Study on the Simple Surface Treatments of N, P Dual-doped Carbon as Metal-free Catalyst for Metal-air Batteries

Article abstract of DOI:10.1002/cctc.202001319

N, P dual-doped porous carbon with high specific surface area of ～2000 m² g^?1 was prepared via foaming-carbonization process. It was demonstrated that the simple surface treatment and modification can conveniently further improve the catalytic performance of oxygen reduction reaction (ORR). The results showed that chemisorbed phosphorus can promote the surface treatment of sodium borohydride (NaBH₄). The modified novel sites, adsorbing water molecules of a semi-free state, promoted the offensive key steps of H₂O molecules in ORR. After post-treatment, the limit current density increased from 5.0 to 6.1 mA cm^?2 with a decreased Tafel slope and the onset potential positively shifted. This surface-modified catalyst was applied to Zn-air battery and Al-air battery, and had exhibited good applicability and excellent performances. This work illustrates that the simple post-treatments can conveniently and effectively improve the ORR performance of metal-free carbon catalysts.

Full text of DOI:10.1002/cctc.202001319

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Study on the Simple Surface Treatments of N, P Dual-doped
Carbon as Metal-free Catalyst for Metal-air Batteries
Authors: Heping Yu, Hui Zhang, and Zhongyi Zhang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202001319
Link to VoR: https://doi.org/10.1002/cctc.202001319
01/2020

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
Study on the Simple Surface Treatments of N, P Dual-doped
Carbon as Metal-free Catalyst for Metal-air Batteries
Heping Yu, Hui Zhang, Zhongyi Zhang*
H.Yu, Prof. H. Zhang, Prof. Z. Zhang*
College of Chemistry and Chemical Engineering
Qingdao University
Qingdao, Shandong, 266071, People’s Republic of China
E-mail: Zhangzy@qdu.edu.cn
any further post-treatments process.[13,31-36] The post-treatment
Abstract: N, P dual-doped porous carbon with high specific surface
area of ~2000 m2 g-1 was prepared via foaming-carbonization process.
method which can further improve the ORR performance without
It was demonstrated that the simple surface treatment and
modification can conveniently further improve the catalytic
introducing transition metal elements is worthy of researchers'
attention. For example, the subtraction strategies such as
cleaning the catalyst surface with suitable organic solvents may
better expose the ORR active sites. The addition strategies such
as the proper surface modification may promote the kinetic steps
of ORR. These post-treatments are still lack of researches,
although they have the potentiality for further improving ORR
performance.
Herein, P, N dual-doped porous carbon (PNC) was prepared
from cellulose through the foaming-carbonization processes,
creating an ultra-high specific surface area of about 2000 m2 g-1.
The post-treatments towards the carbonized product were
investigated, including solvent washing and surface modification.
Interestingly, after the NaBH4 treatment, the limiting current
density of ORR was increased from 5.0 mA cm-2 to 6.1 mA cm-2,
and the onset potential also positively shifted. Structural
characterizations confirmed that the surface was modified with
pyridinic-N-B(OH)3. The novel sites could adsorb water molecules
in a semi-free state, which can promote the kinetic steps in ORR
reaction. This surface-modified catalyst was applied to Zn-air
battery and Al-air battery, and has exhibited good applicability.
Therefore, it is strongly reminded that the performance of ORR
can be further improved simply through a convenient post-
treatment process.
performance of oxygen reduction reaction (ORR). The results showed
that chemisorbed phosphorus can promote the surface treatment of
sodium borohydride (NaBH4). The modified novel sites, adsorbing
water molecules of a semi-free state, promoted the offensive key
steps of H2O molecules in ORR. After post-treatment, the limit current
density increased from 5.0 to 6.1 mA cm-2 with a decreased Tafel
slope and the onset potential positively shifted. This surface-modified
catalyst was applied to Zn-air battery and Al-air battery, and had
exhibited good applicability and excellent performances. This work
illustrates that the simple post-treatments can conveniently and
effectively improve the ORR performance of metal-free carbon
catalysts.
Introduction
With the development of human society, fossil fuels have been
consumed in large quantities and become increasingly depleted.
It is urgent to develop alternative green energy sources and
devices to meet various needs.[1,2] Metal-air batteries such as
zinc-air batteries and aluminum-air batteries have exhibited
remarkable application potentials in the new generation of green
energy devices.[3-6] Oxygen reduction reaction (ORR) is the key
electrode reaction, which is critical to the performance of metal-
air batteries.[7-9] The important key point for future applications is
to reduce the material resource costs of the ORR catalyst as
much as possible. The traditional noble metal catalysts, such as
Pt based catalyst, are costly which limited their applications.[10-12]
At the same time, although some non-noble metal-based
catalysts are reported as alternatives, the ultimate choice of the
lowest cost is the metal-free catalyst. [13-18]
Heteroatom-doped carbon catalysts, as the metal-free ORR
catalysts, have been numerously studied.[8,16-20] The carbon
matrix doped with nitrogen, phosphorus, sulfur or boron has the
possibilities of possessing excellent ORR catalytic activity. Dual-
doping of heteroatoms was demonstrated to be more effective to
achieve high ORR activity compared to the nitrogen doping.[19-24]
Recently, in order to pursue much better ORR performance, a
number of transition metal elements were introduced again into
heteroatom-doped carbon matrix, in which the presence of metal-
heteroatom-carbon bonds can further prompt the ORR catalytic
activities.[25-27] However, the introduction of transition metal
elements weakens the ultimate cost advantage over metal-free
carbon-based catalysts. [28-31] After all transition metals are also
important resources, especially with the increasing depletion. In
addition, it has been ignored by researchers that almost all
reported ORR catalysts were used directly after synthesis without
Results and Discussion
The preparation strategy is shown in Figure 1a. Cellulose was
dissolved in phosphoric acid/polyphosphoric acid to form a
viscous solution. The C3N4 powder, as the nitrogen precursor,
was dispersed in this solution to form a yellowish-brown
o
dispersion. The mixture was foamed at 150 C to get the initial
foam (Figure S1).
