Journal of the American Chemical Society
Article
1
0−13,16−19
iron-based catalysts.
However, because of the
7890A GC with a 5975C mass-selective detector) was used to study
the synthetic mechanism.
relatively harsh conditions used in preparing iron carbide
NPs, typical FTS processes use iron oxide NPs as the catalyst
precursors. The catalysts are treated in CO/syngas, at the end
of which the oxide NPs are proposed, with controversy, to be
transformed into the catalytically active phase, iron carbides,
2.3. Catalyst Preparation. To test the catalytic performance of
Fe C NPs, 80 mg of as-prepared Fe C NPs dispersed in ethanol was
5
2
5 2
impregnated with SiO support and dried at room temperature. A
2
supported Fe O catalyst was prepared by the impregnation method:
2
3
an ethanol solution of iron nitrate was added to the SiO support,
2
1
1,14,35
normally as the mixture of various iron carbide phases.
In
which was then dried and calcined at 433 K in air. For both catalysts,
addition, it was reported that the activity of iron-based catalyst
remained almost constant for particle sizes larger than 6 nm,
and thus, the intrinsic activity of the iron catalyst is considered
the loading of Fe was ∼9 wt %.
2.4. FTS Reaction Conditions. The FTS reaction was carried out
on a fixed-bed flow reactor with a gas mixture containing 32% CO,
1
4,36
63% H , and 5% Ar at a temperature of 543 K. An 80 mg sample of the
to be related more to the phase than to the surface area.
2
supported catalyst with ∼9 wt % iron was loaded in a stainless steel
tube lined with a quartz layer. The gas hourly space velocity (GHSV)
Therefore, it would be desirable to synthesize single-phase
Fe C in order to verify its catalytic activity in FTS.
5
2
3
−1
−1
of the reaction was set at 15 000 cm h gcat . For the supported
Fe C catalyst, no H or CO activation process was adapted. The
Herein we report a facile chemie douce route for the
5
2
2
synthesis of Fe C NPs that involves the reaction of iron
5
2
pressure of the reactor was set at 3 MPa, and the temperature was
ramped from 303 to 543 at 3 K/min (the products can be detected
immediately as long as the aim temperature was reached). The
carbonyl, Fe(CO) , with octadecylamine in the presence of
5
bromide under mild temperatures (up to 623 K). The size of
the iron carbide NPs can be tuned by tailoring the
supported Fe O catalyst was first reduced in H at 653 K under 0.1
2
3
2
MPa for 16 h prior to the FTS reaction. When the temperature was
reduced to 543 K, the gas flow was switched to syngas, and the
pressure was raised to 3 MPa to begin the reaction.
The temperature-programmed surface reaction (TPSR) experi-
ments were performed on the same fixed-bed reactor under the same
conditions as for the FTS reaction. The temperature was raised from
303 to 543 K at 3 K/min and then kept at 543 K. The gas flow of the
reactor was analyzed using a mass spectrometer (Hiden HPR20). We
concentration of Fe(CO) . Most interestingly, the as-
5
synthesized Fe C NPs were used as an FTS catalyst, and the
5
2
results demonstrated that Fe C possesses higher activity and
5
2
selectivity than a conventional reduced-hematite catalyst. More
importantly, the induction period observed with conventional
iron oxide catalysts was not observed with this catalyst, which
clearly indicates that Fe C is an active phase for FTS.
5
2
detected the mass numbers 2 for H , 15 for methane, 18 for water, 28
2
for CO, 41 for propene (to represent the hydrocarbon products), and
2. EXPERIMENTAL SECTION
4
4 for CO2.
A classic hot injection strategy was employed to synthesize the Fe C2
NPs. In a typical procedure, octadecylamine was used as both the
solvent and surfactant, while cetyltrimethylammonium bromide
2.5. Product Analysis. The product and reactant in the gas phase
5
were detected online using an Agilent 6890 GC. C −C -ranged
1
4
alkanes were analyzed using a Plot Al O capillary column with a flame
2
3
(CTAB) was used as the inducing agent and Fe(CO)5 as the
ionization detector (FID); CO, CO , CH , and Ar were analyzed using
2 4
precursor. Notably, various kinds of bromides were applicable as the
inducing agent (Figure S1 in the Supporting Information).
a Porapak Q- and 5A molecular sieve-packed column with a thermal
conductivity detector. The 5% Ar in the syngas was used as an internal
standard for the calculation of CO conversion. The product with large
molecular weight was collected in a cold trap. Hydrocarbons were
analyzed using a 6820 GC with an HP-5 capillary column and an FID,
and oxygenates in water were analyzed using a 6820 GC with an HP-
INNOWax capillary column and an FID. The selectivity of the
products was all on a carbon basis.
