Synthesis of novel multi-functional fluorene-based benzoxazine resins: Polymerization behaviour, curing kinetics, and thermal properties

Article abstract of DOI:10.1016/j.reactfunctpolym.2019.104344

In this paper, a new series of novel aromatic multi-functional fluorene-based benzoxazine monomers (AMFB) were synthesized from the reaction of 9,9-bis(4-aminophenyl)-2,7-dihydroxy-fluorene, aniline, paraformaldehyde, and unsubstituted or substituted phen

Full text of DOI:10.1016/j.reactfunctpolym.2019.104344

http://pubs.acs.org/journal/aelccp
Cite This: ACS Energy Lett. 2019, 4, 2500−2507
Impacts of Imidazolate Ligand on
Performance of Zeolitic-Imidazolate
Framework-Derived Oxygen Reduction
Catalysts
Hao Wang,†,§,∥,⊥ Lauren R. Grabstanowicz,†,∥,∇ Heather M. Barkholtz,†,○ Dominic Rebollar,†,‡
Zachary B. Kaiser,† Dan Zhao,# Biao-Hua Chen,§ and Di-Jia Liu*,†
†Chemical Sciences and Engineering Division, Argonne National Laboratory, Lemont, Illinois 60439, United States
‡Department of Chemistry and Biochemistry, Northern Illinois University, DeKalb, Illinois 60115, United States
#Department of Chemical and Biomolecular Engineering, National University of Singapore, 117585 Singapore
§College of Environmental and Energy Engineering, Beijing University of Technology, Beijing 100124, PR China
S
* Supporting Information
ABSTRACT: Carbon-hosted Fe−N-coordinated (Fe/N/C) materials, espe-
cially those derived from thermolysis of iron-added zinc-based zeolite-
imidazolate frameworks (ZIFs), have emerged as the most promising platinum
group metal-free (PGM-free) oxygen reduction reaction (ORR) catalysts.
However, the impacts of the precursory ZIF structure and their conversion
chemistry during thermal activation to the final catalytic activity remain to be
further explored in the process of continuously refining the catalyst
performance. Herein, we synthesized a series of Fe-doped ZIFs with different
imidazolate ligands and systematically studied the correlation between the
crystal structures with the final ORR catalytic activity in an alkaline
electrolyte. We also investigated the ZIF decomposition chemistry during
pyrolysis using a thermogravimetric mass-spectroscopic analysis. We
demonstrated that imidazole side-chain substitution alters the ZIF’s
decomposition during pyrolysis, influencing the elemental compositions, surface properties, graphitization levels, and
ultimately the catalytic performance. The Zn(mIm)2TPIBP catalyst affords the highest ORR activity with a half-wave
potential of 0.93 V vs RHE, representing the best among all PGM-free catalysts studied.
received a significant amount of attention in recent years
because of their promising catalytic activity.6−10
nvironmentally friendly, all electric vehicles (EVs) have
recently garnered much attention because of worldwide
Ideally, an effective ORR catalyst should have high volumetric
active site density exposed by high surface area without inactive
support such as amorphous carbon used in the traditional
electrocatalyst. Using metal−organic frameworks (MOFs), such
as zeolitic imidazolate frameworks (ZIFs), as the precursors
exemplifies such an approach. Since the original studies of ZIF-
based PGM-free ORR catalysts,6,11 studies on ZIFs as catalytic
precursors have undergone a rapid expansion with many
promising results,12 particularly for Fe/N/C catalysts prepared
from ZIFs containing coordinated13−15 or added iron.16−18 As
the precursors, ZIFs are attractive because of their tunable
structure, inherent porosity, high surface area, and diverse
Econcerns regarding fossil fuel consumption and emission
from internal combustion engine vehicles.1 One of the most
promising EVs is the hydrogen-powered fuel cell with the
advantage of long cruising range and short refueling time.2 Fuel
cell commercialization relies on the cost reduction, of which
reducing or removing platinum group metal (PGM) usage in the
electrocatalysts represents a particularly important path.3
Compared to the anode, the cathodic oxygen reduction reaction
(ORR) has significantly higher kinetic barrier and requires more
PGM catalysts. If PGMs can be replaced by earth-abundant
elements with a simple synthetic route, the fuel cell cost will then
be substantially lowered.4 Exploration of transition metal/
nitrogen coordinated carbon (TM/N/C) as an ORR catalyst
was initiated by Jasinski in 1964 using Co phthalocyanine as
precursors.5 Ever since, the research in this direction has
attracted a great following. In particular, Fe/N/C catalysts have
Received: August 9, 2019
Accepted: September 19, 2019
Published: September 19, 2019
© 2019 American Chemical Society
2500
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
Figure 1. (a) Ligation structure for tris(1,10-phenanthroline) iron(II) perchlorate (TPI) and Zn-ligand coordination structures in (b)
Zn(mIm)2, (c) Zn(eIm)2, (d) Zn(4abIm)2, and (e) Zn(Im)2. Black, carbon; blue, nitrogen; red, iron; and green, zinc.
