Low temperature chemical reduction of fusional sodium metasilicate nonahydrate into a honeycomb porous silicon nanostructure

Article abstract of DOI:10.1039/c4cc00888j

Honeycomb porous silicon (hp-Si) has been synthesized by a low temperature (200 °C) magnesiothermic reduction of Na₂SiO₃· 9H₂O. This process can be regarded as a general synthesis method for other silicide materials. Significantly, hp-Si features excellent electrochemical properties after graphene coating. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cc00888j

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Low temperature chemical reduction of fusional
sodium metasilicate nonahydrate into a
honeycomb porous silicon nanostructure†
Cite this: Chem. Commun., 2014,
50, 6856
Received 3rd February 2014,
Accepted 17th March 2014
Jianwen Liang,a Denghu Wei,a Ning Lin,a Youngchun Zhu,*a Xiaona Li,a
Jingjing Zhang,a Long Fana and Yitai Qian*ab
DOI: 10.1039/c4cc00888j
www.rsc.org/chemcomm
Honeycomb porous silicon (hp-Si) has been synthesized by a low the reaction described as: 2C(s) + SiO2(s) = 2CO(g) + Si(l) and
temperature (200 8C) magnesiothermic reduction of Na2SiO3Á9H2O. requires the use of a temperature well above silicon’s melting
This process can be regarded as a general synthesis method point (>2000 1C).12 Recently, Sandhage et al. demonstrated that
for other silicide materials. Significantly, hp-Si features excellent silicon oxide could be reduced to Si with gaseous Mg via a
electrochemical properties after graphene coating.
so-called magnesiothermic reduction reaction (MRR).8 Using
this method, three-dimensional silicon oxide micro-assemblies
Over the past few years, Si nanostructured materials have were formed into porous Si by the following reaction: 2Mg(g) +
received extensive interest due to their importance in the field SiO2(s) = 2MgO(s) + Si(s) at a temperature higher than 650 1C.
of functional electronic devices such as photovoltaic devices,1,2 Since then, various high purity Si or Si-based nanostructured
electronic devices3 and lithium ion batteries (LIBs).4 Meanwhile, materials have been synthesized by the same MRR process.13–16
the production of pure Si nanostructured materials has been the Nonetheless, these methods are usually of high temperature
subject of many investigations. Typically preparation methods (>650 1C) and not easy to scale up.
such as chemical physical deposition,5 laser ablation,6 chemical
It is well known that solid-state chemical reaction processes
or electrochemical reduction,7–9 thermal reduction10 and others2,11 with coupled reaction–diffusion phenomena at solid/gas or
have gradually appeared in recent years. However, these growth solid/liquid interfaces, lead to a higher reaction rate than
mechanisms have some limitations, as they generally need high solid/solid interfaces only.17,18 The solid/gas or solid/liquid
temperature or a high vacuum, templates and complex equipment, interface provides a larger surface area to boost the reaction
or they employ hazardous and expensive Si precursors. The desire or diffusion and it also avoids the formation of a kinetically
for low-cost Si-based electronic devices has encouraged the quest limited process in the reaction. Therefore, if a much lower
to exploit more effective approaches to reduce the synthesis costs melting point material is considered as a Si precursor, the
for industry as well as overcoming the present technological reaction temperature for producing a Si nanostructured material
difficulties. Therefore, developing a new effective method should therefore be reduced markedly.
for the facile synthesis of Si beyond the current systems is
extremely urgent.
Herein, we present a new synthetic approach for direct
synthesis of honeycomb porous nanostructured silicon (hp-Si)
Direct solid-state reduction of silicon oxide is one of the via a low temperature chemical reduction reaction using sodium
least expensive and easily controllable methods for the synthesis metasilicate nonahydrate (Na2SiO3Á9H2O) as the Si precursor and
of pure silicon materials.7 The carbothermic reduction of silicon Mg as the reductant. Due to the low melting point of Na2SiO3Á
oxide has been studied for several decades and it is still a 9H2O (42 1C), crystalline Si can be obtained by this approach
significant way to synthesise Si in industry, which is based on at a temperature as low as 200 1C, only approximately a tenth
of that applied in the earlier carbothermic reduction and third
of the magnesiothermic reduction process. Neither a high
temperature special furnace nor a high purity protection inter
gas is required. At the same time, this process enables the
syntheses of microporous honeycomb nanocrystalline silicon
microassemblies with mesoporous shapes. Moreover, this
chemical reaction process can be regarded as a general low
temperature synthesis method for different kinds of silicide
based material such as SiC and MgSiN2.
