DOI: 10.1002/cssc.
201501498 Full Papers
Ammonia Decomposition with Manganese Nitride–Calcium
Imide Composites as Efficient Catalysts
Pei Yu,[a, b] Jianping Guo,[a] Lin Liu,[a] Peikun Wang,[a, b] Guotao Wu,[a] Fei Chang,[a, b] and
Ping Chen*[a, c, d]
Ammonia has high gravimetric and volumetric hydrogen den- 11CaNH composite catalyst is about an order of magnitude
sities and is, therefore, considered a promising carrier for the higher than that of Mn6N5 and comparable to the highly active
production of COx-free molecular H2 for forthcoming energy Ni/SBA-15 and Ru/Al2O3 catalysts. Analysis by means of temper-
systems. Alkaline earth metals are generally regarded as struc- ature-programmed decomposition (TPD), X-ray diffraction
tural promoters of catalysts and employed in numerous cata- (XRD), and X-ray absorption near edge spectroscopy (XANES)
lytic processes. Here, we report that calcium imide (CaNH) has reveal that CaNH participates in the catalysis via forming
a strong synergistic effect on Mn6N5 in catalyzing the decom- a [Ca6MnN5]-like intermediate, thus altering the reaction path-
position of NH3, leading to a ca. 40 % drop in apparent activa- way and energetics. A two-step catalytic cycle, accounting for
tion energy. At 773 K, the H2 formation rate over a Mn6N5- the synergy between CaNH and Mn6N5, is proposed.
Introduction
Molecular hydrogen generated from carbonaceous substances by using specific adsorbers (e.g., Calgon-URC, Grace Davison
(e.g., methanol and methane) inevitably contains traces of COx Grade 514).
(x = 1, 2), which degrades cell electrodes even at extremely low Up to now, transition metals, metal carbides, and metal ni-
concentrations.[1] Ammonia (NH3), with high gravimetric trides have been reported as catalytically active for NH3 de-
(17.8 wt %) and volumetric (121 kgH2 m¢3 in liquid form) H2 den- composition.[3] Detailed studies on the catalytic activities of
sities and an annual production volume of over a hundred various metals (Ru, Ir, Ni, Rh, Pd, Pt, Fe)[4] on different supports
megatons, is therefore considered a promising energy carrier (SiO2, zeolite, TiO2, Al2O3, MgO, superbase, activated carbon,
for the production of COx-free hydrogen to mediate forthcom- carbon nanotubes, nitrogen-modified carbon, electrides)[4c, 5]
ing hydrogen energy systems.[1, 2] Although ammonia is corro- have revealed that alkali-metal-promoted ruthenium on carbon
sive and harms proton exchange membrane fuel cells, the nanotubes (Ru/CNTs) possesses the highest activity. However,
level of unconverted NH3 can be reduced to less than 200 ppb the high cost of ruthenium hinders its practical applications,
and the development of highly active non-noble-metal cata-
[a] P. Yu, Dr. J. Guo, Dr. L. Liu, P. Wang, Dr. G. Wu, F. Chang, Prof. P. Chen lysts is urgently needed.[6]
Dalian National Laboratory for Clean Energy Transition metal nitrides are gaining increasing attention
Dalian Institute of Chemical Physics
Chinese Academy of Sciences
owing to their noble-metal-like catalytic properties in hydrode-
Dalian 116023 (PR China) nitrogenation (HDN), hydrodesulfurization (HDS), Fischer–
Fax: (+ 86) 411-84379583 Tropsch (F–T), and ammonia synthesis chemistries.[7] Neat
E-mail: [email protected] binary 3d transition metal nitrides have limited activities to-
[b] P. Yu, P. Wang, F. Chang wards ammonia decomposition,[8] although increasing the sur-
University of Chinese Academy of Sciences
Beijing 100049 (PR China) face area of VN,[9] MoNx,[10] and Fe3N[11] leads to improved cata-
[c] Prof. P. Chen lytic activities. Ternary nitrides, such as Co–Mo–N and Ni–Mo–
State Key Laboratory of Catalysis N, containing elements from the right and left sides of the vol-
Dalian Institute of Chemical Physics cano curve and having optimal nitrogen binding energies,[12]
Chinese Academy of sciences
exhibit better performances towards NH3 decomposition.[13]
Dalian 116023 (PR China)
Recently, Guo et al. found that TiN, VN, CrN, and MnN, which
[d] Prof. P. Chen
Collaborative Innovation Center of Chemistry for Energy Materials have very limited activities in neat form, exhibit extraordinarily
Dalian Institute of Chemical Physics high activities upon forming composites with Li2NH. In particu-
Chinese Academy of Sciences lar, the hydrogen production rate of a MnN–Li2NH catalyst
Dalian 116023 (PR China)
proved superior to that of a Ru/CNT catalyst.[6c] Notably, unlike
Supporting Information for this article is available on the WWW under
potassium and cesium, lithium oxide or hydroxide does not ex-
http://dx.doi.org/10.1002/cssc.201501498.