Figure 2. RRDE results from (A) ORR linear sweep voltammetry polarization curve, (B) Tafel plot of mass-corrected kinetic current, and (C)
the number of electrons transferred as the function of polarization potential. (D) Percentage loss of kinetic current density against the initial
value measured at 0.90 V (vs RHE) as a function of voltage cycle.
composition, all originating from the ordered 3D framework
formed by repetitive M−N4 ligations.19 Upon carefully
controlled heat-treatment, Fe-containing ZIFs can be converted
into a highly porous carbonaceous ORR catalyst uniformly
decorated by Fe/N/C active sites.
conversion, one expects that a majority of the pentagonal
imidazolate rings will be transformed to hexagonal graphitic
structure through decomposition and rearrangement.11 The side
functional group of the imidazolate will participate and influence
the conversion pathway and consequently the final catalyst
porosity and surface functionality. Although significant progress
has been made in improving the ORR performance, the
In the ZIF structure, the imidazole ligand controls the
porosity, surface area, and topology.20,21 During the thermal
2501
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
influence of the ZIF ligand on catalyst activity has been explored
only by a limited yet insightful investigation.22
Zn(mIm)2TPIBP has the highest ImK of 18.89A·g−1; Zn-
(eIm)2TPIBP and Zn(4abIm)2TPIBP also possess excellent
ImK of 10.62 and 7.46 A·g−1, respectively, whereas Zn-
(Im)2TPIBP underperforms. The RRDE Tafel plots are
shown in Figure 2B, of which the slope characterizes the ORR
kinetics. Zn(mIm)2TPIBP afforded a Tafel slope as low as 89
mV·dec−1 at low overpotentials, indicating a faster current surge
with the increase of polarization potential, similar to that of Pt/C
(Figure S1B). We also estimated the ORR turnover frequency
(TOF) at 0.90 V (vs RHE) over Zn(mIm)2TPIBP by assuming
all Fe sites were atomically dispersed.23 The TOF was found to
be 0.21 s−1·atom−1 for Zn(mIm)2TPIBP, which is in general
agreement with those reported in the literature (Supporting
Information).
The RRDE method was also used to measure the number of
electrons transferred, n, at different polarization potential drawn
from the disk and ring currents (Figure 2C). As n = 4 represents
the complete conversion of oxygen to water (O2 + 2H2O + 4e−
→ 4OH−), lower electron-transfer numbers are the result of the
formation of 2e− intermediate H2O2. Hydrogen peroxide can
corrode the PGM-free catalysts and ion-conductive membrane
in the fuel cell and therefore should be minimized. Zn-
(mIm)2TPIBP achieved excellent n values, ranging from 3.94
to 3.99 within the polarization potential range. Zn(eIm)2TPIBP
and Zn(4abIm)2TPIBP also showed good n values ranging from
3.90 to 3.99 and 3.86 to 3.99, respectively. At the preferred fuel
cell operating voltage >0.7 V, all the catalysts displayed n ≈ 4 and
low H2O2 production, except for Zn(Im)2TPIBP (Figure S3).
The durability of the PGM-free catalysts was evaluated under
repeated potential cycling from 0 to 1.20 V in O2-saturated 0.1
M KOH (Figures S4 and S5). LSVs were recorded after a
designated number of cycles, and the kinetic current densities
(jK) at 0.90 V were measured and normalized against that of the
fresh catalyst (Figure 2D). The retention of jK after 10 000 cycles
follows the order of Zn(Im)2TPIBP (73.4%) > Zn(eIm)2TPIBP
(71.8%) > Zn(mIm)2TPIBP (66.8%) > Zn(4abIm)2TPIBP
(61.5%). The catalyst with the lowest activity, Zn(Im)2TPIBP,
exhibited the highest durability, while the others showed similar
deactivation rates.
Our study suggests that the structure of the imidazole ligand
in ZIF influences the ORR activity and durability of the final
catalyst. The catalyst with the best ORR activity is from ZIF with
the 2-methylimidazolate ligand. The activity decreases only
slightly with higher alkyl side-chain substitution, as are the cases
for 2-ethylimidazolate and 4-azabenzimidazolate, but Zn-
(eIm)2TPIBP showed better durability than Zn(mIm)2TPIBP.
The imidazolate ligand without the side chain leads to the lowest
activity for Zn(Im)2TPIBP. The ligand size affects the
composition, structure, and graphitization level, all of which
affect the catalyst’s activity and durability. As we will discuss
below, these catalyst properties are influenced by ligand-
determined ZIF surface property and decomposition chemistry.
The catalysts were characterized by a variety of techniques. X-
ray diffraction (XRD) patterns of catalysts after synthesis and
ball milling to NH3 activation were also included in Figure S6,
compared to simulated XRD patterns.20,24 The addition of the
iron complex did not alter the crystalline structure of ZIFs
because TPI remains outside the ZIF cavity as the result of its
large size. After ball milling, Zn(mIm)2TPIB exhibited extensive
peak broadening due to the loss of long-range crystal
ordering.25,26 In contrast, Zn(Im)2TPIB, Zn(eIm)2TPIB, and
Zn(4abIm)2TPIB showed only minor peak broadening,
attributed to their lattice rigidity in the absence of porosity.