a Hefei National Laboratory for Physical Science at Micro-scale,
Department of Chemistry, University of Science and Technology of China, Hefei,
Anhui 230026, PR China. E-mail: ychzhu@ustc.edu.cn, ytqian@ustc.edu.cn
b School of Chemistry and Chemical Engineering, Shandong University, Jinan,
Shandong 250100, PR China. E-mail: ytqian@ustc.edu.cn
† Electronic supplementary information (ESI) available: Experimental section;
detailed synthesis and analysis of SiC and MgSiN2 materials; auxiliary analysis
such as different magnification SEM images, TEM images and Raman spectra.
See DOI: 10.1039/c4cc00888j
6856 | Chem. Commun., 2014, 50, 6856--6859
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
smooth surface morphology (Fig. S4, ESI†). The transmission elec-
tron microscopy (TEM) image (Fig. 1c) of the hp-Si product clearly
reveals the presence of mesopores in the honeycomb framework.
Nitrogen sorption measurements show that hp-Si is highly porous
with a Brunauer–Emmett–Teller (BET, Fig. S5, ESI† for the nitrogen
adsorption and desorption isotherms) surface area of 19.3 m2 gÀ1
and a distributional Barrett–Joyner–Halenda (BJH, Fig. 1f) mesopore
size of 2.5 nm. The high-resolution transmission electron micro-
scopy (HRTEM, Fig. 1d) image shows the lattice fringe images of
0.312 nm that can be indexed to the (111) diffraction peaks of cubic
Si, confirming the crystalline nature of hp-Si.
For a more specific understanding of this reaction, different
reduction temperature have also been analysed. With a temperature
higher than 350 1C, MgO and Mg2Si become the main products
due to magnesium hydroxide dehydration and magnesium silicon
alloying (Fig. S6, ESI†), which would lead to an uncontrollable
reaction process and a low yield of Si production. For example,
when the reaction temperature is up to 400 1C, the yield of Si is as
low as 1%. XRD patterns of the products synthesized at different
reaction temperatures clearly show that silicon can be synthesized
from a minimum of 200 to 350 1C. Next, the autogenic pressure
originating from hydrogen produced in this reaction process was
found to play a critical role in the formation of hp-Si. Fig. 2 shows
the autogenic pressure and temperature as a function of time during
reaction under in situ detection. When the temperature increases to
near to 200 1C, the autogenic pressure sharply increases from 4.7
(at 191.5 1C) to 15.0 MPa (at 199.7 1C) and eventually saturates
around 15.5 MPa (when temperature hold at 200 1C for 10 hours).
What’s more, the autogenic pressure reduces to about 13 MPa when
the temperature decreases to room temperature naturally. This
slight reduction is possibly due to the decrease of temperature
(DP = nRDT/V, the ideal gas equation). On the other hand, the
high pressure remained in the reaction system indicating that it
is spontaneously derived from the chemical reaction process.
It is significant that crystalline Si can be directly fabricated
Fig. 1 (a) SEM image of agglomerated as-prepared reacted products. (b)
SEM image, (c) TEM image, (d) HRTEM image, (e) XRD pattern and (f) BJH
pore diameter distribution of the hp-Si product.
Fusional Na2SiO3Á9H2O was converted into NaOH/Si as-prepared
products by the following reaction with Mg in a stainless steel
autoclave (Fig. S1, ESI†) at a temperature of 200 1C and an autogenic
pressure of 15 MPa:
Na2SiO3Á9H2O(l) + 5Mg(s) = 2NaOH(s) + 5Mg(OH)2(s) + 3H2(g) + Si(s).
(1)
As shown in Fig. 1a (also large scale SEM image in Fig. S2 in via a direct solid-phase precursor reduction reaction from
ESI†), the as-prepared reacted products are agglomerated into silicate at such a modest temperature (200 1C), at which the
dozens of micron-sized and well-dispersed wizened vesicles in the formation of high reactivity species (such as Mg2Si) and volatile
surface. Thus, suggesting that the synthesis process of silicon here silicon gas species (such as Mg(g), SiO(g), Si(g)) was suppressed.
may involve some kind of melting and gas escape process. The Although a few authors have previously examined the synthesis
X-ray diffraction (XRD) pattern (Fig. S3, ESI†) of the reacted
products shows several intense diffraction peaks. Most of these
can be indexed to Mg(OH)2 and MgO, which is the main compo-
nent of the reacted products. A trace amount of Na2MgSiO4 was
also detected from the XRD pattern, which was associated with
non-reduced silicate. MgO and Mg(OH)2 were both removed by 1 M
HCl solution, and Na2MgSiO4 was removed by 10% HF solution.
The XRD pattern after this treatment is shown in Fig. 1e. There is
one sharp reflection around 2y = 28.51 and four weak reflections at
47.41, 56.21, 69.31 and 76.61, corresponding to the diffraction peaks
of the (111), (220), (311), (400) and (331) lattice planes of cubic Si
(JPCDS 77-2111), respectively. This suggests that NaSiO3Á9H2O was
successfully converted into Si after reduction at 200 1C. The yield of
this Si material is expected to be higher than 10%.
The resulting Si product shows a porous honeycomb morphology
with a pore size of several hundred nanometers (hp-Si, Fig. 1b) and
Fig. 2 The curves of autogenic pressure (top) and temperature (bottom)
as a function of time during the reaction under the in situ detection.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 6856--6859 | 6857

View Article Online
ChemComm
Communication
of Si from silicon oxide with magnesium, the low temperature
synthesis for Si via magnesiothermic reduction by such a
reaction was not examined. Besides Si, there is numerous other
silicides that possess unusual properties and have attracted
considerable interest as nanostructures for fabrication with a
low temperature synthesis approach.15,19 With this in mind, we
also tested the synthesis of SiC and MgSiN2 materials following
a procedure similar to that for Si. Experimental details are
shown in the ESI.† XRD analysis (Fig. S7 and S8, ESI†) clearly
reveals that the crystalline phase of MgSiN2 and SiC can be
analogously synthesized by using a corresponding fusional
Na2SiO3Á9H2O as a precursor at low temperature (200–300 1C).
Si materials have been considered as the most promising
anode candidates for advanced LIBs due to their high theoretical
capacity and natural abundance.20 However, the practical applica-
tion of Si anodes has been greatly impeded due to their huge
volume fluctuation during cycling, resulting in pulverization, loss of
electrical continuity, and fast capacity fading.21–23 Thus, researchers
have concentrated on constructing hybrid anodes by integrating
micro/nano-structure with carbon matrices.24,25 Here, in order to
use the hp-Si anodes in LIBs, a graphene coating was employed.
We used the covalent bonding of an aniline polymer as the
adhesive to assemble the graphene layer on the surface of the
hp-Si (hp-Si@G).26 The resulting assembly process on GO and RGO
was checked by TEM. Fig. S9 (ESI†) shows a typical TEM image of
an aniline polymer after a diazotization reaction.26 The aniline
polymer was uniformly distributed on the surface of hp-Si, demon-
strating the well-behaved assembly process. Fig. 3a and b show the
SEM and TEM images of hp-Si@G (an additional SEM image
showing the large morphology characteristics of hp-Si@G is given
in Fig. S10, ESI†). As expected, the intact morphology of hp-Si and
graphene layer is due to the assembly process. The carbon content
Fig. 3 Electrochemical performance of a hp-Si@G electrode. (a) SEM and
(b) TEM image of hp-Si@G. (c) Typical voltage curves at 1st, 2nd, 3rd cycles
in the voltage window of 0.005–1.0 V versus Li/Li+ at 3.6 A gÀ1. (d) Cycling
stability and Coulombic efficiency of the hp-Si@G anode at 3.6 A gÀ1
(e) Electrode cycled at various current densities.