hibit significant electronic promotion to transition metal cata-
This publication is part of a Special Issue on the “Chemistry for Energy
Conversion and Storage” conference in Berlin, Germany. To view the lysts in NH3 decomposition.[5b, 14] However, when lithium is in
complete issue visit: http://dx.doi.org/10.1002/cssc.v9.4. the form of Li2NH, its role in enhancing catalytic activity is pro-
ChemSusChem 2016, 9, 364 – 369 364 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1864564x, 2016, 4, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cssc.201501498 by Sichuan University, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Full Papers
nounced. Experimental data show that Li2NH is an NH3-trans- evidenced by X-ray absorption near edge spectroscopy
mitting agent and synergizes with 3d nitrides, creating (XANES), shown in Figure 6. A Mn6N5-CaO catalyst, as a refer-
a unique catalytic pathway.[6c] Hence, we are interested in in- ence, was synthesized by ball-milling Mn6N5 with CaO. The
vestigating whether its alkaline earth analogues, such as CaNH, manganese contents in all catalysts were determined by induc-
can function similarly in ammonia decomposition. tive coupled plasma–atomic emission spectroscopy (ICP–AES).
Alkaline earth metal oxides (i.e., MgO, CaO, SrO, and BaO) The apparent activation energies of the samples were mea-
have often been used as supports or structural promoters to sured in a flow of pure NH3, as shown in Table 1. In addition,
enhance the dispersion of active sites of catalysts. In particular, the results of N2 physisorption evaluated by the Brunauer–
CaO has been employed in the catalytic synthesis of methane Emmett–Teller (BET) method show that the specific surface
through CO2 hydrogenation,[15] methane–carbon dioxide re- area of Mn6N5-15CaNH-1 (27.8 m2 g¢1) is comparable to that of
forming,[16] and ammonia synthesis processes. In industrial neat Mn6N5 (28.8 m2 g¢1).
fused-iron ammonia synthesis catalysts, CaO is believed to pre- Depicted in Figure 1 is the temperature dependence of NH3
vent iron from sintering by avoiding direct contact between conversions over neat CaNH and manganese-based catalysts.
the iron particles. Lendzion-Bielun et al. found that calcium
and aluminum incorporate into a magnetite particle and form
a (CaOAl2O3)-Al-phase, which can prevent iron from aggrega-
tion.[17]
The calcium and nitrogen atoms in CaNH are arranged in
the motif of the NaCl-type structure; hydrogen atoms of the
imide groups are disordered within the calcium octahedra, and
occupy the six-fold split positions.[18] CaNH has been investigat-
ed as a hydrogen storage candidate,[19] but no attention was
given to its catalytic effects in NH3 decomposition.
Herein, we found that CaNH can exert a strong synergistic
effect on Mn6N5, leading to greatly enhanced catalytic activity
that is comparable to the highly active Ni/SBA-15 and Ru/Al2O3
catalysts. Our experimental results reveal that CaNH is unlikely
to be the promoter but rather participates in the catalytic reac- Figure 1. Temperature dependence of NH3 conversions over CaNH and Mn-
based catalysts. Reaction conditions: sample loading 50 mg, flow rate of
tion by forming a [Ca6MnN5]-like intermediate, thereby altering
pure NH3 30 mL min¢1.
the reaction pathway and energetics. A two-step catalytic cycle
is proposed, to account for the synergy between CaNH and
Mn6N5. Neat CaNH and Mn6N5 show negligible catalytic activities at
temperatures below 773 K. Mn6N5-18CaO also exhibits a very
low activity. However, there is a remarkable increase in catalytic
Results and Discussion activity for the Mn6N5-15CaNH-1 composite catalyst.