2.1. Synthesis of 20 nm Fe5C2 NPs. In a four-neck flask, a
mixture of octadecylamine (14.5 g) and CTAB (0.113 g) was stirred
sufficiently and degassed under a flow of N . The mixture was heated
2
to 393 K, and then Fe(CO) (0.5 mL, 3.6 mmol) was injected under a
5
N blanket. The mixture was heated to 453 K at 10 K/min and kept at
2
this temperature for 10 min. A color change from orange to black was
observed during the process, implying the decomposition of Fe(CO)5
and the nucleation of Fe nanocrystals. Subsequently, the mixture was
further heated to 623 at 10 K/min and kept there for 10 min before it
was cooled to room temperature. The product was washed with
ethanol and hexane and collected for further characterization. The
synthetic procedure for 10 nm NPs was similar to that for 20 nm
Fe C NPs except that the Fe(CO) was dissolved in 2.5 mL of hexane
3
. RESULTS AND DISCUSSION
3.1. Fe Morphology and Structure. The size and
C
5 2
morphology of the samples were characterized by TEM. Figure
1a shows that the Fe C NPs were ∼20 nm in diameter (Figure
5
2
5
2
5
S2a). The HRTEM image of an isolated 20 nm Fe C NP
5 2
before injection. The as-synthesized NPs were kept in an Ar-filled
glovebox to avoid exposure to air before further characterization.
(
Figure 1b) reveals the core−shell structure. The lattice spacing
in the core was 0.205 nm, corresponding to the (510) plane of
Fe C , while the shell structure appeared to be amorphous.
2.2. Characterization. Transmission electron microscopy (TEM)
5
2
was carried out on an FEI Tecnai T20 microscope. High-resolution
TEM (HRTEM) was carried out on an FEI Tecnai F20 microscope.
X-ray diffraction (XRD) patterns were obtained using a Rigaku
DMAX-2400 X-ray diffractometer equipped with Cu Kα radiation.
The accelerating voltage and current were 40 kV and 100 mA,
respectively. Extended X-ray absorption fine structure (EXAFS) was
characterized on beamline BL14W1-XAFS at the Shanghai Synchro-
tron Radiation Facility (SSRF), with the storage ring being operated at
Moreover, the particle size could be tuned by changing the
concentration of Fe(CO) in the mixture (Figure S2b,c). The
5
structure of Fe C NPs was verified by XRD analysis. Figure 1c
5
2
shows the XRD pattern of iron carbide NPs (average diameter
of 20 nm), which is consistent with that of Fe C (JCPDS no.
5
2
36-1248). The average particle size estimated from the Scherrer
equation was 23.1 nm, which is consistent with the statistical
analysis in Figure S2a. EXAFS was used to investigate the fine
structure of the samples. For verification, the theoretical Fourier
transformed EXAFS (FT-EXAFS) pattern of the first two
coordination shells of Fe C was fitted using the program
3.5 GeV and 300 mA. X-ray photoelectron spectroscopy (XPS)
measurements were carried out on an Axis Ultra imaging photo-
electron spectrometer (Kratos Analytical Ltd.) using a monochrom-
atized Al Kα anode, and the C 1s peak at 284.8 eV was taken as an
internal standard. Raman spectroscopy was recorded on a Renishaw
5
2
IFEFFIT. In Figure 2a, the gray fitted line is a joint contribution
of Fe−C scattering in the first Fe−C coordination shell (red
circles) and Fe−Fe scattering in the first Fe−Fe coordination
shell (blue triangles). The experimental curve is shown in
1000 Raman imaging microscope system with an excitation wave-
length of 632.8 nm. Fourier transform IR spectroscopy (FTIR) was
carried out on a Nicolet Magna-IR 750 FTIR spectrometer. A gas
chromatography−mass spectrometry (GC−MS) instrument (Agilent
B
dx.doi.org/10.1021/ja305048p | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX