Science 14Feb2020

SCIENCE sciencemag.org

GRAPHIC: A. KITTERMAN/

SCIENCE

By Liyu Chen^1 and Qiang Xu1,2

S

urface defects of nanomaterials can

serve as active sites for adsorption

and chemical transformations in het-

erogeneous catalysis ( 1 , 2 ). However,

the defects in catalyst supports can

also induce carbon deposition to de-

activate the catalysts. This issue is particu-

larly relevant for supported metal catalysts,

a major category of heterogeneous catalysts,

which deactivate because of the formation of

carbon-based materials on catalyst surfaces

after prolonged use, through a process called

coking ( 3 ). Developing supported metal cata-

lysts with coking and sintering resistance

with high catalytic activity in high-tempera-

ture applications remains a great challenge

( 4 ). On page 777 of this issue, Song et al. ( 5 )

address this critical issue by choosing a de-

fect-free single-crystalline magnesium oxide

(MgO) as a support and then blocking the

active step edges with nickel-molybdenum

(Ni–Mo) nanocatalysts, achieving coke- and

sintering-resistant activity in quantitative

production of synthesis gas from dry reform-

ing of methane (CH 4 ).

The conversion of carbon dioxide (CO 2 )

into value-added products can not only con-

trol excess CO 2 emissions but also decrease

the depletion of fossil resources toward a

more sustainable energy economy ( 6 ). CO 2

conversion through dry reforming with

methane (DRM) can produce synthesis gas

for the synthesis of hydrogen, methanol, and

various hydrocarbon-based fuels. The DRM

is a strongly endothermic reaction with op-

erating temperatures of 800° to 1000°C,

which requires heterogeneous catalysts to

convert CH 4 and CO 2 effectively. Numerous

supported metal catalysts have been investi-

gated, in which Ni-based catalysts supported

on oxides [such as MgO or zirconium diox-

ide (ZrO 2 )] have attracted great attention

because of their low cost and high activity

( 7 ). However, deactivation of the supported

Ni catalysts in reforming reactions owing

to coke formation and metal sintering have

been widely reported.

The defects of supports and the sizes and

compositions of Ni-based nanoparticles can

influence coke formation and metal sintering

( 8 ). Intensive efforts have been made to de-

sign supported Ni catalysts with coking and

sintering resistance to minimize catalyst de-

activation. Coke formation is often addressed

by passivating the active sites with traces of

sulfur and phosphites ( 9 ). The sintering of

metal nanoparticles at high temperatures

can be prevented through steric stabilization

by an overlayer of inorganic oxides ( 10 ). How-

ever, these methods often lead to a compro-

mise of the inherent activity.

To develop an efficient dry reforming cata-

lyst based on Ni/MgO, Song et al. selected a

highly crystalline MgO solid as a support on

the hypothesis that a less crystalline MgO

with defects could result in carbon deposition

on the support (see the figure). The authors

prepared the single-crystalline MgO through

reduction of CO 2 with Mg chips through an

autothermal reaction. Subsequently, they

used a polyol-mediated reductive growth

method with hydrazine as a reducing agent

and polyvinylpyrrolidone (PVP) as a surfac-

tant to load metal nanoparticles. Ni salt and

Mo salt were simultaneously mixed to load

Ni–Mo on MgO (denoted as NiMoCat), on

the consideration that bimetallic nanopar-

ticles may have a synergistic effect between

the two metals.

The NiMoCat the authors prepared exhib-

ited high activity and durability under dry

reforming conditions, giving a steady quan-

titative conversion of CH 4 and CO 2 for up to

850 hours. NiMoCat showed no detectable

coking under the reactive gas flow, which

helps explain the impressive activity. NiMo-

Cat showed far superior activity and durabil-

ity compared with those of many industrial

and research catalysts. In the control experi-

ments, using the same polyol synthesis for

Ni–Mo nanoparticles but commercial MgO

as a support, the authors observed severe

coke formation and lower conversion yields.

Using a wet impregnation for synthesis of

Ni–Mo nanoparticles on crystalline MgO

also resulted in heavy coking. MgO and Ni–

Mo nanoaparticles alone showed deactiva-

tion under the reaction condition because of

coke formation.

For NiMoCat, the as-synthesized particles

with an average size of 2.88 nm grow and

form Mo–Ni alloys with an average size of

17.30 nm and then are stabilized on the high-

energy step edges of the 111 MgO crystal lat-

tice plane under the reactive gas flow. The

stabilization of Ni–Mo nanoparticles on the

high-energy step edges of MgO can not only

prevent further sintering but also eliminate

the active sites of MgO to prevent carbona-

ceous deposition that deactivates the catalyst.

Song et al. call this phenomenon the Nano-

catalysts On Single Crystal Edges (NOSCE)

technique. To verify the NOSCE behavior,

NiMoCat was crushed through ball milling

to expose new step edges of MgO. The ball-

milled NiMoCat showed severe coking under

the same reaction condition, proving that the

step edges of MgO are active sites for carbon

deposition. In addition, the Mo doping is crit-

ical for high activity and stability. Mo doping

can enhance the oxidative stability of Ni and

facilitate the mobility of Ni particles to move

to the step edges of MgO.

The development of supported metal

catalysts with high activity and durability in

high-temperature applications is of vital im-

portance to the chemical industry. Defects in

catalyst support can have a passive effect to

accumulate carbon deposition that deacti-

vates the catalyst. The NOSCE technique can

eliminate the support in the catalytic reac-

tion to prevent carbon deposition, leading to

stable catalysts with coke- and sintering-re-

sistant activity. This addresses the long-term

challenge in catalysis science. This discovery

will likely drive the rapid development of ac-

tive and stable nanocatalysts for many chal-

lenging reactions, which may pave the way

for industrial application. j

REFERENCES AND NOTES

1. M. Behrens et al., Science 336 , 893 (2012).


2. N. Tsumori et al., Chem 4 , 845 (2018).


3. C. H. Bartholomew, R. J. Farrauto, Fundamentals of
   Industrial Catalytic Processes (Wiley-VCH, 2011).


4. Q.-L. Zhu, Q. Xu, Chem 1 , 220 (2016).


5. Y. Song et al., Science 367 , 777 (2020).


6. J. Artz et al., Chem. Rev. 118 , 434 (2018).


7. S. Li, J. Gong, Chem. Soc. Rev. 43 , 7245 (2014).


8. S. De et al., Energy Environ. Sci. 9 , 3314 (2016).


9. B. Abdullah et al., J. Clean. Prod. 162 , 170 (2017).


10. S. Das et al., Appl. Catal. B 230 , 220 (2018).
    ACKNOWLEDGMENTS
    We thank AIST and the National Natural Science Foundation of
    China (NSFC-21875207) for financial support.
    10.1126/science.aba6435

CATALYSIS

Fewer defects, better catalysis?

Defect-free magnesium oxide provides a better route

for carbon dioxide conversion

(^1) AIST-Kyoto University Chemical Energy Materials Open
Innovation Laboratory (ChEM-OIL), National Institute of
Advanced Industrial Science and Technology (AIST), Yoshida,
Sakyo-ku, Kyoto 606-8501, Japan.^2 School of Chemistry and
Chemical Engineering, Yangzhou University, Yangzhou 225002,
China. Email: [email protected]; [email protected]
Ni–Mo catalyst
MgO sheet
MgO crystal
Coking
Dry reforming
Dry reforming
14 FEBRUARY 2020 • VOL 367 ISSUE 6479 737
Catalysis without the coking
Defect-free MgO crystals (bottom) avoid the
reaction-killing carbon buildup from MgO sheets (top).
Published by AAAS