To demonstrate the effect of CaNH on the catalytic per- Mn6N5-15CaNH-1 can catalyze NH3 decomposition at tempera-
formance of Mn6N5, composite catalysts were prepared with tures just above 673 K, which is ca. 150 K lower compared to
different Mn6N5/CaNH molar ratios and by different synthetic Mn6N5 and CaNH. At ca. 800 K, its activity is ca. 33 times that of
methods. Table 1 shows the properties of catalysts employed CaNH and 10 times that of Mn6N5. The Mn6N5-11CaNH-2 and
in this study. The catalyst Mn6N5-CaNH-1 was synthesized by Mn6N5-111CaNH-2 composite catalysts, prepared by ball-milling
ball-milling Mn6N5 with CaNH. The Mn6N5-CaNH-2 catalysts MnCl2 with CaNH, feature better contact between Mn6N5 and
were made by ball-milling MnCl2 with CaNH. During this treat- CaNH than Mn6N5-15CaNH-1. As expected, NH3 conversions
ment the chemical reaction 6 MnCl2 + 6 CaNH!Mn6N5 + over Mn6N5-111CaNH-2 and Mn6N5-11CaNH-2 catalysts were
6 CaCl2 + NH3 + 3/2 H2 took place. The formation of Mn6N5 was higher (Figure 1). The apparent activation energies of Mn6N5-
15CaNH-1 and Mn6N5-11CaNH-2 are about 85 kJ mol¢1. These
values are substantially lower than that of neat Mn6N5
Table 1. Properties of catalysts employed in ammonia decomposition (138 kJ mol¢1, see Figure 2). These results imply that the nature
of the active sites of Mn6N5-CaNH composite catalysts is differ-
Catalyst Components Loading (gMn/gcat.)[a] Ea[b]
[%] [kJ mol¢1] ent from those of pure Mn6N5.
To get the optimal H2 formation rate, we carried out several
Mn6N5-15CaNH-1 Mn6N5, CaNH 26.80 84
Mn6N5-11CaNH-2 Mn6N5, CaNH, CaCl2 20.00 86 experiments separately by varying NH3 flow rates while main-
Mn6N5-111CaNH-2 Mn6N5, CaNH, CaCl2 4.58 76 taining a catalyst loading of 50 mg. Figure 3 shows the activi-
Mn6N5-18CaO Mn6N5, CaO 23.33 109 ties of the Mn6N5-11CaNH-2 catalyst at different gas hourly
Mn6N5 Mn6N5 82.46 138 space velocity rates for NH3 (GHSVNH3 ) and temperatures. At
CaNH CaNH – –
723 K, the GHSVNH3 has a relatively weak effect, and the maxi-
[a] Manganese actual loadings determined by ICP-AES. [b] Activation mum H2 formation rate is ca. 4.5 mmolH2 gcat.¢1 min¢1. The
energy was measured under a flow of pure NH3.
effect of GHSVNH3 becomes more pronounced at increased re-
ChemSusChem 2016, 9, 364 – 369 www.chemsuschem.org 365 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1864564x, 2016, 4, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cssc.201501498 by Sichuan University, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Full Papers
Figure 2. Arrhenius plots of Mn6N5-11CaNH-2 (squares), Mn6N5-15CaNH-1 (cir-
cles), and Mn6N5 (triangles). Figure 4. XRD patterns of (a) as-prepared, and (b) post-activity test Mn6N5-
11CaNH-2 composite catalysts
lected after an activity test at 873 K. Only CaNH phase can be
identified in the as-prepared catalyst. The absence of any
phase related to manganese is likely due to the poor crystallin-
ity of the sample (poor signal-to-noise ratio of curve a in
Figure 4). The XANES characterization shows that the chemical
state of manganese in the as-prepared catalyst is identical to
that of neat Mn6N5 (Figure 6). The XRD pattern of Mn6N5-
11CaNH-2 catalyst collected after the activity test reveals
Mn6N5, CaNH, and Ca4NHCl6 phases (see Figure S1). The signifi-
cantly enhanced crystallinities of Mn6N5 and CaNH indicate par-
ticle aggregation and recrystallization at elevated tempera-
tures. The presence of Ca4NHCl6 is unavoidable under the
Figure 3. Effects of reaction temperature and GHSVNH3 on the catalytic per- preparation method applied here. Little catalytic function of
formance of Mn6N5-11CaNH-2. this substance is observed.
