-angle annular dark-field scanning TEM (HAADF-STEM) images

were obtained on a Talos F200X electron microscope (Thermo Fisher, America). H2-TPR was tested

by an Xianquan TP-5080-D chemisorber. Before the test, 100 mg 20-40 mesh samples were pretreated

for 1 h at 300 °C in an Ar atmosphere. After cooling to room temperature, the Ar atmosphere was

switched to 5% H2/Ar (30 mL/min), the mixture was heated to 500 °C at a heating rate of 10 °C/min,

and the H2 content was detected by a thermal conductivity detector (TCD). The H2-TPD test was

conducted on a Builder AMI-300 (Germany). First, 100 mg of sample (40-60 mesh) was pretreated
at 300 °C under a He atmosphere for 1 h. After cooling to room temperature, 10% H2/He (30 mL/min)

was introduced and fully adsorbed for 60 min. After the He was purged to a stable baseline, the sample

was heated to 600 °C at a heating rate of 10 °C/min and the H2 content was detected by TCD. The

CO2-TPD characterization was performed on a Bayer BELCAT-B. 50 mg of sample (40-60 mesh)

was activated at 300 °C for 1 h in a 10% H2/N2 atmosphere. After cooling to room temperature, CO2

(30 mL/min) was introduced and maintained for 60 min. After sufficient adsorption, the sample was

purged with N2 until the baseline was stable. Finally, the temperature was ramped up to 500 °C at a

rate of 10 °C/min, after which the CO2 content was detected by TCD. The X-ray photoelectron

spectroscopy (XPS) spectra were measured by an Escalab 250Xi (Thermo Fisher Scientific)

instrument under Al Kα radiation, and the binding energy was calibrated by measuring the C 1s peak

at 284.8 eV. Inductively coupled plasma emission spectrometry (ICP-OES, Agilent 7700) was used

to determine the content of indium and palladium in the catalysts.

In situ diffuse reflectance Fourier transform spectroscopy (in situ DRIFTS) was performed on a

Bruker VERTEX 70V. We adopted two strategies. In strategy I, all the catalysts were pretreated at

300 °C and 10% H2/N2 (50 mL/min) for 1 h, after which the background information was collected

after cooling to 150 °C. Then, the atmosphere was changed to CO2/H2 (1:4, 10 mL/min) + N2 (10

mL/min), and the amount of adsorbed substances during CO2 hydrogenation was recorded. The

temperature range was 150−330 °C, and the temperature interval was 20 °C. At each temperature, the

mixture was kept for 10 min. The same preprocessing procedure was used for strategy II, but the

signal acquisition process was carried out at 270 °C. First, the catalyst was kept at 30% CO2/N2 (10

mL/min) for 30 min to allow the catalyst to fully adsorb CO2. Then, N2 (10 mL/min) was passed for

5 min to remove the weakly adsorbed CO2 from the surface of the catalyst. Finally, the environment
was shifted to 10% H2/N2 (10 mL/min), and the catalyst was kept in this atmosphere for 30 min to

analyze the species transformation process.

Catalytic performance of CO2 hydrogenation

The performance of the catalysts for CO2 hydrogenation was evaluated in a fixed bed reactor. First,

0.15 g of the catalyst (40-60 mesh) was mixed with 0.35 g of inert quartz sand (25-50 mesh) and then

put in the reactor. The catalyst is subjected to in situ reduction prior to the reaction (see the Preparation

of catalysts for details of the reduction conditions). After the temperature was reduced to the reaction

temperature, the atmosphere was switched to a H2/CO2 mixture (H2/CO2 = 3:1), the pressure was

increased to 4 MPa, and the GHSV = 8000 mL/gcat/h. The temperature evaluation range was

230−310 °C, with a temperature interval of 20 °C. An online analysis of the product gas stream was

performed via gas chromatography (GC, SCION, 456C) equipped with a thermal conductivity

detector (CH3OH, CO, CH4, and CO2). CO2 conversion (XCO2) and CH3OH selectivity (SCH3OH) were

calculated as follows:
f𝐶𝑂2 A𝐶𝑂2,out
𝑋𝐶𝑂2 = 1 − × 100%, (1)
f𝐶𝑂2 A𝐶𝑂2,in
fCH3OH ACH3OH，out
𝑆𝐶𝐻3𝑂𝐻 = × 100%, (2)
f𝐶𝑂2 A𝐶𝑂2,in − f𝐶𝑂2 A𝐶𝑂2,out

where f and A represent the chromatographic peak area and the correction factor for the

corresponding substance, respectively.