Herein, we prepared four ZIF-derived catalysts using our
previously demonstrated one-pot ZIF synthesis method16 and
systematically investigated their ORR activities in alkaline
media. Four ZIFs were prepared through the solid-state reaction
between ZnO with imidazole (HIm), 2-methylimidazole
(HmIm), 2-ethylimidazole (HeIm), or 4-azabenzimidazole
(H4abIm). Tris (1,10-phenanthroline)iron(II) perchlorate
(TPI) was added to incorporate both iron and ligated nitrogen
into the precursors. ZIFs with imidazole of different side chains
are denoted as Zn(ligand)2. The coordination structures of TPI,
Zn(Im)2, Zn(mIm)2, Zn(eIm)2, and Zn(4abIm)2 are given in
Figure 1. Samples containing TPI are denoted as Zn-
(ligand)2TPI and, after ball-mill, Zn(ligand)2TPIB. The
resulting powder samples were subject to pyrolysis to produce
the final catalysts, denoted as Zn(ligand)2TPIBP. Elemental
analysis showed these catalysts contain a similar amount of Fe at
ca. 1 wt % (Table S1). To better understand the ZIF-to-catalyst
conversion process, we further investigated the catalyst structure
as a function of the ligand and the pyrolysis chemistry. The
results demonstrated that the imidazolate ligand in ZIF impacts
strongly on the final catalyst’s surface property and composition.
The catalyst derived from Zn(mIm)2TPIBP showed the best
ORR catalytic activity, correlated to its higher surface area and
pyridinic-N content. Detailed procedures on the catalyst
synthesis can be found in the Supporting Information.
The catalytic activity and durability were tested using RRDE
methods in an alkaline solution (0.1 M KOH). All the potentials
reported are in reference to the reversible hydrogen electrode
(RHE). Figure 2A displays linear sweep voltammograms (LSVs)
of all four catalysts. The onset potentials (E0) in Table 1 were
Table 1. Onset Potential (E0), Half-Wave Potential (E1/2),
Mass-Corrected Kinetic Current (ImK) at 0.90 V (vs. RHE),
and Tafel Slope for All Four PGM-Free Electrocatalysts
E0
E1/2
ImK
Tafel slope
electrocatalyst
[V vs RHE] [V vs RHE] [A·g−1
]
[mV·dec−1
]
Zn(mIm)2TPIBP
Zn(eIm)2TPIBP
Zn(4abIm)2TPIBP
Zn(Im)2TPIBP
1.14
1.12
1.12
1.11
0.93
0.92
0.90
0.86
18.89
10.62
7.46
89
98
95
2.97
124
measured from the LSVs at the current density of 22 μA·cm−2.
Additionally, half-wave potentials (E1/2) were recorded at the
half point between onset and diffusion-limited current densities
and are also listed in Table 1. A commercial Pt/C catalyst was
also tested as a reference with optimized E0 of 1.05 V and E1/2 of
0.88 V at the platinum loading of 60 μgPt·cm−2 (Figures S1A and
Figure S2). For PGM-free catalysts, Zn(mIm)2TPIBP exhibited
an E0 of 1.14 V and an E1/2 of 0.93 V, representing the best ORR
catalytic activity for a PGM-free catalyst in the alkaline
electrolyte according to the literature (Table S2). Zn-
(eIm)2TPIBP and Zn(4abIm)2TPIBP also displayed excellent
activities with E1/2 of 0.92 and 0.90 V, respectively. All three
catalysts surpassed the Pt/C benchmark in E0 and E1/2
.
Zn(Im)2TPIBP, the catalyst derived from imidazolate without
side chain, showed a lower activity with an E0 of 1.11 V and an
E
1/2 of 0.86 V.
To assess the ORR reaction kinetics, the mass-corrected
kinetic current (ImK) of the PGM-free catalysts was calculated at
0.90 V and is listed in Table 1. Among the PGM-free catalysts,
2502
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
Figure 3. (A) Relative pyridinic-N content and BET specific surface area (SSA) plotted as a function of mass-corrected kinetic current density at
0.90 V (vs RHE). (B) Diffusion-limited current density at 0.60 V (vs RHE) plotted as a function of BET SSA. (C) Tafel slope and half-wave
potential (E1/2) plotted as a function of micropore volume. (D) Relationship between the durability and graphitization level of the catalysts.
After thermal activation, no crystallinity was found in all catalysts
except Zn(Im)2TPIBP, which displayed a graphite crystallite
feature. Figure S7 shows the Raman spectra of the catalysts. The
spectra between 800 and 2000 cm−1 can be fitted into four peaks,
two of which represent defect-activated (D band, 1347 cm−1)
and planar motion of sp2-carbon atoms (G band, 1593 cm−1).27
The intensity ratio of these two bands, IG/ID, characterizes the
catalyst graphitization level with higher value being more
graphitized. We found that this ratio follows Zn(Im)2TPIBP
(0.53) > Zn(eIm)2TPIBP (0.52) > Zn(mIm)2TPIBP (0.51) >
Zn(4abIm)2TPIBP (0.46), suggesting a decreasing order of the
graphitization. Increasing the sizes of imidazole alkyl sub-
stitution clearly led to more amorphous catalyst (Table S3).
