Inorganic Chemistry
Article
2
1
air contamination. The H NMR spectra demonstrate only
15NH4 was observed when using 15N2 as the feeding gas
(Figure S6). The static experiment with an intermittent N2
supply reveals the high dependence of the reduction current
highly linear relationship between the ammonia production
and reaction time. These results convincingly confirm the NH3
is produced from N2 electrochemical reduction on ERGO-0.4
in the ambient environment.
eV, which are attributed to the C−C, C−O, CO, and O−
CO bonds, respectively.32 The CO/C−C ratio decreases
from 1.18 to 1.08 and further to 0.22 when the reduction
potential changes from −0.2 to −0.4 and −0.6 V.
Correspondingly, the O/C atomic ratio decreases from 0.43
to 0.40 and further to 0.18. The relative O−H peak intensity
gets smaller as the applied potential turns from −0.2 to −0.4 V
signal becomes significantly less when the reduction potential
increases from −0.4 to −0.6 V (Figure S15). Combing this
with the above electrocatalytic performance results, we
suppose the CO group is the active site for NRR activity
on GO. In addition, Raman and FTIR spectra are collected for
further vibration information analysis. In Figure 3e, the peaks
at 1320 and 1590 cm−1 belong to the characteristic D band and
G band of carbon materials, representing the defects of carbon
lattice atoms and the plane vibration of ordered graphite
structure, respectively.33 The ID/IG ratios for ERGO-0.4 and
ERGO-0.6 are 1.31 to 1.26, revealing more oxygen species are
reduced under a higher reducing potential. They are both
substantially larger than pure GO nanosheets (Figure S16),
indicating the decrease in size of the in-plane sp2 domains.31
Fourier transform infrared (FTIR) spectra confirm that the
majority of the CO groups located at 1796 cm−1 are reduced
(Figure 3f).34,35 These are consistent with the above XPS
discussions.
To distinguish the effects of an individual oxygenated
functional group on N2 reduction and identify the actual sites
for NH3 generation, we used the chronoamperometry strategy
at different potentials to prepare ERGO with controllable
oxygenated groups. Figure 3a shows the original peak intensity
Density functional theory (DFT) calculations are further
performed to identify the potential of individual oxygen species
for NRR and to understand the underlying reaction
mechanism. As the first starting step, efficient molecular N2
fixation and activation are the prerequisites for following the
ammonia synthesis pathway. A variety of N2 adsorption
configurations on the graphene-based substrates is constructed.
After optimization, it is found the N2 molecule is far from the
pristine graphene surface with the distance larger than 3 Å. The
similar phenomenon is observed on the graphene sheets with
OH and O−C−O modifications, manifesting that the N2
molecules cannot be adsorbed on a graphene surface with
these oxygenated group modifications (Figure S17). In
contrast, Figure 4a shows the CO group on graphene can
interact with N2 tightly, with the two N atoms combining with
the C and O atoms, respectively, in the side-on adsorption
configuration. The average distance between the N2 and CO
groups is 1.45 Å, indicating a strong adsorption at the interface.
Figure 4b suggests the NN bond length is elongated from
1.10 Å of free N2 molecular to 1.24 Å in the adsorbed case,
accompanied by the CO bond length increasing from 1.22
to 1.46 Å.
Meanwhile, the vibration frequency of the CO bond
decreases from 1769.4 to 1768.9 cm−1 after the N2 adsorption.
Besides, the charge density difference (CDD) calculation
shown in Figure 4c presents a two-way charge transfer process
for the adsorption with charge accumulation and depletion
simultaneously occurring on the adsorbed N2 molecule. This
phenomenon is in accordance with the “acceptance-donation”
electron reservoir for NRR as described before.36 Figure 4d
shows the corresponding projected density of states (PDOS)
calculation, in which the CO group provides more
hybridization with N2. These are consistent with our
experimental observations, illustrating that CO plays a
critical role for N2 fixation and activation. It is noted that the
COOH group also contains the CO counterpart, and Figure
S18 shows COOH can also combine with N2 strongly, further
Figure 3. (a) CV curves on a series of ERGO materials. (inset) The
magnification of CV curves measured on ERGO-0.6. (b) NH3 yield
comparison between ERGO-0.2 (three times), ERGO-0.4 (five
times), and ERGO-0.6 (three times). XPS spectra of C 1s for (c)
ERGO-0.4 and (d) ERGO-0.6. (e) Raman and (f) FTIR spectra for
ERGO-0.4 and ERGO-0.6.
will become smaller when a larger reduction potential is
applied. As the potential increases to −0.6 V, this reduction
peak gets flat and almost disappears (inset in Figure 3a). The
electrochemically active surface area (ECSA) is evaluated by
running the CV at different scanning rates in the double-layer
potential region. Figure S10 illustrates that ERGO-0.6
possesses a much smaller charging/discharging capacitance
when compared with ERGO-0.4, which confirms the
realization of a controllable and partial reduction of the
surface oxygen species. The NRR performance on ERGO-0.2
3b compares the NH3 yields of ERGO-0.2, -0.4, and -0.6, all of
which are collected from at least three individual measure-
ments based on separated electrodes. ERGO-0.4 attains a
comparable performance as that of ERGO-0.2, while ERGO-
0.6 exhibits significantly reduced NRR activity. Electro-
chemical impedance spectroscopy (EIS) data present an
apparently smaller semicircle on ERGO-0.4 than on ERGO-
0.6, confirming the significant activity decrease from −0.4 to
−0.6 V reduction (Figure S13). These results indicate the
changes in surface structure and chemical environment from
−0.4 to −0.6 V reduction are essential to determine the NRR
activity.
X-ray photoelectron spectroscopy (XPS) measurement is
further conducted to investigate the chemical composition
change and environment variation during the potential-
dependent electroreduction process. For the C 1s spectra as
deconvoluted peaks located at 284.8, 285.6, 287.2, and 288.8
C
Inorg. Chem. XXXX, XXX, XXX−XXX