The enhanced catalytic activities and significantly reduced
action temperatures. At 823 K, the H2 formation rate still has apparent activation energies manifest the strong promoting
room to increase at the GHSVNH3 higher than effect of CaNH on Mn6N5. We thus investigated the interaction
72 000 mLNH3 gcat.¢1 h¢1. It is worth highlighting that the H2 for- between these two compounds by means of temperature-pro-
mation rate of the Mn6N5-11CaNH-2 catalyst reaches ca. grammed decomposition (TPD) and XANES. A minor amount
16 mmolH2 gcat.¢1 min¢1 at 773 K under the GHSVNH3 of of H2 was observed in the TPD profile of neat CaNH at temper-
60 000 mLNH3 gcat.¢1 h¢1, which is comparable to the highly atures above 650 K (Figure 5 a). For the Mn6N5-15CaNH-
active Ni/SBA-15 and Ru/Al2O3 catalysts under the similar reac- 1 sample, however, a significantly increased amount of H2 was
tion conditions.[5a, 6a] A MnN-0.2Li2NH catalyst reported by us re- released at ca. 550 K, peaking at ca. 670 K. In the meantime
cently showed a higher NH3 conversion rate substantial amounts of N2 and minor amounts of NH3 were
(25 kgNH3 kgcat.¢1 h¢1, 773 K).[6d] However, owing to the relatively generated also (Figure 5 b), showing significant differences
low melting point of LiNH2 (below 673 K), the catalyst has to
have a high manganese content to maintain structural stability.
CaNH, on the other hand, is in solid form under the reaction
conditions so that the manganese content can be largely re-
duced in the present case. When considering the NH3 conver-
sion rate based on the mass of manganese, the NH3 conver-
sion rate of the Mn6N5-11CaNH-2 catalyst (50 kgNH3 kgMn¢1 h¢1,
773 K) is markedly faster than that of the MnN-0.2 Li2NH cata-
lyst (34 kgNH3 kgMn¢1 h¢1, 773 K). The stability of the Mn6N5-
11CaNH-2 catalyst was also investigated. As shown in Figure S3
(Supporting Information), the activity experiences a slight drop
from 95 % to 90 % over a period of 12 h at 823 K. We ascribe
this drop to aggregation of CaNH and Mn6N5 (as evidenced in
Figure 4 shown below).
Figure 4 shows X-ray diffraction (XRD) patterns of the as-pre- Figure 5. TPD measurements of neat (a) CaNH, and (b) Mn6N5-15CaNH-
pared Mn6N5-11CaNH-2 composite catalyst and the catalyst col- 1 samples. 1 Torr = 133.3 Pa.
ChemSusChem 2016, 9, 364 – 369 www.chemsuschem.org 366 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1864564x, 2016, 4, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cssc.201501498 by Sichuan University, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Full Papers
The mechanistic interpretation of the Mn6N5-CaNH system is,
in principle, the same as the one proposed for the Li2NH-3d
transition metal nitride system.[6c] Notably, bulk-phase Ca6MnN5
was not observed in the post-TPD Mn6N5-15CaNH-1 and
Mn6N5-11CaNH-2 samples (Supporting Information, Figure S2),
which is obviously different from the observation of ternary
Li7MnN4 phase in the Li2NH-MnN composite catalyst.[6d] We ten-
tatively ascribe this difference to the slow migration rates of
large and highly charged ions (Ca2 + ) in the solid-state reaction
of Mn6N5 and CaNH. From this information we deduce that the
catalysis may be carried out within a few layers of the Mn6N5/
CaNH interface.
Figure 6. Mn K-edge XANES spectra of manganese-containing samples. NH3 decomposition proceeds through successive dehydro-
genation followed by recombinative desorption of adsorbed
nitrogen and hydrogen atoms on transition metals or metal ni-
from neat CaNH and evidencing interactions between CaNH trides. Previous kinetic studies indicate that at low tempera-
and Mn6N5. tures, the recombinative desorption of adsorbed nitrogen
XANES was employed to characterize the chemical state of atoms is the rate-determining step, while at high temperatures,
manganese. We prepared a sample with low manganese con- the dissociation of the N¢H bond is a slow step.[22] The stabili-
tent (Mn6N5-212CaNH-2) for characterization, to ensure a uni- ties of intermediate nitrogen species are closely related to the
form chemical environment of manganese. Figure 6 shows the kinetic barrier of the rate-determining step according to the
manganese K-edge XANES spectra of Mn6N5-212CaNH-2 col- Brønsted-Evans-Polanyi (BEP) relation. For neat Mn6N5, the in-
lected under different conditions and neat Mn3N2 and Mn6N5 termediate species Nad forming on the surface of Mn6N5 may
for comparison. Unfortunately, the inflection points of the be unstable. The related kinetic barrier of the rate-determining
main absorption edges in many spectra overlap with pre-edge step is expectedly high, which is partially supported by the
features. However, the absorption edge level at about 0.6 of high apparent activation energy of neat Mn6N5 shown in
the full edge jump is still visible. We found that the order of Figure 2 and by the fact that a nitrogen-rich MnNx (x 1) sub-
manganese oxidation state w as post-TPD > post-testing > stance is not thermodynamically stable. The presence of CaNH
fresh = Mn6N5 > Mn3N2. In other words, the manganese species changes the situation by forming the nitrogen-rich intermedi-
in the post-TPD sample (i.e., the sample collected after the in- ate species Ca6MnN5. The Mn¢N distance in Ca6MnN5 is
teraction between Mn6N5 and CaNH occurred) has the highest 1.796 æ[21] considerably shorter than the Mn¢N bond length in
oxidation state or is in a nitrogen-rich environment. In addi- Mn6N5(2.128 æ),[23] reflecting the stabilization of the nitrogen
tion, the most pronounced and best separated pre-edge peak atom in the vicinity of manganese and calcium. As a conse-
for the post-TPD sample is visible at around 6540 eV, which in- quence, the kinetic barrier of rate-determining step is de-
dicates there is a distorted [MnIIIN3]6¢ group that lacks centro- creased according to the BEP relation.
symmetry in the post-TPD sample.[20] Gregory and co-workers The function of CaNH goes beyond the role of structure pro-
reported that the ternary nitride Ca6MnN5 consists of distorted moter. It offers synergy with Mn6N5 and participates in the cat-
[MnIIIN3]6¢ anions.[21] We, therefore, tentatively assign the man- alytic reaction by forming a [Ca6MnN5]-like intermediate, and
ganese species in the post-TPD sample to a [Ca6MnN5]-like en- substantially reduces the kinetic barrier from 138 kJ mol¢1 (neat
vironment, in which the oxidation state of manganese is Mn6N5) to 85 kJ mol¢1 (Figure 2). All of the experimental results
higher than that in Mn6N5. presented herein lead us to propose that CaNH is a co-catalyst.
Combined with the TPD and XANES results, Reaction 1 (see Our experimental results also show that when CaO is mixed
below) likely occurs in the TPD treatment. Under the reaction with Mn6N5, no distinct increase in catalytic activity is observed
condition, that is, in a pure NH3 flow, the [Ca6MnN5]-like spe- (Figure 1), which further evidences that the chemical cycle (Re-
cies may go back to Mn6N5 and CaNH following Reaction 2, actions 1 and 2) can not be achieved without CaNH.
which can be indicated by the fact that the reaction of the
post-TPD sample with NH3 gives rise to a manganese species
resembling that of the as-prepared sample (see Figure 6).
Conclusions
Through the combination of Reactions 1 and 2, the neat Reac-
tion 3 (NH3 decomposition) can be fulfilled over the Mn6N5- Mn6N5, which is almost inactive towards ammonia decomposi-
CaNH composite. tion, shows a relatively high catalytic activity through synergy
with CaNH. The role of CaNH goes beyond that of structural
Mn6 N5 þ 36 CaNH ¼ 6 Ca6 MnN5 þ 11=2 N2 þ 18 H2 ðReaction 1Þ promoter: it participates in the catalysis and produces a stable
[Ca6MnN5]-like intermediate, and thus lowers the activation
6 Ca6 MnN5 þ 12 NH3 ¼ Mn6 N5 þ 36 CaNH þ 1=2 N2 ðReaction 2Þ energy for ammonia decomposition. A two-step catalytic cycle
accounting for the synergistic effect between CaNH and Mn6N5
2 NH3 ¼ 3 H2 þ N2 ðReaction 3Þ is proposed.
ChemSusChem 2016, 9, 364 – 369 www.chemsuschem.org 367 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1864564x, 2016, 4, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cssc.201501498 by Sichuan University, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Full Papers
Experimental Section Acknowledgements
Preparation of catalysts: Ca(NH2)2 was prepared by reacting calci- The authors thank financial support from the Project of National
um metal with liquid ammonia overnight. CaNH was obtained by Science Funds for Distinguished Young Scholars (51225206), the
decomposing Ca(NH2)2 under vacuum at 673 K for 7 h ENRE- National Natural Science Foundation of China (U1232120), and
F 28 ENREF 2. Mn6N5 (usually referred to as MnN; there is nitro- the Shanghai Synchrotron Radiation Facility (SSRF) for providing
gen deficiency in MnN, so that the actual stoichiometry of MnN is
the beam time.
Mn6N5.[23]) was prepared as follows: lithium amide (267 mg) and
manganese(II) chloride (733 mg) were firstly ball milled in an agate
jar at 323 K (150 rpm for 3 h) on a Retsch planetary ball mill Keywords: ammonia · calcium · heterogeneous catalysis ·
(PM 400), followed by heating the sample at 573 K for 2 h under manganese · nitrides
pure NH3 atmosphere. Then the sample was purified by stirring it
in dry tetrahydrofuran (THF) followed by centrifugation for 3 times
to remove LiCl. The Mn6N5-CaNH composite catalysts were synthe- [1] S. F. Yin, B. Q. Xu, X. P. Zhou, C. T. Au, Appl. Catal. A 2004, 277, 1 – 9.
sized by ball-milling CaNH with Mn6N5 or manganese(II) chloride in [2] a) T. V. Choudhary, C. Sivadinarayana, D. W. Goodman, Catal. Lett. 2001,
an agate jar at 150 rpm for 5 h on a Retsch planetary ball mill 72, 197 – 201; b) A. Klerke, C. H. Christensen, J. K. Nørskov, T. Vegge, J.
(PM 400).The Mn6N5-CaNH-1 denotes the sample with Mn6N5 as the Mater. Chem. 2008, 18, 2304 – 2310; c) F. Schìth, R. Palkovits, R. Schlçgl,
D. S. Su, Energy Environ. Sci. 2012, 5, 6278 – 6289.
precursor and the Mn6N5-CaNH-2 represents MnCl2 as the precur-
[3] J. C. Ganley, F. S. Thomas, E. G. Seebauer, R. I. Masel, Catal. Lett. 2004,
sor. All the catalysts were stored in a glove box filled with argon.
96, 117 – 122.
The amount of metal was determined by ICP-AES. The symbol x in [4] a) S. F. Yin, B. Q. Xu, C. F. Ng, C. T. Au, Appl. Catal. B 2004, 48, 237 – 241;
Mn6N5-xCaNH composite catalysts represents for the molar ratio b) W. H. Chen, I. Ermanoski, T. E. Madey, Abstr. Am. Chem. Soc. 2004, 228,
between CaNH and Mn6N5. 474; c) X. K. Li, W. J. Ji, J. Zhao, S. J. Wang, C. T. Au, J. Catal. 2005, 236,
181 – 189; d) A. Vavere, R. S. Hansen, J. Catal. 1981, 69, 158 – 171; e) A.