The formula for calculating the STY of CH3OH is as follows:

𝜇𝐶𝑂2 × 𝑋𝐶𝑂2 × 𝑆𝐶𝐻3𝑂𝐻 × 𝑀𝐶𝐻3𝑂𝐻
𝑆𝑇𝑌 = , (3)
𝑚𝑐𝑎𝑡 × 𝑉𝑚

where μCO2 is the flow rate of CO2 (L/h), XCO2 is the CO2 conversion, SCH3OH is the CH3OH

selectivity, MCH3OH represents the molecular weight of CH3OH (32.04 g/mol), mcat is the mass of the

catalyst (g), and Vm is the molar volume of CO2 under standard conditions (22.4 L/mol).
Density function theory calculations

Density function theory (DFT) calculations were performed with the Vienna ab initio simulation

package (VASP) employing the Perdew–Burke–Ernzerhof (PBE) density functional. Core electrons

were represented with projector augmented-wave (PAW) pseudopotentials, while valence electrons

were described by plane waves with a kinetic cutoff energy of 520 eV. The bulk lattice parameters of

m-ZrO2, In2O3, Pd, and In3Pd2 were optimized with a kinetic energy cutoff of at least 700 eV. One

outermost layer was allowed to relax, while the bottommost layer was kept fixed, mimicking bulk

positions. A vacuum region of 20 Å was added between slabs, and we applied a dipole correction

along the z axes to account for the asymmetry in the relaxations.To ensure the validity of DFT

calculation results, we constructed the models based on the characterization results adhering to the

following principles:

(ⅰ) For each component, we selected the most stable bulk and the experimentally characterized

surface structure as the foundation for the models. m-ZrO2 and Pd are represented by their most stable

terminations (−111) and (111),2 respectively. In the TEM analysis, the In2O3 (222) facet was identified

(Figure 1e), and thus In2O3 is represented by the (222) termination. For In3Pd2, the (102) facet was

confirmed in both XRD and TEM (Figure 1c,e). Thus, the (102) facet was identified as the termination

of In3Pd2.

(ii) To better express the interaction between In3Pd2 and In2O3, we place the clusters relatively close

to each other on the substrate. HRTEM imaging revealed that these two phases are closely positioned

(Figure 1e). Therefore, when modeling In3Pd2−In2O3, we aimed to maintain the distance between the

two phases as close as possible to reflect actual observations. At the same time, we strictly avoided

the phenomenon of oxygen escape.

 The energy (ΔE) of adsorption is defined as follows:

E = EAB − EA − EB (4)

Where EAB represents the overall energy, EA is the energy of the support and active components,

and EB is the energy of the adsorbed molecules.The Gibbs free energy (ΔG) correction method is as

follows:

G = E + EZPE - TS (5)

Where E is the reaction energy difference between reactants and products of each reaction; EZPE

and S is the change in zero point energy and entropy for each reaction. The zero point energy and

entropy of adsorbates and free molecules were obtained from the vibrational frequency calculations.
Figure S1. XRD patterns of In3Pd2/m-ZrO2 and In3Pd2/SiO2.

Figure S2. N2 adsorption−desorption isotherms (a) and BJH pore size distributions (b) of

indium−palladium intermetallic catalysts.

Figure S3. SEM images of different ratios of indium−palladium intermetallic catalysts.
Figure S4. (a-c) CO2 hydrogenation performance of indium−palladium intermetallic catalysts

(reaction conditions: T = 230−310 oC, P = 4 MPa, H2/CO2 = 3, GHSV = 8000 mL/gcat/h). (d) CH3OH

STY versus time in stability test of InPd(2:1)/m-ZrO2 at 270 oC.