.
of hp-Si@G is about 17.3 wt% by elemental analysis measurement. from 0.36 A gÀ1 to 7.2 A gÀ1. As shown in Fig. 3e, the hp-Si@G
The Raman spectra of the hp-Si show a characteristic Si(TO) band anode exhibits a considerably good rate capability and still
at 521 cmÀ1 (Fig. S11a, ESI†). After the assembly process, it has delivers a reversible capacity of 0.4 A h gÀ1 at a high current
significant additional two broad peaks at 1345 cmÀ1 (D band, density of 7.2 A gÀ1. There is a difference in the capacities
disordered) and at 1580 cmÀ1 (G band, graphene) (Fig. S11b, ESI†). between Fig. 3d (B1.2 A h gÀ1) and Fig. 3e (B0.7 A h gÀ1) at the
The strong Raman signals of the D and G bands were thought to same current density of 3.6 A gÀ1 around the 30th cycle. This
stem from the randomly distributed graphene sheets.
may be due to the different current density set before 3.6 A gÀ1
.
The electrochemical properties of the hp-Si@G anode were Lower current density means longer cycling times during the
investigated in CR2016 coin cells with lithium foil as a counter test, more electrolyte decomposition and thicker SEI (solid
electrode. Fig. 3c shows typical voltage curves at 1st, 2nd, 3rd electrolyte interface) film formation will appear. The SEI film
cycles in the voltage window of 0.005–1.0 V versus Li/Li+ at increases ionic resistance and causes electrical isolation of
3.6 A gÀ1. The plateau located at around 0.2 V relates to the electrode materials, which can lead to capacity degradation.22
alloy formation process between Li and Si. The cycling stability This results in lower capacity performance when the current
and Coulombic efficiency of the hp-Si@G anode at 3.6 A gÀ1 are density goes to 3.6 A gÀ1. Similar phenomenon can also be
shown in Fig. 3d. This composite anode shows the cycling found in many previous reports.27–29 Notably, the capacity
stability with reversible capacity at B0.6 A h gÀ1 after 360 cycles reversibly recovers to approximately 1.5 A h gÀ1 once the current
(all of the specific capacity is based on the total mass of the rate goes back to 0.4 A gÀ1 and is stable at B1 A h gÀ1 after
Si–G composites). The hp-Si@G anode delivers a relatively high 160 cycles. It is a normal characteristic of the electrochemical
initial Coulombic efficiency of 76.0%, then increases to performance whether the capacity could be recovered after the
B93.7% and finally stabilizes at nearly 100% in the subsequent high rate measurements.23,24,27 In the high current density, due
cycles. The high value of the Coulombic efficiency is attributed to the limitation of electronic conduction and ionic conduction,
to the porous honeycomb structure of silicon and a thin the activity of electrode materials did not show completely. Thus,
graphene layer for protection. Furthermore, the rate capability when the current density decreases, the capacity can be recovered.
of hp-Si@G was investigated with increasing current densities The high capacity and stability of the hp-Si@G composite
6858 | Chem. Commun., 2014, 50, 6856--6859
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
H. W. Abernathy Iii, C. J. Summers, M. Liu and K. H. Sandhage,
Nature, 2007, 446, 172–175.
9 T. Nohira, K. Yasuda and Y. Ito, Nat. Mater., 2003, 2, 397–401.
10 J. P. Murray, G. Flamant and C. J. Roos, Sol. Energy, 2006, 80,
1349–1354.
11 Y. Zhong, F. Peng, X. Wei, Y. Zhou, J. Wang, X. Jiang, Y. Su, S. Su,
S.-T. Lee and Y. He, Angew. Chem., Int. Ed., 2012, 51, 8485–8489.
12 M. Nagamori, I. Malinsky and A. Claveau, Metall. Trans. B, 1986, 17,
503–514.
13 H. Jia, P. Gao, J. Yang, J. Wang, Y. Nuli and Z. Yang, Adv. Energy
Mater., 2011, 1, 1036–1039.
14 M. Dasog, L. F. Smith, T. K. Purkait and J. Veinot, Chem. Commun.,
2013, 49, 7004–7006.
15 Y. Shi, F. Zhang, Y.-S. Hu, X. Sun, Y. Zhang, H. I. Lee, L. Chen and
G. D. Stucky, J. Am. Chem. Soc., 2010, 132, 5552–5553.
16 J. K. Yoo, J. Kim, Y. S. Jung and K. Kang, Adv. Mater., 2012, 24,
5452–5456.