This foam was carbonized under Ar/H2 atmosphere to obtain
the P, N dual-doped carbon (PNC). Various post-treatment
processes (Figure 1a) towards PNC were carried out, including
NaBH4 treatment, ethanol washing, ethanol washing + NaBH4
treatment, and the corresponding samples were named as PNC-
B, PNC-E and PNC-E/B, respectively. The X-ray diffraction
spectrum (XRD) and Raman spectra of PNC, PNC-E, PNC-B and
PNC-E/B are very similar (Figure 1b and Figure 1c), which have
expressed the characteristics of carbon materials. In the
meantime, Brunauer-Emmett-Teller (BET) measurements also
confirm that they have similar pore structure and specific surface
area (Figure 1d-g). The isotherms curves of them have similar
hysteresis loops. The specific surface areas of PNC, PNC-B,
PNC-E, and PNC-E/B are 2020.9, 1949.5, 1929.7, and 1911.6
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
Figure 1. (a) The schematic diagram of preparation strategy, (b) XRD spectra and (c) Raman spectra of the four samples. (d) PNC, (e) PNC-B, (f) PNC-
E, (g) PNC-E/B for N2 adsorption–desorption isotherms and corresponding Pore size distribution curves (inset), respectively.
m2 g-1, respectively. The ultra-high specific surface areas of them
are around 2000 m2 g-1, with a difference of less than 5%.
In fact, these samples also retain similar morphologies despite
undergoing different surface treatments. Herein, PNC-B was
taken to clarify their morphological features. The scanning
electron microscope (SEM) image (Figure 2a) of PNC-B exhibits
the appearance of porous carbon. Transmission electron
microscope (TEM) image (Figure 2b) suggests the interior of
PNC-B is full of dense porous structures. The high-resolution
transmission electron microscope (HTEM) image (Figure 2c)
exhibits that the PNC-B tends to be amorphous with many chaotic
stripes. These results imply a super-high specific surface area of
PNC-B evolved by such abundant pores. Figure 2d is a STEM
image of a certain area of PNC-B. The elemental distribution (in
Figure 2e) shows that the N, P and B elements are evenly
Figure 2. (a) SEM, (b)TEM and (c) HRTEM images of PNC-B, (d) STEM
distributed on the carbon background. The Figure 2f illustrates the
image and (e) EDS mapping of PNC-B, (f) EDS mapping of C and P for PNC
differences in the distribution of C and P elements in the PNC and
PNC-E. The dotted lines mark the boundary of the carbon
and PNC-E, respectively.
distribution of PNC and PNC-E, respectively. For PNC, the
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
distribution of C and P elements is consistent. For the sample
PNC-E of after ethanol washing, the P distribution became scarce
at the boundary of carbon. This appearance suggests that ethanol
may have dissolved and removed a part of P at the surface.
Perhaps not all the P elements contained in the PNC are
embedded in the carbon matrix.
approximately similar. In particular, the N/C ratio normalized to
carbon is almost equal to 0.02 for each material. This reflects that
the nitrogen element is incorporated into the carbon matrix, so
that its content is not changed by ethanol washing or NaBH4
treatment. However, this is not the case for the phosphorus
elements. As seen in Table 1, they can be divided into two groups
of different levels of phosphorus content. PNC and PNC-B have
similar phosphorus contents of ~1.40%, while PNC-E and PNC-
E/B have similar phosphorus contents of ~0.78%. The unanimous
decreasing in the phosphorus contents of PNC-E and PNC-E/B
demonstrates that ethanol washing can remove a portion of the
phosphorus while leaving other elements unchanged.
It is explained that a part of the phosphorus elements on the
surface is just chemisorbed. They are not actually doped
into the bulk phase of the PNC. What's more, it is interesting that
the boron content in PNC-B is significantly higher than that of
PNC-E/B. Thereby, the presence of chemisorbed phosphorus can
promote the surface modification of NaBH4 treatment.
Table 1 XPS element relative percentage content of different
materials
Atomic
percentage
PNC
C
O
N
N/C
P
P/C
B
at.%
--
at.%
at.% at.%
at.%
93.26 3.32
93.44 3.54
2.01 0.022 1.41 0.015
2.24 0.024 0.78 0.008
PNC-E
--
PNC-E/B
PNC-B
90.69 3.33
88.10 3.40
1.96 0.022 0.79 0.009 3.23
1.95 0.022 1.40 0.016 5.15
Therefore, XPS was used to further study the differences in
surface elements accurately. Table 1 provides the atomic
contents of different samples obtained from the XPS analysis. The
contents of C, O and N elements in all of these samples are
Figure 3. High-resolution XPS spectra of (a) P 2p, (b) B 1s and (c) N 1s.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
The high-resolution XPS spectra of the P 2p of the four samples
were shown in Figure 3a. The P 2p spectra center of PNC, PNC-
E, PNC-B, and PNC-B/E are located at 133.5, 133.4, 133.1, and
133.0 eV, respectively. Therefore, after NaBH4 treatment, the
spectra centers of the obtained PNC-B and PNC-B/E are
obviously shifted to lower binding energy for decreasing ~0.4 eV
compared to PNC and PNC-E (seen the inset of Figure 3a).
Commonly, the binding energy of P-C, P-N and P-O bonds are at
132.7 eV, 133.4 eV and 135.2 eV, respectively.[15,26,40-42]
Accordingly, the P 2p spectra of PNC and PNC-E were fitted, as
shown in Figure 3a. For PNC, the P element mainly exists in the
form of P-N bonds. After washed with ethanol, the P-N bond was
significantly weakened, which made the proportion of P-N bond
and P-C bond in PNC-E tend to be comparable. The decreasing
of P-N peak of PNC-E implies that the P atoms removed by
ethanol are mainly chemisorbed on the N atoms sites. However,
the situation is different from PNC-B, because its phosphorus
content is not decreased compared to PNC, as seen in Table 1.