Catalyst test: Ammonia decomposition reactions were performed Amano, H. Taylor, J. Am. Chem. Soc. 1954, 76, 4201 – 4204; f) M. Gros-
on a continuous-flow fixed-bed quartz reactor at atmospheric pres- man, D. G. Loffler, React. Kinet. Catal. Lett. 1987, 33, 87 – 92; g) X. Z.
sure. Typically, 50 mg of catalyst powder was loaded and tested in Duan, G. Qian, X. G. Zhou, Z. J. Sui, D. Chen, W. K. Yuan, Appl. Catal. B
the temperature range of 573 K to 873 K under pure NH3 at a flow 2011, 101, 189 – 196.
rate of 30 mL min¢1. Activity was measured at a 25 K interval. Am- [5] a) S. F. Yin, B. Q. Xu, W. X. Zhu, C. F. Ng, X. P. Zhou, C. T. Au, Catal. Today
monia conversion data reported here were collected after 20 min 2004, 93 – 95, 27 – 38; b) K. Nagaoka, T. Eboshi, N. Abe, S. Miyahara, K.
Honda, K. Sato, Int. J. Hydrogen Energy 2014, 39, 20731 – 20735; c) S.
time-on-stream at corresponding temperatures. Gas flow rate was
Armenise, L. Roldan, Y. Marco, A. Monzon, E. Garcia-Bordeje, J. Phys.
regulated by a mass flow controller (Brooks instrument, model
Chem. C 2012, 116, 26385 – 26395; d) F. R. Garca-Garca, J. Alvarez-Rodri-
5850 E). Gas composition was analyzed using an online gas chro- guez, I. Rodriguez-Ramos, A. Guerrero-Ruiz, Carbon 2010, 48, 267 – 276;
matograph (GC-2014C, Shimadzu) equipped with a Porapak N e) F. Hayashi, Y. Toda, Y. Kanie, M. Kitano, Y. Inoue, T. Yokoyama, M. Hara,
column and thermal conductivity detector (TCD). H. Hosono, Chem. Sci. 2013, 4, 3124 – 3130.
[6] a) H. C. Liu, H. Wang, J. H. Shen, Y. Sun, Z. M. Liu, Appl. Catal. A 2008,
BET measurements: A Micromeritics ASAP 2010 automated physi- 337, 138 – 147; b) J. Zhang, H. Y. Xu, X. L. Jin, Q. J. Ge, W. Z. Li, Appl.
sorption instrument was used to measure the N2 adsorption iso- Catal. A 2005, 290, 87 – 96; c) J. P. Guo, P. K. Wang, G. T. Wu, A. A. Wu,
therms of samples at liquid N2 temperature (77 K). The specific sur- D. Q. Hu, Z. T. Xiong, J. H. Wang, P. Yu, F. Chang, Z. Chen, P. Chen,
face area was determined from the linear portion of the BET plot. Angew. Chem. Int. Ed. 2015, 54, 2950 – 2954; Angew. Chem. 2015, 127,
Before each measurement, the sample was heated to 573 K and 2993 – 2997; d) J. P. Guo, F. Chang, P. K. Wang, D. Q. Hu, P. Yu, G. T. Wu,
kept at this temperature for 5 h. Z. T. Xiong, P. Chen, ACS Catal. 2015, 5, 2708 – 2713.
[7] J. S. J. Hargreaves, Coord. Chem. Rev. 2013, 257, 2015 – 2031.
XRD measurements: XRD patterns of samples were performed by [8] C. R. Lotz, F. Sebba, Trans. Faraday Soc. 1957, 53, 1246 – 1252.
using a PANalytical X’pert diffractometer with CuKa radiation at [9] a) S. T. Oyama, J. Catal. 1992, 133, 358 – 369; b) J. G. Choi, M. K. Jung, S.
Choi, T. K. Park, I. H. Kuk, J. H. Yoo, H. S. Park, H. S. Lee, D. H. Ahn, H. S.
40 kV and 40 mA. A self-made sample cell was used to protect
Chung, Bull. Chem. Soc. Jpn. 1997, 70, 993 – 996.
samples from air contamination.
[10] a) W. Zheng, T. P. Cotter, P. Kaghazchi, T. Jacob, B. Frank, K. Schlichte, W.