The investigations on the catalyst surface properties revealed
that Zn(mIm)2TPIBP has the highest Brunauer−Emmett−
Teller (BET) specific surface area (SSA) of 1139 m2·g−1 among
the four (Figure S8 and Table S4). The total pore volume
follows the same order as the surface area with Zn(mIm)2TPIBP
being the highest (0.62 cm3·g−1) and Zn(Im)2TPIBP the lowest
(0.30 cm3·g−1). The catalysts’ morphology and composition
were also studied by transmission electron microscopy (TEM),
showing a uniform and porous Zn(mIm)2TPIBP in the absence
of agglomerated metal particles (Figure S9A,B). The elemental
mapping of Fe and N showed that they are both populated
densely and uniformly within the carbon matrix (Figure S9C−
E). TEM images also found that Zn(eIm)2TPIBP and
Zn(4abIm)2TPIBP (Figure S10A,B) are significantly more
amorphous than Zn(Im)2TPIBP, of which the graphitic carbons
such as nanotubes and iron carbide particles were detected
(Figure S11). Zn(mIm)2TPIBP was found to be more porous
with less Zn compared to Zn(eIm)2TPIBP after the same
treatment (Figure S12), suggesting that the more porous ZIF
precursor leads to more complete Zn elimination and higher
porosity in final catalyst.
A number of studies attributed the ORR active site in the Fe/
N/C composite to a single iron atom ligated by four pyridinic or
pyrrolic nitrogens embedded on the graphitic surface, Fe/N4/
C.23,28 The empirical connection between the catalyst chemical/
structural properties and its performance could shed light on the
key descriptors of the active center. To this end, we first studied
different nitrogen and carbon moieties in our ZIF-derived
catalysts using X-ray photoelectron spectroscopy (XPS)
(Figures S13−S15) and tabulated their relative concentrations
(Tables S5 and S6). We then examined the correlation between
these compositions with the ORR polarization curve at both the
kinetic-controlled region (high voltage) and a diffusion-
controlled region (high current). As the descriptors for ORR
kinetics, both E0 and iK at 0.90 V (vs RHE) increase linearly as
the pyridinic-N density increases (Figure 3A), whereas they
correlate weakly, if at all, with the pyrrolic-N and graphitic-N
contents (Figure S16). This observation is in good agreement
with the previous reports on pyridinic-N in promoting the ORR
activity more than other N forms.29−35 The amount of each N
moiety as the function of the total oxidized carbons (CxOy) is
also plotted (Figure S17). Only pyridinic-N showed a positive
correlation with CxOy, which is consistent with the report by
Artyushkova et al.36 on formation of pyridinic-N by carbon
oxidation. Figure 3B shows a linear dependence of the oxygen
2503
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
Figure 4. (A) TGA curves of all PGM-free catalysts, ramping from room temperature to 1000 °C in Ar atmosphere. Mass spectra recorded at (B)
mass = 2 (H2); (C) mass = 15 (methyl); (D) mass = 17 (NH3); (E) mass = 28 (N2, CH2=CH2), with inset mass = 27 (·CH=CH2); and (F) mass =
29 (·CH2−CH3). Square, Zn(mIm)2TPIBP; circle, Zn(eIm)2TPIBP; up-triangle, Zn(4abIm)2TPIBP; down-triangle, Zn(Im)2TPIBP.
diffusion-limited current density (jD) at 0.60 V (vs RHE) on the
BET SSA. Because our catalysts are uniform without the
distinction between the catalytic and the support surfaces, the
higher surface area affords a greater number of exposed N-
containing active sites and, consequently, a higher current
density. Figure 3C indicates that micropore volume has a
negative relationship with Tafel slope and a positive relationship
with E1/2, respectively. This figure indicates two important
implications of microporosity for ORR kinetics. Because
micropores dominate the majority fraction of the total pore
volume and hence the SSA in ZIF-derived catalysts,16 higher
micropore volume provides more exposed active sites therefore
higher kinetic current and lower Tafel slope under the same
amount of catalyst, for the similar reason described for the
limiting current study in Figure 3B. Half-wave potential E1/2 is
associated with both reaction barrier (E0) and oxygen transport.
The catalyst with higher micropore volume will entrap and
concentrate more solvated O2, leading to lower reaction barrier
and upshift of E1/2 as well as E0 (Table 1). Figure 3D reveals the
positive correlation between the catalyst durability and its
graphitization level. Higher carbonization leads to less
amorphous catalyst of lower SSA (Figure S18), which tends to
be more stable but less active (Figure 3B,C). The Fe/N4/C type
ORR catalyst is prone to deactivation in the acidic media from
the demetalation and peroxide oxidation.37−39 The former is
prompted by the proton binding to the carbonaceous nitrogen
leading to deligation of iron, and the latter is caused by the
oxidation of the catalyst framework by peroxide formed during
ORR. In the alkaline medium, demetalation becomes a lesser
issue because of high pH. Oxidative corrosion by the peroxide,
on the other hand, becomes an overriding cause. Attack by
peroxide is more prevalent on the amorphous region, such as at
the edge of the graphene plane. The retention of the ORR
activity after 10 000 voltage cycles shown in Figure 3D suggests
that higher IG/ID ratio, or higher graphitization, led to better
catalyst durability at the expense of lower surface area.
Therefore, the rational design and processing of the Fe/N/C
composite require a trade-off between SSA and graphitization.