17 C. E. Birchenall, J. Electrochem. Soc., 1956, 103, 619–624.
18 Y. Yin, R. M. Rioux, C. K. Erdonmez, S. Hughes, G. A. Somorjai and
A. P. Alivisatos, Science, 2004, 304, 711–714.
19 R. G. Blair, A. Anderson and R. B. Kaner, Chem. Mater., 2005, 17,
2155–2161.
demonstrate the benefits of such an open, porous, 3D structure
with such synthetic materials.
In summary, porous honeycomb silicon can be synthesized
with a yield of 10–15% by the low temperature magnesiothermic
reduction of Na2SiO3Á9H2O at only 200 1C. Significantly, the
as-prepared hp-Si features excellent electrochemical storage
properties after graphene coating, including a high specific
reversible capacity (ca. 1 A h gÀ1) and a stable cycle retention
(several hundred cycles), as promising anode materials in LIBs.
Also, this low temperature magnesiothermic reduction process
can be regarded as a general method for other silicide materials.
This work was supported by the 973 Project of China (No.
2011CB935901), the National Natural Science Fund of China
(No. 91022033, 21201158), Anhui Provincial Natural Science
Foundation (1208085QE101) and the Fundamental Research
Funds for the Central Universities (No. WK 2340000027).
20 J. Szczech and S. Jin, Energy Environ. Sci., 2011, 4, 56–72.
21 M. McDowell, S. Lee, W. Nix and Y. Cui, Adv. Mater., 2013, 25,
4966–4985.
Notes and references
1 X. Li, Y. Xiao, J. H. Bang, D. Lausch, S. Meyer, P.-T. Miclea, J.-Y. Jung, 22 M. McDowell, I. Ryu, S. Lee, C. Wang, W. Nix and Y. Cui, Adv. Mater.,
S. L. Schweizer, J.-H. Lee and R. B. Wehrspohn, Adv. Mater., 2013, 25,
3187–3191.
2 S. K. Cho, F.-R. F. Fan and A. J. Bard, Angew. Chem., Int. Ed., 2012, 51,
12740–12744.
3 S. Su, X. Wei, Y. Guo, Y. Zhong, Y. Su, Q. Huang, C. Fan and Y. He,
Part. Part. Syst. Charact., 2013, 30, 326–331.
4 J. R. Szczech and S. Jin, Energy Environ. Sci., 2011, 4, 56–72.
2012, 24, 6034–6041.
23 H. Wu, G. Chan, J. Choi, I. Ryu, Y. Yao, M. McDowell and Y. Cui, Nat.
Nanotechnol., 2012, 7, 310–315.
24 J. Deng, H. Ji, C. Yan, J. Zhang, W. Si, S. Baunack, S. Oswald, Y. Mei
and O. Schmidt, Angew. Chem., 2013, 125, 2382–2386.
25 Z. Han, N. Yabuuchi, K. Shimomura, M. Murase, H. Yui and
S. Komaba, Energy Environ. Sci., 2012, 5, 9014–9020.
5 E. Mullane, T. Kennedy, H. Geaney, C. Dickinson and K. M. Ryan, 26 C. Du, M. Chen, L. Wang and G. Yin, J. Mater. Chem., 2011, 21,
Chem. Mater., 2013, 25, 1816–1822. 15692–15697.
6 F. Erogbogbo, K.-T. Yong, I. Roy, G. Xu, P. N. Prasad and 27 T. Bogart, D. Oka, X. Lu, G. Meng, C. Wang and B. Korgel, ACS Nano,
M. T. Swihart, ACS Nano, 2008, 2, 873–878. 2014, 8, 915–922.
7 J.-I. Lee, K. T. Lee, J. Cho, J. Kim, N.-S. Choi and S. Park, Angew. 28 Y. Xiao, D. Hao, H. Chen, Z. Gong and Y. Yang, ACS Appl. Mater.
Chem., Int. Ed., 2012, 51, 2767–2771. Interfaces, 2013, 5, 1681–1687.
8 Z. Bao, M. R. Weatherspoon, S. Shian, Y. Cai, P. D. Graham, 29 J. Chang, X. Huang, G. Zhou, S. Cui, P. B. Hallac, J. Jiang,
S. M. Allan, G. Ahmad, M. B. Dickerson, B. C. Church, Z. Kang,
P. T. Hurley and J. Chen, Adv. Mater., 2014, 26, 758–764.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 6856--6859 | 6859