Phosphorus, as a polyvalent element, can be chemisorbed in a
multi-binding states (bonded to N and C atoms simultaneously).
The way of NaBH4 treatment has weakened P-N bonds and
strengthened P-C bonds, which causes the spectra center of
PNC-B moves to lower binding energy, as well as the decreasing
of the P-N bond peak and enhancement of P-C peak. The two
cases ultimately lead to the highest P-C peak of PNC-E/B
compared to other samples.
between them. While, comparing the N spectra of PNC and PNC-
B, it is found that the shoulder peak of PNC-B at 395-400 eV is
significantly smaller than that of PNC. This area belongs to the
region of pyridinic-N and pyrrolic-N, which are located at the
edges or defects of the carbon matrix. The phenomenon
demonstrates the evidence that NaBH4 reacts with N species at
the edges or defects. The same difference also appears in the
comparison of PNC-E and PNC-E/B, but becomes less obvious
compared to the difference between PNC and PNC-B. This
indicates that the phosphorus atoms adsorbed on the surface of
PNC can promote the reaction between NaBH4 and the edge N
sites. This is because the steric hindrance of the N sites is very
big in the carbon matrix. The P atoms adsorbed have a catalytic
effect to welcome B groups and introduce them to the N sites,
consequently promote the modification of PNC because of borate
(Figure 4b). This discovery has potential significance for the
modification and surface reconstruction of carbon materials.
Infrared spectroscopy (IR) was also applied to reveal the
differences between samples, as seen in Figure 4a. The IR
spectra of PNC and PNC-E are extremely similar, indicating that
ethanol washing does not cause the changes of the surface
organic groups. While, the IR spectra of the samples (PNC-B and
PNC-E/B) treated with NaBH4 exhibit the significant changes. For
the PNC-B sample obtained by PNC treated with NaBH4 solution,
its IR spectrum is red curve in Figure 4a. The peak at 887 cm-1
corresponds to the out-of-plane bending vibration of BOH. The
peaks at 1053 cm-1 and 1090 cm-1 are corresponding to the in-
plane bending vibration of BOH.[47,48] The 1389 cm-1 peak
The high-resolution XPS spectra of B 1s (Figure 3b) were fitted
by two peaks. The peaks at 190.5 eV and 192.0 eV are
corresponding to the B-N and B-O bond, respectively.[10,25,43,44] It
can be seen that the peak of B-N bond is dominated no matter in
PNC-B or PNC-E/B. This indicates that the generated borides
after NaBH4 treatment are linked to the N atoms at the edges of
carbon matrix. The generation of the B-N bonds would weaken
the P-N bonds, thereby strengthening the P-C bond of the N-P-C
chemical environments. It also has explained the mechanism of
the P 2p spectra shifts to the lower binding energy, for PNC-B and
PNC-E/B.
3-
corresponds to the asymmetric stretching vibration of BO3 of
B(OH)3.[47-49] The peaks at 1567 cm-1 and 1630 cm-1 are
corresponding to the bending vibration of H2O molecules, which
do not occur to tetrahydroxy borate B(OH)4. Therefore, it can be
confirmed that the modification is in the form of B(OH)3.
Considering that the B spectrum of XPS is mainly composed of B-
N bonds and B-O bonds, as well as the previous discussion on
the XPS of pyridinic-N, the IR results further illustrate the structure
of the generating site is mainly pyridinic-N-B(OH)3.
The high-resolution XPS spectra of N 1s (Figure 3c) were
deconvolved into four peaks. The peaks at 398.6 eV, 399.8 eV,
401.2 eV, and 403.5 eV are attributed to pyridinic-N, pyrrolic-N,
graphitic-N and oxidized-N, respectively.[33,40,45,46] Comparing the
N 1s spectra of PNC and PNC-E, no significant difference is found
In addition, the peak covering a range of 3017-3660 cm-1 is
significantly broader for PNC-B (Figure 4a) than the other
samples. Compared to PNC and PNC-E, this peak of PNC-B has
expanded toward a much lower wave numbers, which is relate to
the weakly bonded H2O molecules. This situation reveals the
Figure 4. (a) IR spectrums of various sample and (b) Schematic diagram of structural changes.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
weak hydrogen bonds have formed between H2O molecules and
pyridinic-N-B(OH)3 and implies that some H2O molecules could
attach to such sites in a semi-free state. PNC-E/B is obtained by
treating with NaBH4 solution, beforehand removing the surface P
atoms chemisorbed. The peaks related to BOH and H2O in PNC-
B are obviously sharper, compared to PNC-E/B. This proves that
the presence of phosphorus chemisorbed can catalyze the
formation of pyridinic-N-B(OH)3 and it is in good agreement with
the XPS results. It is manifested that the surface P atoms can
welcome B species and introduce it to the pyridinic-N sites, thus
overcoming the steric hindrance (Figure 4b).
In summary, P, N dual-doped porous carbon was prepared. It
was found that a part of P was actually chemisorbed at N sites,
but not embedded in the carbon matrix. These P atoms were
demonstrated that they can catalyze the surface modification of
carbon material, resulting the novel sites of pyridinic-N-B(OH)3.
Interestingly, this surface modification can attach H2O molecules
in a semi-free state and close to the catalytic site of the ORR. This
is benefit to promote the offensive steps by H2O molecules in the
ORR process. In the reaction equation, the step (2) and step (4)
of ORR involve the attack of H2O molecules. Ingenious interface
modifications can increase the speed of these steps, which may
increase the apparent catalytic activity of ORR.
the electrolyte was O2-saturated, the CV curve of PNC-B
exhibited an obvious O2 reduction peak appeared at about 0.76
V, indicating that the PNC-B had excellent ORR activity. The
linear sweep voltammetry (LSV) curves (Figure 5b) were
performed to accurately investigate the ORR performance of
various samples. The catalytic performance of PNC is the worst
due to the negative onset potential (0.94 V vs. RHE) and half-
wave potential (0.75 V), with the smallest limit current density (5.0
mA cm-2). After washing with ethanol, the onset potential (0.95 V)
of the obtained PNC-E, especially the half-wave potential (0.78 V),
was positively shifted with a slightly increased limit current density
(5.3 mA cm-2). After further treatment with NaBH4, the limiting
current density of the obtained PNC-E/B was increased to 5.5 mA
cm-2, compared to the untreated PNC-E. Interestingly, the PNC-B
obtained by directly treating PNC with NaBH4 has the largest
increased limiting current density (6.1 mA cm-2), and the onset
potential (0.95 V) and half-wave potential (0.76 V) are also
significantly positive. Therefore, PNC-B has approached the
onset potential and half-wave potential of commercial Pt/C (20%)
catalyst (0.97 V and 0.83 V), especially the limit current density
has greatly exceeded that of commercial Pt/C (20%) catalyst (5.3
mA cm-2). This confirms that the surface modification significantly
improves the catalytic performance of ORR, especially the
catalytic current density is greatly improved at the high
overpotential. That can be more clearly expressed on the Tafel
slopes (Figure 5c), which reflects the reaction kinetics in the
electrocatalytic process. The Tafel slopes of PNC, PNC-E, PNC-
E/B and PNC-B are 103.9 mV dec-1, 87.3 mV dec-1, 86.1 mV dec-
1 and 64.3 mV dec-1, respectively. The Tafel slope of PNC-B is the
lowest among all these samples, and is close to the Tafel slope
of 20% Pt/C (63.7 mV dec-1). This result is related to a high
exchange current density at the electrode interfaces (in Table S2),
implying a faster kinetic process of PNC-B.
O2 + * → O2*
(1)
O2* + H2O + e – → HOO* + OH – (2)
HOO* + e – → O* + OH –
O* + H2O + e – → HO* + OH –
HO* + e – → OH – + *
(3)
(4)
(5)
The test results of the catalytic performances are consistent
with the above analysis, as discussed below. The electrocatalytic
activity was initially detected by cyclic voltammetry (CV) at a scan
rate of 10 mV s-1 in the potential range of 0.177-1.177 V (vs. RHE).
As shown in Figure 5a, there are not any redox peaks when
samples were tested in Ar-saturated 0.1 M KOH solution. After
Figure 5. (a) CV curves of PNC-B and Pt/C catalyst in Ar and O2-saturated 0.1 M KOH solution at a scan rate of 10 mV s-1; LSV curves (b) and Tafel slopes (c)
of various samples in O2-saturated 0.1 M KOH solution at 1600 rpm; (d) LSV curves at different rotation speeds and (e) the corresponding Koutecky-Levich plots of
PNC-B; (f) The i-t curves of PNC-B and Pt/C at 0.68 V (vs. RHE); and the inset shows durability test for the PNC-B by cyclic voltammograms scanning at 3000
cycles at 1600 rpm in O2-saturated 0.1 M KOH solution.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
To further investigate the catalytic mechanism of PNC-B, the
polarization curve measurements were performed on RDE at the
rotation speeds from 400 to 2025 rpm (Figure 5d). The
corresponding Koutechy-Levich (K-L) plots at 0.40-0.55 V (vs.
RHE) are exhibited in Figure 5e. The electron transfer number (n)
calculated from K-L plots is 3.99 3.79, 3.77 and 3.63 at the
corresponding potential of 0.40, 0.45, 0.50 and 0.55 V,
respectively. This indicates that the PNC-B catalyzes ORR
through a quasi-four electron transfer pathway. Meanwhile,
Rotating Ring Disk Electrode (RRDE) tests (Figure S6) were also
performed to calculate the H2O2 percent yields. As shown in
The performances of this surface-modified catalyst PNC-B in
energy devices had been further investigated, including its
application in Zn-air batteries and Al-air batteries. This is
important for evaluating the practicality of this surface
modification strategy.
The Zn-air battery consists of a catalyst-supported carbon cloth
(1.0 mg cm-2) as an air electrode, a piece of zinc foil as the anode
and 6 M KOH solution as electrolyte. As shown in Figure 6a, the
PNC-B based Zn-air battery remains at the open circuit voltage
(1.33 V) for more than 600 seconds, indicating a lower self-
discharging property. The discharge polarization curve and power
density curve of Zn-air batteries are shown in Figure 6b. The
maximum power density of the PNC-B based battery achieves
91.3 mW cm-2 at 0.70 V, which is higher than that of the Pt/C
based battery (85.2 mW cm-2). The PNC-B based battery
encountered long-term constant current discharge at 10 mA cm-2
for more than 14 hours without significant voltage drop observed
(the inset of Figure 6c), which indicates that the discharge process
has excellent stability. Meanwhile, the specific capacity of PNC-B
based battery achieves 767 mAh g−1 (in Figure 6c), which is
obtained at the discharge current density of 10 mA cm−2 and
normalized to the mass of zinc consumed. It is significantly higher
than that of Pt/C based battery (717 mAh g−1). PNC-B based Zn-
air batteries were also tested for specific capacities at different
current densities from 5 to 100 mA cm-2 (Figure S8). With the
increase of current density, the specific capacity of zinc-air battery
decreases continuously. More important, the Zn-air battery
assembled with PNC-B electrode also shows an excellent rate
capability with a quick dynamic response at current densities
ranging from 5 to 100 mA cm-2, as shown in Figure 6d. The Zn-air
battery equipped with PNC-B still maintains a high voltage with
almost no voltage decay by using the galvanostatic discharge
tests and the cell voltage of Zn-air battery with PNC-B electrode
can be restored to its initial level (Figure 6d). In Figure 6e, the Zn-
air battery packs of PNC-B light up the LED screen with red
Chinese characters “Qingdao University”.
-
Figure S7, PNC-B exhibits a HO2 yield less than 20% and the
calculated electron transfer number is 3.5-3.9, within the potential
range of 0.3-0.8 V (vs. RHE). This is consistent with the results of
RDE tests.
The catalytic stability of PNC-B was further investigated. As
seen in the inset of Figure 5f, PNC-B was tested by circulating
between -0.8 V and 0.2 V at a scanning rate of 50 mV s−1 in O2-
saturated 0.1 M KOH solution. The loss of limiting current density
was only 3% after 3000 cycles, and the onset potential did not
change significantly. As shown in Figure 5f, PNC-B was tested by
chronoamperometric curve (i-t) under the constant applied
potential (0.68 V vs. RHE). After 9 hours of testing, the current
density of PNC-B remained above 95% and only decreased about
5%. However, the current density decreased more than 15% for
the Pt/C catalyst. These results indicate that PNC-B has good
electrocatalytic stability towards ORR.
Based on the above analysis, the surface modification by
NaBH4 treatment has effectively improved the catalytic
performance of ORR. The surface modification not only preserves
the active sites, but also promotes the offensive steps of H2O
molecules, thereby improving the kinetic process, making the
ORR reaction take place in a quasi-four electron pathway and
exhibiting good stability.
The application performances in the devices
Figure 6. (a) Open circuit plots of Zn-air batteries; (b) The discharge polarization curves and power density curves of Zn-air batteries; (c) The galvanostatic
discharge curves of Zn-air batteries at 10 mA cm-2 (inset) and corresponding specific capacity is normalized to the mass of the consumed Zn; (d) Rate capability
behavior of the Zn-air battery assembled with PNC-B electrode; (e) LED screen powered by Zn–air batteries.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
The Al-air battery consists of a catalyst-supported carbon cloth Conclusion
(1.0 mg cm-2) as an air electrode, an aluminum foil as the anode,
and a solution containing 6 M KOH, 0.01 M NaSnO3, 0.5 mM
In(OH)3 and 7.5 mM ZnO as the electrolyte. As shown in Figure
7a, the PNC-B based Al-air battery remains at the open circuit
voltage (1.40 V) for more than 600 seconds, which indicates PNC-
B has a lower self-discharging property. Figure 7b shows the
discharge polarization curve and power density curve of the
assembled Al-air battery. The maximum power density of PNC-B
based Al-air battery is 75.4 mW cm-2, which is larger than the
maximum power density of Pt/C-based battery (67.1 mW cm-2). In
the meantime, there is no significant voltage drop for the Al-air
battery of PNC-B (the inset of Figure 7c) after a constant
discharge for more than 240 minutes under the current density of
10 mA cm-2. Under the constant discharge current density of 10
mA cm-2, the specific capacity of PNC-B based battery is 957 mAh
g−1 (Figure 7c), which normalizes to the mass of aluminum foil
consumed. Its specific capacity is significantly higher than that of
Pt/C based aluminum-air battery (718 mAh g−1). Specific
capacities of PNC-B based aluminum-air batteries at different
current densities of 5 to 100 mA cm-2 were also tested (Figure S9).
The specific capacity of aluminum-air battery decreases with the
increase of current density. In addition, the PNC-B based Al-air
battery also shows excellent rate capability and fast dynamic
response at current densities from 5 to 100 mA cm-2 (Figure 7d).
Through constant current discharge test, the Al-air battery
configured with PNC-B has almost no voltage attenuation while
maintaining high voltage, which can restore the battery voltage of
to the initial level. In Figure 7e, the Al-air battery packs of PNC-B
light up the LED screen with blue Chinese characters “Qingdao
University”.
The metal-free P, N dual-doped porous carbon was prepared
via the foaming-carbonization processes, achieving an ultra-high
specific surface area of about 2000 m2 g-1. Subsequently, several
simple surface post-treatment methods were compared. Either
ethanol washing (subtraction to make surfaces cleaner) or NaBH4
treatment (addition to modify the surface) can improve the
catalytic activity of ORR. Some of the chemisorbed phosphorus is
not actually bonded into the carbon matrix and can be dissolved
in ethanol. These P atoms chemisorbed on the carbon surface
can be used to catalyze the surface modification via NaBH4
treatment to create a novel catalytic site of pyridinic-N-B(OH)3.
The order of ORR activity is PNC-B> PNC-E/B> PNC-E>PNC.
Interestingly, PNC-B has the best catalytic performance because
the generated novel sites can attach H2O molecules of a semi-
free state, which is benefiting to promote the offensive steps by
H2O molecules in the ORR process. PNC-B exhibited the largest
limiting current density (6.1 mA cm-2), and the onset potential
(0.95 V) and half-wave potential (0.76 V) are significantly positive.
Furthermore, PNC-B was demonstrated to be applicable to some
devices, such as Zn-air battery and Al-air battery, and had
exhibited outstanding performances.
This work illustrates that in the heteroatom-doped metal-free
catalysts, some heteroatoms may actually not be bonded into the
bulk phase. In addition, the appropriate post-treatments can
conveniently and effectively improve the ORR performance of
metal-free catalysts. These two aspects require the widespread
attentions in this field.
Figure 7. (a) Open circuit plots of Al-air batteries; (b) The discharge polarization curves and power density curves of Al-air batteries; (c) The galvanostatic
discharge curves of Al-air batteries at 10 mA cm-2 (inset) and corresponding specific capacity is normalized to the mass of the consumed Al; (d) Rate capability
behavior of the Al-air battery assembled with PNC-B electrode ; (e) LED screen powered by Al-air batteries.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
Experimental Section
a reversible hydrogen electrode (RHE), and the potential
conversion formula is as follows:
Preparations
ERHE = EAg/AgCl +0.0592V × pH+ E0
(1)
Ag/AgCl
PNC was prepared as follows: The polyphosphoric acid and
phosphoric acid were mixed with a mass ratio of 3:2. The cellulose
powder was added and completely dissolved to form a viscous
solution. After that, the carbon nitride (C3N4) powder was
dispersed in the above solution. The mixture was then foamed at
a constant temperature of 150 °C for 1 hour to get an initial foam.