To better understand the impact of imidazolate ligands on the
surface and catalytic properties during thermal activation, we
investigated the pyrolysis chemistry of the four ZIFs using
thermogravimetric analysis (TGA) in Ar atmosphere coupled
with mass spectroscopy (MS). ZIFs, particularly Zn(mIm)2, are
now frequently used in preparing various porous electro-
chemical and electrocatalytic materials through thermolysis.12
However, little is known about the chemistry of ZIF ligand-to-
carbon transformation under the elevated temperature. Probing
decomposition products during the thermolysis could reveal the
conversion chemistry and guide the catalyst optimization
process. Figure 4A shows the weight loss of all four ZIFs as
the function of pyrolysis temperature. The weight-loss of
Zn(eIm)2TPIB starts at the lowest temperature ca. 500 °C,
suggesting it as the least stable among the four. Multiple mass
fragments were detected simultaneously, indicating several
conversion reactions occurring at the same time (Figure 4B−
F). For example, hydrogen (mass = 2, Figure 4B) produced from
dehydrogenation reaction, starting just below 500 °C and
completing slightly above 800 °C, was detected on all four
samples, as is expected during the organics-to-carbon con-
version. C1 groups such as methyl (mass = 15, Figure 4C) and
methane (mass = 16, Figure S19A) were found from
Zn(eIm)2TPIB (560 °C); Zn(mIm)2TPIB (600 °C); and, to
much lesser degree, Zn(Im)2TPIB (650 °C). These groups were
most likely formed from the breaking of the methyl group in
Zn(mIm)2TPIB or C−C bond of the ethyl group in
Zn(eIm)2TPIB. Because ethyl C−C bond breaking can lead to
rearrangement of α carbon with imidazolate to form a more
stable 6-member ring in the graphitic plane, the process should
2504
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
be more facile and therefore contributes to the higher methyl
production by Zn(eIm)2TPIB over Zn(mIm)2TPIB. A small
amount of methyl fragments from Zn(Im)2TPIB can be
generated only from decomposition of imidazole ring because
of its lack of side chain. No methyl was detected from
Zn(4abIm)2TPIB, suggesting that the imidazole is stabilized
by the adjacent azabenzyl group. Similarly, C2 alkyl group such
as ·CH2−CH3 (mass = 29) (Figure 4F) or C2H2 (mass = 26)
(Figure S19B) were also detected, with Zn(eIm)2TPIB being
the most dominating contributor at lower temperature among
all four samples. This can be attributed to detachment of the
ethyl group from imidazolate during pyrolysis. Note that both
C1 and C2 productions are within a relatively narrow
temperature window between 500 °C to 700 °C.
activity of Zn(Im)2TPIBP. For the other three PGM-free
precursors, TGA shows that N2 emerged at ca.700 °C and
continually increases with temperature after removal of the
ethylene background using ethenyl as reference. N2 evolutions
are less facile in Zn(mIm)2TPIB, Zn(eIm)2TPIB, and Zn-
(4abIm)2TPIB compared to Zn(Im)2TPIB, which may be
attributed to a competitive path of rearrangement of the
imidazole ring with the side chains informing amorphous
carbon. For these precursors, however, the nitrogen loss
transitioned from ammonia to dinitrogen with the increase of
temperature. Investigation of the temperature window for N loss
and hydrocarbon fragments revealed the mechanistic insight of
the decomposition chemistry of different imidazolates in ZIFs
promoted by the presence of iron from TPI. We should point
out, however, that we could not identify any fragmented piece
from TPI because of its relatively low concentration in the
precursors.
In summary, four TPI-added ZIFs containing imidazolate
ligands of different side chains (HIm, HmIm, HeIm, and
H4abIm) were prepared through a solid-state synthesis,
followed by thermolysis to produce Fe/N/C ORR catalysts.
The side-chain substitution clearly affects the key structure and
composition factors of the final catalyst such as pyridinic-N,
SSA, total micropore volume, graphitization level, etc., all of
which impact the ORR activity and durability. Additionally, we
investigated the imidazolate-to-carbon conversion during the
thermolysis and observed a number of temperature-dependent
fragmentations associated with the key constituents of the final
catalyst, such as nitrogen content. This study provides several
findings that can be applied in the rational design, synthesis, and
optimization for ZIF-derived materials in energy storage and
conversion applications.
In addition to C1 and C2 groups, we also observed higher
molecular weight fragments during pyrolysis. For example, we
detected the fragments of mass = 44 at temperature above 550
°C, which can be attributed to C3H8 (Figure S19C). This
fragment showed the highest intensity from Zn(Im)2TPIB
compared to other samples. Its emission intensity and
temperature patterns match well with that of N2 fragment
(discussed below), suggesting a concerted decomposition of the
imidazolate ring through which formation of dinitrogen triggers
the coproduction of propane. Because imidazolate in Zn-
(Im)2TPIB does not have any sacrificial side chain, the
decomposition causes the complete fragmentation of the
aromatic ring to smaller hydrocarbon pieces, which subse-
quently reassimilate into the carbon with higher graphitization
level and less nitrogen content. This explains why we observed a
higher IG/ID ratio and the highly graphitized carbon species such
as carbon nanotubes in pyrolyzed Zn(Im)2TPIB than in others.
Another important fact of higher molecular fragmentation is that
the process continued with the rise of temperature up to 1000
°C, suggesting a continuous conversion before the carbonization
completed. It is noteworthy to point out that we did not observe
any fragment with the mass greater than 44 in our study.
Because the carbonaceous nitrogen content in the PGM-free
catalyst directly correlates to the ORR activity, knowing their
formation kinetics is crucial in controlling and optimizing the
catalyst synthesis. Nitrogen embedded in carbon is metastable.
Retaining a higher level of carbonaceous-N while improving the
conductivity through graphitization requires the balance of
thermolysis temperature and time. Main decomposition
products containing nitrogen include ammonia (mass = 17)
and N2 (mass = 28), both of which were observed in our
experiment. Figure 4D shows the NH3 emission profile with
Zn(eIm)2TPIB having the lowest and Zn(4abIm)2TPIB the
highest onset temperature. The temperature band is relatively
broad, from 500 °C to over 900 °C, suggesting a continuous
process. The N2 signal, however, is convoluted by the presence
of ethylene, which has the same mass (Figure 4E). To separate
ethylene’s contribution, we resorted to the signal of ethenyl
(mass = 27, Figure 4E inset) as an indicator because they
appeared conjointly in the MS spectra. For example, the sharp
peak observed for Zn(eIm)2TPIB at temperature below 600 °C
is most likely due to ethylene because it was found in both
spectra. Similarly matching peaks observed between 600 °C to
800 °C of all the samples in both spectra can also be attributed to
ethylene except for Zn(Im)2TPIB, of which the peak intensity of
mass 28 is significantly higher than that of mass 27 at 800 °C.
This further confirms the aforementioned concerted formation
of N2 and C3H8 from fragmentation of the imidazolate ring in
Zn(Im)2TPIB. The massive N loss gave rise to poor ORR
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acsenergy-
lett.9b01740.
Experimental details, electrochemical test results, physical
characterization results, structure−property relationship
of the studied materials, and ORR activity comparison
with the literature (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: djliu@anl.gov (D.-J.L.).
ORCID
Hao Wang: 0000-0003-0674-0811
Dan Zhao: 0000-0002-4427-2150
Biao-Hua Chen: 0000-0002-9871-9560
Di-Jia Liu: 0000-0003-1747-028X
Present Addresses
⊥H.W.: National Renewable Energy Laboratory, 15013 Denver
West Parkway, Golden, CO 80401.
∇L.R.G.: Pittsburgh Glass Works, 30 Isabella St #500,
Pittsburgh, PA 15212.
○H.M.B.: Wisconsin Department of Transportation, 3502
Kinsman Blvd., Madison, WI 53704.
Author Contributions
∥H.W. and L.R.G. contributed to this work equally.
2505
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
(14) Zhao, S.; Yin, H.; Du, L.; He, L.; Zhao, K.; Chang, L.; Yin, G.;
Zhao, H.; Liu, S.; Tang, Z. Carbonized nanoscale metal−organic
frameworks as high performance electrocatalyst for oxygen reduction
reaction. ACS Nano 2014, 8 (12), 12660−12668.
(15) Qian, Y.; Cavanaugh, J.; Khan, I. A.; Wang, X.; Peng, Y.; Hu, Z.;
Wang, Y.; Zhao, D. Fe/Fe3C/N-Doped Carbon Materials from Metal−
Organic Framework Composites as Highly Efficient Oxygen Reduction
Reaction Electrocatalysts. ChemPlusChem 2016, 81 (8), 718−723.
(16) Zhao, D.; Shui, J. L.; Grabstanowicz, L. R.; Chen, C.; Commet, S.
M.; Xu, T.; Lu, J.; Liu, D. J. Highly Efficient Non-Precious Metal
Electrocatalysts Prepared from One-Pot Synthesized Zeolitic Imidazo-
late Frameworks. Adv. Mater. 2014, 26 (7), 1093−1097.
(17) Wang, X.; Zhang, H.; Lin, H.; Gupta, S.; Wang, C.; Tao, Z.; Fu,
H.; Wang, T.; Zheng, J.; Wu, G.; et al. Directly converting Fe-doped
metal−organic frameworks into highly active and stable Fe-NC
catalysts for oxygen reduction in acid. Nano Energy 2016, 25, 110−119.
(18) Zhang, C.; Wang, Y. C.; An, B.; Huang, R.; Wang, C.; Zhou, Z.;
Lin, W. Networking pyrolyzed zeolitic imidazolate frameworks by
carbon nanotubes improves conductivity and enhances oxygen-
reduction performance in polymer-electrolyte-membrane fuel cells.
Adv. Mater. 2017, 29 (4), 1604556.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by U.S. Department of Energy, Fuel
Cell Technologies Office through Office of Energy Efficiency
and Renewable Energy. The works performed at Argonne
National Laboratory’s Center for Nanoscale Materials, a U.S.
Department of Energy Office of Science User Facility, is
supported by Office of Science, U.S. Department of Energy
under Contract DE-AC02-06CH11357. Authors thank Pine
Research Instrumentation for providing the alkaline resistant
electrochemical cell for this experiment. The views and opinions
of the authors expressed herein do not necessarily state or reflect
those of the United States Government or any agency thereof;
neither the United States Government nor any agency thereof,
nor any of their employees, makes any warranty, expressed or
implied, or assumes any legal liability or responsibility for the
accuracy, completeness, nor usefulness of any information,
apparatus, product, or process disclosed, or represents that its
use would not infringe privately owned rights.
(19) Morozan, A.; Jaouen, F. Metal organic frameworks for
electrochemical applications. Energy Environ. Sci. 2012, 5 (11),
9269−9290.