Zhang, D. S. Su, F. Schueth, R. Schlçgl, J. Am. Chem. Soc. 2013, 135,
Temperature-programmed decomposition: The gaseous products
3458 – 3464; b) V. Tagliazucca, K. Schlichte, F. Schueth, C. Weidenthaler,
from samples during heat treatment were measured by using
J. Catal. 2013, 305, 277 – 289; c) V. Tagliazucca, M. Leoni, C. Weidenthal-
a home-made TPD system with an online mass spectrometer er, Phys. Chem. Chem. Phys. 2014, 16, 6182 – 6188.
(Hiden QIC-20), which recorded the signals of H2 (m/z = 2), N2(m/ [11] Z. Schnepp, M. Thomas, S. Glatzel, K. Schlichte, R. Palkovits, C. Giordano,
z = 28), and NH3 (m/z = 17) simultaneously. Pure argon was used as J. Mater. Chem. 2011, 21, 17760 – 17764.
the carrier gas (30 mL min¢1). In each test, 50 mg of sample was [12] A. Boisen, S. Dahl, J. K. Nørskov, C. H. Christensen, J. Catal. 2005, 230,
loaded and heated from room temperature to 773 K at a ramping 309 – 312.
rate of 5 K min¢1. [13] a) C. H. Liang, W. Z. Li, Z. B. Wei, Q. Xin, C. Li, Ind. Eng. Chem. Res. 2000,
39, 3694 – 3697; b) X. Z. Duan, G. Qian, X. G. Zhou, D. Chen, W. K. Yuan,
XANES measurements: The manganese local environment was Chem. Eng. J. 2012, 207, 103 – 108; c) J. Ji, X. Z. Duan, G. Qian, X. G.
characterized by the extended X-ray absorption fine structure tech- Zhou, G. S. Tong, W. K. Yuan, Int. J. Hydrogen Energy 2014, 39, 12490 –
nique. X-ray absorption data were collected on BL14W beam line 12498.
[14] L. Li, Z. H. Zhu, G. Q. Lu, Z. F. Yan, S. Z. Qiao, Carbon 2007, 45, 11 – 20.
(Synchrotron Radiation Facility, Shanghai, China) operating at an
[15] W. G. Gao, H. Wang, C. Han, W. Y. Liu, Y. H. Wang, M. Y. Jia, Chem. Ind.
electron energy of 3.5 GeV and a ring current of 300 mA. XANES Eng. Prog. 2014, 33, 2963 – 2969.
signals of self-supporting wafers were measured at the Mn K edge [16] N. N. Sun, X. Wen, N. Zhao, F. K. Xiao, W. Wei, Y. H. Sun, The 238th ACS
in fluorescence mode. Samples were pressed into pellets and then National Meeting, 2009, Aug. 16th..
sealed with KAPTON films on both sides to avoid air contamina- [17] Z. Lendzion-Bielun, Pol. J. Chem. 2007, 81, 433 – 440.
tion. [18] T. Sichla, H. Jacobs, Z. Anorg. Allg. Chem. 1996, 622, 2079 – 2082.
ChemSusChem 2016, 9, 364 – 369 www.chemsuschem.org 368 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1864564x, 2016, 4, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/cssc.201501498 by Sichuan University, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Full Papers
[19] a) S. Hino, T. Ichikawa, H. Y. Leng, H. Fujii, J. Alloys Compd. 2005, 398, [23] K. Suzuki, T. Kaneko, H. Yoshida, Y. Obi, H. Fujimori, H. Morita, J. Alloys
62 – 66; b) K. Tokoyoda, S. Hino, T. Ichikawa, K. Okamoto, H. Fujii, J. Compd. 2000, 306, 66 – 71.
Alloys Compd. 2007, 439, 337 – 341.
[20] R. Niewa, Z. W. Hu, R. Kniep, Eur. J. Inorg. Chem. 2003, 1632 – 1634.
[21] D. H. Gregory, M. G. Barker, P. P. Edwards, D. J. Siddons, Inorg. Chem.
1995, 34, 5195 – 5198. Received: November 6, 2015
[22] W. Tsai, W. H. Weinberg, J. Phys. Chem. 1987, 91, 5302 – 5307. Published online on January 14, 2016
ChemSusChem 2016, 9, 364 – 369 www.chemsuschem.org 369 Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