Figure S5. CO2 hydrogenation performance of In3Pd2/m-ZrO2 (reaction conditions: T = 230−310 °C,

P = 4 MPa, H2/CO2=3, GHSV = 8000 mL/gcat/h).

Figure S6. DFT calculations of InPd(2:1)/m-ZrO2 models for intermediates. Color code of DFT

models: light blue (Zr), red (O), brown (Pd), yellow (In), grey (C), light green (H).
Figure S7. DFT calculations of indium−palladium intermetallic catalyst models of CO2 adsorption

by different components. Color code of DFT models: light blue (Zr), red (O), brown (Pd), yellow (In),

grey (C).
Figure S8. DFT calculations of indium−palladium intermetallic catalyst models of CO adsorption by

different components. Color code of DFT models: light blue (Zr), red (O), brown (Pd), yellow (In),

grey (C).
Figure S9. DFT calculations of indium−palladium intermetallic catalyst models of H2 adsorption by

different components. Color code of DFT models: light blue (Zr), red (O), brown (Pd), yellow (In),

grey (C), light green (H).

Table S1. Textural characteristics and components of indium−palladium intermetallic catalysts.

Samples SBET/(m2·g-1)a In(wt.%)b Pd(wt.%)b

m-ZrO2 162 / /

In/m-ZrO2 110 7.9 /

InPd(3:1)/m-ZrO2 119 / /

InPd(2:1)/m-ZrO2 117 7.9 3.0

InPd(1:1)/m-ZrO2 112 7.7 5

InPd(1:2)/m-ZrO2 114 / /

Pd/m-ZrO2 153 / 2.7

a
Based on BET data.

b
Valued by ICP-OES.
Table S2. XPS results of m-ZrO2 and indium−palladium intermetallic catalysts.

Ov/(Ov+OL) In2O3 Contents

Samples
(%) (%)
m-ZrO2 24.9 /

In/m-ZrO2 36.9 79.4

InPd(2:1)/m-ZrO2 38.3 71.3

Pd/m-ZrO2 35.5 /
Table S3. Free energy of intermediates in CO2 hydrogenation of InPd(2:1)/m-ZrO2.

HCOO* path Free energy (eV) RWGS+CO path Free energy (eV)

CO2* 2.05 CO2* 2.05

HCOO* 0.64 COOH* 1.06

HCOOH* 2.04 CO*+OH* 2.88

H2COOH* 1.66 CO*+H2O* 2.16

Table S4. Adsorption energy of reactants in different components of indium−palladium intermetallic

catalysts.

Eads (eV)
Components
CO2 CO H2

m-ZrO2 −0.36 −0.02 −0.01

Pd/m-ZrO2 −0.04 −2.60 −0.63

In3Pd2/m-ZrO2 −0.19 −1.20 −0.13

In2O3/m-ZrO2 −0.57 −0.06 −0.05

In3Pd2−In2O3/m-ZrO2 −0.50 −0.22 −0.18

REFERENCES

(1) Zheng, X. H.; Yu, P. J.; Liu, Y. X.; Ma, Y. D.; Cao, Y. N.; Cai, Z. P.; Zhou, L. S.;

Huang, K.; Zheng, S. T.; Jiang, L. L. Efficient Hydrogenation of Methyl Palmitate to

Hexadecanol over Cu/m-ZrO2 Catalysts: Synergistic Effect of Cu Species and Oxygen Vacancies.

ACS Catal. 2023, 13 (3), 2047-2060.

(2) Araújo, T. P.; Morales-Vidal, J.; Giannakakis, G.; Mondelli, C.; Eliasson, H.; Erni, R.;

Stewart, J. A.; Mitchell, S.; López, N.; Pérez-Ramírez, J. Reaction-Induced Metal-Metal Oxide

Interactions in Pd-In2O3/ZrO2 Catalysts Drive Selective and Stable CO2 Hydrogenation to

Methanol. Angew. Chem., Int. Ed. 2023, 62 (42), e202306563.