̂
́
(20) Park, K. S.; Ni, Z.; Cote, A. P.; Choi, J. Y.; Huang, R.; Uribe-
Romo, F. J.; Chae, H. K.; O’Keeffe, M.; Yaghi, O. M. Exceptional
chemical and thermal stability of zeolitic imidazolate frameworks. Proc.
Natl. Acad. Sci. U. S. A. 2006, 103 (27), 10186−10191.
REFERENCES
■
(1) Jacobson, M.; Colella, W.; Golden, D. Cleaning the air and
improving health with hydrogen fuel-cell vehicles. Science 2005, 308
(5730), 1901−1905.
(21) Tan, J. C.; Bennett, T. D.; Cheetham, A. K. Chemical structure,
network topology, and porosity effects on the mechanical properties of
Zeolitic Imidazolate Frameworks. Proc. Natl. Acad. Sci. U. S. A. 2010,
107 (22), 9938−9943.
(22) Armel, V.; Hindocha, S.; Salles, F.; Bennett, S.; Jones, D.; Jaouen,
F. Structural descriptors of zeolitic−imidazolate frameworks are keys to
the activity of Fe−N−C catalysts. J. Am. Chem. Soc. 2017, 139 (1),
453−464.
(23) Zitolo, A.; Goellner, V.; Armel, V.; Sougrati, M.-T.; Mineva, T.;
Stievano, L.; Fonda, E.; Jaouen, F. Identification of catalytic sites for
oxygen reduction in iron-and nitrogen-doped graphene materials. Nat.
Mater. 2015, 14 (9), 937.
(24) Hayashi, H.; Cote, A. P.; Furukawa, H.; O’Keeffe, M.; Yaghi, O.
M. Zeolite A imidazolate frameworks. Nat. Mater. 2007, 6 (7), 501.
(25) Cao, S.; Bennett, T. D.; Keen, D. A.; Goodwin, A. L.; Cheetham,
A. K. Amorphization of the prototypical zeolitic imidazolate framework
ZIF-8 by ball-milling. Chem. Commun. 2012, 48 (63), 7805−7807.
(26) Bennett, T. D.; Cao, S.; Tan, J. C.; Keen, D. A.; Bithell, E. G.;
Beldon, P. J.; Friscic, T.; Cheetham, A. K. Facile mechanosynthesis of
amorphous zeolitic imidazolate frameworks. J. Am. Chem. Soc. 2011,
133 (37), 14546−14549.
(2) Das, V.; Padmanaban, S.; Venkitusamy, K.; Selvamuthukumaran,
R.; Blaabjerg, F.; Siano, P. Recent advances and challenges of fuel cell
based power system architectures and control−A review. Renewable
Sustainable Energy Rev. 2017, 73, 10−18.
(3) Berger, D. J. Fuel cells and precious-metal catalysts. Science 1999,
286 (5437), 49.
(4) Debe, M. K. Electrocatalyst approaches and challenges for
automotive fuel cells. Nature 2012, 486 (7401), 43−51.
(5) Jasinski, R. A new fuel cell cathode catalyst. Nature 1964, 201
(4925), 1212−1213.
̀
(6) Proietti, E.; Jaouen, F.; Lefevre, M.; Larouche, N.; Tian, J.;
Herranz, J.; Dodelet, J.-P. Iron-based cathode catalyst with enhanced
power density in polymer electrolyte membrane fuel cells. Nat.
Commun. 2011, 2, 416.
(7) Shui, J.; Chen, C.; Grabstanowicz, L.; Zhao, D.; Liu, D.-J. Highly
efficient nonprecious metal catalyst prepared with metal−organic
framework in a continuous carbon nanofibrous network. Proc. Natl.
Acad. Sci. U. S. A. 2015, 112 (34), 10629−10634.
(8) Zitolo, A.; Goellner, V.; Armel, V.; Sougrati, M.-T.; Mineva, T.;
Stievano, L.; Fonda, E.; Jaouen, F. Identification of catalytic sites for
oxygen reduction in iron-and nitrogen-doped graphene materials. Nat.
Mater. 2015, 14 (9), 937−942.
(9) Chung, H. T.; Cullen, D. A.; Higgins, D.; Sneed, B. T.; Holby, E.
F.; More, K. L.; Zelenay, P. Direct atomic-level insight into the active
sites of a high-performance PGM-free ORR catalyst. Science 2017, 357
(6350), 479−484.
(10) Li, J.; Ghoshal, S.; Liang, W.; Sougrati, M.-T.; Jaouen, F.; Halevi,
B.; McKinney, S.; McCool, G.; Ma, C.; Yuan, X.; et al. Structural and
mechanistic basis for the high activity of Fe−N−C catalysts toward
oxygen reduction. Energy Environ. Sci. 2016, 9 (7), 2418−2432.
(11) Ma, S.; Goenaga, G. A.; Call, A. V.; Liu, D. J. Cobalt imidazolate
framework as precursor for oxygen reduction reaction electrocatalysts.
Chem. - Eur. J. 2011, 17 (7), 2063−2067.
(12) Barkholtz, H. M.; Liu, D.-J. Advancements in rationally designed
PGM-free fuel cell catalysts derived from metal−organic frameworks.
Mater. Horiz. 2017, 4 (1), 20−37.
(13) Zhao, D.; Shui, J.-L.; Chen, C.; Chen, X.; Reprogle, B. M.; Wang,
D.; Liu, D.-J. Iron imidazolate framework as precursor for electro-
catalysts in polymer electrolyte membrane fuel cells. Chem. Sci. 2012, 3
(11), 3200−3205.
(27) Wu, G.; Mack, N. H.; Gao, W.; Ma, S.; Zhong, R.; Han, J.;
Baldwin, J. K.; Zelenay, P. Nitrogen-doped graphene-rich catalysts
derived from heteroatom polymers for oxygen reduction in nonaqueous
lithium−O2 battery cathodes. ACS Nano 2012, 6 (11), 9764−9776.
̀
(28) Kramm, U. I.; Herranz, J.; Larouche, N.; Arruda, T. M.; Lefevre,
M.; Jaouen, F.; Bogdanoff, P.; Fiechter, S.; Abs-Wurmbach, I.;
Mukerjee, S.; et al. Structure of the catalytic sites in Fe/N/C-catalysts
for O2-reduction in PEM fuel cells. Phys. Chem. Chem. Phys. 2012, 14
(33), 11673−11688.
(29) Kundu, S.; Nagaiah, T. C.; Xia, W.; Wang, Y.; Dommele, S. V.;
Bitter, J. H.; Santa, M.; Grundmeier, G.; Bron, M.; Schuhmann, W.;
et al. Electrocatalytic activity and stability of nitrogen-containing
carbon nanotubes in the oxygen reduction reaction. J. Phys. Chem. C
2009, 113 (32), 14302−14310.
(30) Rao, C. V.; Cabrera, C. R.; Ishikawa, Y. In search of the active site
in nitrogen-doped carbon nanotube electrodes for the oxygen reduction
reaction. J. Phys. Chem. Lett. 2010, 1 (18), 2622−2627.
(31) Cui, X.; Yang, S.; Yan, X.; Leng, J.; Shuang, S.; Ajayan, P. M.;
Zhang, Z. Pyridinic-nitrogen-dominated graphene aerogels with Fe-N-
2506
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507

ACS Energy Letters
Letter
C coordination for highly efficient oxygen reduction reaction. Adv.
Funct. Mater. 2016, 26 (31), 5708−5717.
(32) Serov, A.; Artyushkova, K.; Atanassov, P. Fe-N-C oxygen
reduction fuel cell catalyst derived from carbendazim: Synthesis,
structure, and reactivity. Adv. Energy Mater. 2014, 4 (10), 1301735.
(33) Lai, L.; Potts, J. R.; Zhan, D.; Wang, L.; Poh, C. K.; Tang, C.;
Gong, H.; Shen, Z.; Lin, J.; Ruoff, R. S. Exploration of the active center
structure of nitrogen-doped graphene-based catalysts for oxygen
reduction reaction. Energy Environ. Sci. 2012, 5 (7), 7936−7942.
(34) Wang, Y.-C.; Lai, Y.-J.; Wan, L.-Y.; Yang, H.; Dong, J.; Huang, L.;
Chen, C.; Rauf, M.; Zhou, Z.-Y.; Sun, S.-G. Suppression Effect of Small
Organic Molecules on Oxygen Reduction activity of Fe/N/C Catalysts.
ACS Energy Letters 2018, 3, 1396−1401.
(35) Wang, Q.; Ji, Y.; Lei, Y.; Wang, Y.; Wang, Y.; Li, Y.; Wang, S.
Pyridinic-N-dominated doped defective graphene as a superior oxygen
electrocatalyst for ultrahigh-energy-density Zn−air batteries. ACS
Energy Lett. 2018, 3 (5), 1183−1191.
(36) Artyushkova, K.; Serov, A.; Rojas-Carbonell, S.; Atanassov, P.
Chemistry of multitudinous active sites for oxygen reduction reaction in
transition metal−nitrogen−carbon electrocatalysts. J. Phys. Chem. C
2015, 119 (46), 25917−25928.
(37) Varnell, J. A.; Sotiropoulos, J. S.; Brown, T. M.; Subedi, K.;
Haasch, R. T.; Schulz, C. E.; Gewirth, A. A. Revealing the role of the
metal in non-precious-metal catalysts for oxygen reduction via selective
removal of Fe. ACS Energy Lett. 2018, 3 (4), 823−828.
(38) Wang, H.; Yin, F. X.; Liu, N.; Kou, R. H.; He, X. B.; Sun, C. J.;
Chen, B. H.; Liu, D. J.; Yin, H. Q. Engineering Fe−Fe3C@Fe−N−C
active sites and hybrid structures from dual metal−organic frameworks
for oxygen reduction reaction in H2−O2 fuel cell and Li−O2 battery.
Adv. Funct. Mater. 2019, 29, 1901531.
̀
(39) Chenitz, R.; Kramm, U. I.; Lefevre, M.; Glibin, V.; Zhang, G.;
Sun, S.; Dodelet, J.-P. A specific demetalation of Fe−N4 catalytic sites in
the micropores of NC_Ar+ NH3 is at the origin of the initial activity loss
of the highly active Fe/N/C catalyst used for the reduction of oxygen in
PEM fuel cells. Energy Environ. Sci. 2018, 11 (2), 365−382.
2507
DOI: 10.1021/acsenergylett.9b01740
ACS Energy Lett. 2019, 4, 2500−2507