The obtained initial foam was carbonized in a tube furnace to get
the PNC, at 1000 °C for 2 hours under the Ar/H2 atmosphere.
PNC-B was prepared by continuously stirring the PNC powder
in 0.25 M sodium borohydride solution for 2 hours, subsequently
centrifuged and dried at 50 °C .
For the RDE tests, the polarization curves were carried out at a
scan rate of 10 mV s-1 at different rotation rates ranging from 400
to 2025 rpm. The limited diffusion currents (JK) and the electron
transfer number (n) can be obtained according to the K-L plots
which based on the following Koutecky - Levich (K-L) equations:
1
퐽
1
1
1
1
=
+
=
+
퐽
퐾
(2)
(3)
(4)
1/2
퐽
퐿
퐽
퐾
퐵푤
B = 0.2nFC0 (D0)2/3v-1/6
ퟏ
Jk =
풏풌푭푪
ퟎ
PNC-E was prepared by stirring the PNC powder in ethanol for
5 hours, subsequently centrifuged and dried at 50°C .
PNC-E/B was prepared by continuously stirring the PNC-E in
0.25 M sodium borohydride solution for 2 hours, subsequently
centrifuged and dried at 50°C .
The constant 0.2 is adopted when the rotation rate is expressed
in rpm. For the Tafel plot, the kinetic current was calculated from
the mass-transport correction of RDE by:
푱 ×푱
푳
JK =
(5)
푱 −푱
푳
Characterizations
where J represents the measured current density; JK and JL are
the kinetic- and diffusion-limiting current densities, respectively; w
is the linear rotation speed (rpm s-1) of the disk; F is the Faraday
constant (F=96485 C mol-1), C0 is the bulk concentration of O2
(C0=1.2×10-3 mol L-1), D0 is the diffusion coefficient of O2 in 0.1 M
KOH solution (D0=1.9×10-5 cm2 s-1), v is the kinematic viscosity of
the electrolyte (v=0.01 cm2 s-1), and k is the electron-transfer rate
constant. For the RRDE tests, the polarization curves were
measured at a disk rotation rate of 1600 rpm. The potential of the
ring was set at 1.4 V (vs. RHE). The collecting efficiency of the
RRDE (N) was 0.43. The H2O2 yield and the electron transfer
number (n) were calculated as follows:
The apparent morphologies were tested on scanning electron
microscope (SEM, SU8200, Japan). Transmission electron
microscope (TEM) and high-resolution transmission electron
microscopy (HRTEM) images were obtained in a JEOL2100F
instrument at an accelerating voltage of 200 kV. The X-ray
diffraction (XRD) was tested using a Bruker D8 Advance X-ray
diffractometer with Cu Kα radiation (λ = 0.154 nm) in the 2θ range
of 10−90° with scan rate of 20°/min. X-ray photoelectron
spectroscopic (XPS) analysis was carried out on a Thermo
Fisher Scientific ESCALAB 250Xi, and XPSPEAK software was
used for data analysis. Raman spectroscopy (514 nm radiation
laser) was employed using an inVia reflex. For BET surface area
analysis was obtained from an instrument ASAP 2020, and the
samples were degassed in vacuum and then measured at 77 K.
The infrared spectra were obtained by a Fourier transform
microscopy infrared spectrometer (Thermo Nicolet iN10). The 1
mg catalyst sample and 150 mg KBr were ground into a fine
powder and then pressed into a transparent sheet under 10 tons
of pressure. The prepared sheets were put into the spectrometer
to analyze the infrared spectrum.
푰풓/푵
H2O2% = 200×
(6)
(7)
푰풅+푰풓/푵
푰풅
n = 4× 푰 +푰 /푵
풅
풓
Where Id is the disk current and Ir is the ring current. The stability
test was performed at a static potential at room temperature with
the working electrode rotating at 400 rpm in O2-saturated 0.1 M
KOH solution.
Assembly and tests of batteries
Electrochemical measurements
Electrochemical measurements were conducted on an
electrochemical workstation CHI760E in a standard three-
electrode system under the room temperature. The catalyst film-
coated rotating disk electrode (RDE) or ring-disk electrode
(RRDE) served as the working electrode, with a glassy carbon
disk (4.0 mm diameter) and a Pt ring (5.0 mm inner-diameter and
7.0 mm out-diameter). Ag/AgCl (3 mol L-1 KCl) served as the
reference electrode and platinum-wire electrode served as the
counter electrode. 3 mg of the catalyst was dispersed in 1 mL
ultrapure water containing 50 μL of Nafion (5 wt%), and sonicated
for 30 minutes to form a uniform black ink. 12 μL sample ink was
dropped onto the electrode to form the film-coated test electrode.
The catalyst loading on the electrode surface was 0.288 mg cm−2.
The electrochemical tests were in the 0.1 M KOH solution, with
the scan rate was 10 mV s-1. Cyclic voltammetry (CV)
measurement was tested in the potential range from −0.8 V to
+0.2 V (vs. Ag/AgCl) and the RRDE tests were in the range from
−0.8 V to +0.1 V (vs. Ag/AgCl). The potentials were calibrated with
Zinc-air battery was assembled in a mold. The carbon cloth
loaded with catalyst about 1.0 mg cm-2 was used as the cathode.
A piece of polished zinc foil was used as the anode. The solution
of 6 M KOH was used as electrolyte. Aluminum-air battery was
assembled in a similar mold. Carbon cloth loaded with catalyst
about 1.0 mg cm-2 was used as the` cathode. A piece of polished
aluminum foil was used as the anode. The mixed solution was
used as an electrolyte containing 6 M KOH, 0.01 M Na2SnO3, 0.5
mM In(OH)3, 7.5 mM ZnO.[37-39]The discharge polarization curve
was tested by LSV method on a CHI760E electrochemical
workstation. The galvanostatic discharge 10 mA cm-2 was
performed in LAND testing system. The specific capacity (Csp) of
batteries was calculated from as follows:
Csp = (i×t)/(△m)
(8)
Where i is discharge current, t is discharge time and △m is the
consumed weight of zinc or aluminum foil.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
[22] S. B. Ren, X. L. Chen, P. X. Li, D. Y. Hu, H. L. Liu, W. Chen, W. B. Xie, Y.
X. Chen, X. L. Yang, D. M. Han, G. H. Ning, X. H. Xia, J. Power Sources
2020, 461, 228145.
Acknowledgements
[23] Y. Nie, Z. D. Wei, ChemCatChem 2019, 11, 5885-5897
[24] S. L. Zhao, X. Y. Lu, L. Z. Wang, J. Gale, R. Amal, Adv. Mater. 2019, 31,
1805367.
This work was supported by Shandong Provincial Key research
and development plan, China (Grant No. 2017GGX40119),
Shandong Provincial Natural Science Foundation, China (Grant
No. ZR2019MB033, ZR2015BM002), the Program of the Qingdao
Key Lab of solar energy utilization and energy storage technology
(Grant No. QDKLSE201602).
[25] L. Gao, M. C. Zhu, Z. Y. Zhang, G. L. Cui, Electrochim. Acta 2019, 327,
134980.
[26] M. Mooste, E. Kibena‐Põldsepp, V. Vassiljeva, A. Kikas, M. Käärik, J.
Kozlova, V. Kisand, M. Külaviir, S. Cavaliere, J. Leis, A. Krumme, V.
Sammelselg, S. Holdcroft, K. Tammeveski, ChemCatChem 2020，12，1-
15.
[27] Y. Gao, Z. C. Xiao, D. B. Kong, R. Iqbal, Q. H. Yang, L. J. Zhi, Nano Energy
2019, 64, 103879.
Conflict of interest
[28] X. H. Wang, X. B. Gong, L. Peng, Z. Yang, Y. Liu, Bioelectrochemistry 2019,
127, 104-112.
The authors declare no conflict of interest.
[29] Y. J. Ji, H. L. Dong, C. Liu, Y. Y. Li, J. Mater. Chem. A 2018, 6, 13489-
13508.
[30] J. Liu, P. Song, Z. G. Ning, W. L. Xu, Electrocatalysis 2015, 6, 132-147.
[31] Q. L. Wei, M. Cherif, G. X. Zhang, A. Almesrati, J. T. Chen, M. J. Wu, N.
Komba, Y. F. Hu, T. Regier, T.-K. Sham, F. Vidal, S. Sun, Nano Energy
2019, 62, 700-708.
Keywords: Aluminum-air battery • Metal-free electrocatalyst •
Oxygen reduction reaction • Surface treatment • Zinc-air battery
[32] H. J. Yu, L. Shang, T. Bian, R. Shi, G.I. N. Waterhouse, Y. F. Zhao, C. Zhou,
L. Z. Wu, C.H. Tung, T. R. Zhang, Adv. Mater. 2016, 28, 5080-5086.
[33] L. P. Ge, D. Wang, P. X. Yang, H. Xu, L. H. Xiao, G. X. Zhang, X. Y. Lu, Z.
Z. Duan, F. Meng, J. Q. Zhang, M. Z. An, Nanoscale 2019, 11, 17010-
17017.
[1] Y. D. Yan, S. C. Yan, Z. T. Yu, Z. G. Zou, ChemCatChem 2019, 11, 1033-
1038.
[2] Z. Dou, N. Du, X. Liu, Q. H. Wu, Q. Wu, Y. Fu, G. L. Zhang, T. Y. Ma,
ChemSusChem 2020, 13, 4094-4102.
[34] J. Quílez-Bermejo, M. Melle-Franco, E. San-Fabián, E. Morallón, D.
Cazorla-Amorós, J. Mater. Chemistry A 2019, 7, 24239-24250.
[35] J. Y. Sun, X. X. Li, J. L. Yang, ChemCatChem 2019, 6252-6257.
[36] Q. D. Wang, Y. M. Li, K. Wang, J. T. Zhou, L. W. Zhu, L. Gu, J. Hu, X. Cao,
Electrochim. Acta 2017, 257, 250-258.
[3] X. M. Ge, A. Sumboja, D. Wuu, T. An, B. Li, F.W.T. Goh, T.S.A. Hor, Y. Zong,
Z. L. Liu, ACS Catal. 2015, 5, 4643-4667.
[4] Y. G. Li, H. J. Dai, Chem. Soc. Rev. 2014, 43, 5257-5275.
[5] Q. Zhao, X. J. Huang, M. M. Zhou, Z. N. Ju, X. D. Sun, Y. Sun, Z. H. Huang,
H. Li, T. Y. Ma, ACS Appl. Mater. Interfaces 2020, 12, 36072-36081.
[6] X. H. Guo, Y. P. Zhu, T.Y. Ma, J. Energy Chem. 2017, 26, 1107–1116.
[7] L. Jiao, Y. L. Hu, H. X. Ju, C. D. Wang, M. R. Gao, Q. Yang, J. F. Zhu, S. H.
Yu, H. L. J. Mater. Chem. A 2017, 5, 23170-23178.
[37] T. He, Y. Zhang, Y. Chen, Z. Zhang, H. Wang, Y. Hu, M. Liu, C.-W. Pao,
J.-L. Chen, L.Y. Chang, Z. Sun, J. Xiang, Y. Zhang, S. Chen, J. Mater.
Chem. A 2019, 7, 20840-20846.
[38] Y. F. Wang, W. D. Pan, H.Y.H. Kwok, H. M. Zhang, X. Lu, D.Y.C. Leung, J.
Power Sources 2019, 437, 226896.
[8] L. J. Yang, Y. Zhao, S. Chen, Q. Wu, X. Z. Wang, Z. Hu, Chin. J. Catal. 2013,
34, 1986-1991.
[39] Y. Q. Wang, J. Hao, J. Yu, H. Yu, K. K. Wang, X. T. Yang, J. Li, W. Z. Li, J.
Energy Chem. 2020, 45, 119-125.
[9] Y. L. Zhao, X. G. Li, X. B. Jia, S. Y. Gao, Nano Energy 2019, 58, 384-391.
[10] X. X. Liu, Y. H. Wang, L. B. Chen, P. P. Chen, S. P. Jia, Y. Zhang, S. Y.
Zhou, J. B. Zang, ACS Appl. Mater. Interfaces 2018, 10, 37067-37078.
[11] R. P. Zhao, Q. H. Li, Z. J. Chen, V. Jose, X. Jiang, G. T. Fu, J. M. Lee, S.
M. Huang, Carbon 2020, 164, 398-406.
[40] X. L. Xing, R. J. Liu, M. Anjass, K. C. Cao, U. Kaiser, G. J. Zhang, C. Streb,
Appl. Catal., B 2020, 277, 119195.
[41] H. Zhao, Z. P. Hu, Y. P. Zhu, L. Ge, Z. Y. Yuan, Chin. J. Catal. 2019, 40,
1366-1374.
[12] Q. Li, T. He, Y. Q. Zhang, H. Q. Wu, J. J. Liu, Y. J. Qi, Y. P. Lei, H. Chen,
Z. F. Sun, C. Peng, L. Z. Yi, Y. Zhang, ACS Sustainable Chem. Eng. 2019,
7, 17039-17046.
[42] Y. H. Wang, S. J. Wang, R. Q. Li, H. B. Li, Z. J. Guo, B. L. Chen, R. Li, Q.
X. Yao, X. X. Zhang, H. Y. Chen, Y. Li, K. G. Qu, Y. Zheng, Carbon 2020,
162, 586-594.
[13] C.J. Mena-Durán, I.L. Alonso-Lemus, P. Quintana, R. Barbosa, L.C.
Ordoñez, B. Escobar, Int. J. Hydrogen Energy 2018, 43, 3172-3179.
[14] Y. F. Cheng, F. Liao, W. Shen, L. B. Liu, B. B. Jiang, Y. Q. Li, M. W. Shao,
Nanoscale 2017, 9, 18977-18982.
[43] P. Marbaniang, I. Patil, M. Lokanathan, H. Parse, D. C. Sesu, S. Ingavale,
B. Kakade, ACS Sustainable Chem. Eng. 2018, 6, 11115-11122.
[44] S. Y. Mou, T. W. Wu, J. F. Xie, Y. Zhang, L. Ji, H. Huang, T. Wang, Y. L.
Luo, X. L. Xiong, B. Tang, X. P. Sun, Adv. Mater. 2019, 31, 1903499.
[45] Q. Lv, N. Wang, W. Y. Si, Z. F. Hou, X. D. Li, X. Wang, F. H. Zhao, Z. Yang,
Y. L. Zhang, C. S. Huang, Appl. Catal., B 2020, 261, 118234.
[46] X. Y. Xue, H. Yang, T. Yang, P. F. Yuan, Q. Li, S. C. Mu, X. L. Zheng, L. F.
Chi, J. Zhu, Y. G. Li, J. N. Zhang, Q. Xu, J. Mater. Chem. A 2019, 7, 15271-
15277.
[15] R. J. Liu, K. C. Cao, A. H. Clark, P. L. Lu, M. Anjass, J. Biskupek, U. Kaiser,
G. J. Zhang, C. Streb, Chem. Sci. 2020, 11, 1043-1051.
[16] Y. Fu, P. Richardson, K. K. Li, H. Yu, B. Yu, S. Donne, E. Kisi, T. Y. Ma,
Nano-Micro Lett. 2020, 12:65.
[17] S. Dilpazir, R. J. Liu, M. L. Yuan, M. Imran, Z. J. Liu, Y. B. Xie, H. Zhao, G.
J. Zhang, J. Mater. Chem. A 2020, 8, 10865-10874.
[47] G. Kessler, E. Ronsch, H. A. Lehmann, Z. Anorg. Allg. Chem. 1969, 365,
281-287.
[18] D. F. Yan, Y. X. Li, J. Huo, R. Chen, L. M. Dai, S. Y. Wang, Adv. Mater.
2017, 29, 1606459.
[48] J. Gotjreau, H. Kallfass, Z. Anorg. Allg. Chem. 1959, 299, 160-169.
[49] S. Ghose, C. Wan, Am. Mineral. 1978, 63, 160-171.
[19] L. Hao, S. S. Zhang, R. J. Liu, J. Ning, G. J. Zhang, L. J. Zhi, Adv. Mater.
2015, 27, 3190-3195.
[20] X. Xiao, X. H. Li, Z. X. Wang, G. C. Yan, H. J. Guo, Q. Y. Hu, L. J. Li, Y.
Liu, J. X. Wang, Appl. Catal., B 2020, 265, 118603.
[21] H. Singh, S. Q. Zhuang, B. Ingis, B.B. Nunna, E.S. Lee, Carbon 2019, 151,
160-174.
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001319
ChemCatChem
FULL PAPER
Entry for the Table of Contents
Study on the Simple Surface Treatments of N, P Dual-doped
Carbon as Metal-free Catalyst for Metal-air Batteries
Heping Yu, Hui Zhang, Zhongyi Zhang*
Novel active sites generated by simple surface treatment can further improve the activity of oxygen reduction reaction by promoting
the attack from semi-free water molecules, which is applied to the preparation of metal-free catalysts in metal-air batteries.
This article is protected by copyright. All rights reserved.