Observation of reversible pore change in mesoporous tin phosphate anode material during Li alloying/dealloying

Article abstract of DOI:10.1149/1.2212061

We observed the pore expansion and contraction of mesoporous tin phosphate during Li alloying/dealloying using small-angle X-ray scattering and transmission electron microscopy. As-prepared mesoporous tin phosphate showed pore and porewall sizes of 3 and 2 nm, respectively. During lithium alloying (discharging), pore size was slightly contracted, but porewall size was slightly expanded within the range of 1 nm. However, during lithium dealloying (charging), pore and porewall sizes recovered to their original sizes before cycling. The charged sample had a nanoscale pore (～3 nm) array with a more or less uniformly sized open-porewall structure of amorphous lithium phosphates with metallic α-Sn face-centered-cubic nanocrystals ～2 nm in diameter.

Full text of DOI:10.1149/1.2212061

Journal of The Electrochemical Society, 153 ͑9͒ A1633-A1636 ͑2006͒
A1633
0013-4651/2006/153͑9͒/A1633/4/$20.00 © The Electrochemical Society
Observation of Reversible Pore Change in Mesoporous Tin
Phosphate Anode Material during Li Alloying/Dealloying
Hansu Kim,a Gyeong-Su Park,b Eunjin Kim,c Jinyoung Kim,c
,z
Seok-Gwang Doo,a and Jaephil Choc,
*
aMaterials Laboratory and bAnalytical Engineering Center, Samsung Advanced Institute of Technology,
Giheung, Korea
cDepartment of Applied Chemistry, Kumoh National Institute of Technology, Gumi, Korea
We observed the pore expansion and contraction of mesoporous tin phosphate during Li alloying/dealloying using small-angle
X-ray scattering and transmission electron microscopy. As-prepared mesoporous tin phosphate showed pore and porewall sizes of
3 and 2 nm, respectively. During lithium alloying ͑discharging͒, pore size was slightly contracted, but porewall size was slightly
expanded within the range of 1 nm. However, during lithium dealloying ͑charging͒, pore and porewall sizes recovered to their
original sizes before cycling. The charged sample had a nanoscale pore ͑ϳ3 nm͒ array with a more or less uniformly sized
open-porewall structure of amorphous lithium phosphates with metallic ␣-Sn face-centered-cubic nanocrystals ϳ2 nm in
diameter.
© 2006 The Electrochemical Society. ͓DOI: 10.1149/1.2212061͔ All rights reserved.
Manuscript submitted December 22, 2005; revised manuscript received March 22, 2006. Available electronically June 26, 2006.
acetone several times for microstructural analysis. The small-angle
X-ray scattering ͑SAXS͒ patterns were obtained using a Rigaku in-
strument. The nitrogen adsorption isotherm was measured at 77 K
on a Micromeritics ASAP 2010 gas sorption system.
Metal phosphates have been heavily studied in the area of AlPO4
for catalytic applications, but little progress has been made with
other types of phosphates.1-4 Recently, surfactant-mediated SnO2
and tin phosphates have been reported as having possible applica-
tion in Li rechargeable batteries.5-7 Mesoporous and crystalline com-
posites consisting of tin phosphates with a Brunauer–Emmett–Teller
͑BET͒ surface area of 76 m2/g exhibited initial charge capacity of
ϳ600 mAh/g with an irreversible capacity ratio of 25%.5 Further,
these materials showed Ͻ10% capacity fading after 30 cycles com-
pared to other lithium reactive metal oxides that underwent severe
capacity losses. It has been proposed that the excellent capacity
retention was possibly due to reversible pore changes during the Li
reaction. Kim et al. have reported that tin phosphate was first de-
composed to an active Li4.4Sn alloy and an inactive amorphous
lithium phosphate phase.5 A similar mechanism was also proposed
by Courtney and Dahn8,9 and Xiao et al.10 Despite these studies, no
direct evidence for pore and porewall changes or the formation of
tin particles in the porewall framework during Li alloying/
dealloying has been discovered.
Results and Discussion
Figure 1a shows the SAXS patterns of the as-prepared mesopo-
rous tin phosphate after annealing at 500°C for 5 h. The intense
peak of the as-synthesized mesoporous tin phosphate corresponds to
the d-spacing of 5.0 0.2 nm ͑from 2d sin ␪ = ␭͒. Similar single-
peak diffraction patterns with small d-spacing were previously ob-
served in SnO2 and tin phosphates.7,12,13 An as-prepared sample
shows a relatively sharp peak, indicating a high degree of pore or-
dering. However, SAXS patterns of the samples discharged and
charged to 0 and 1.5 V, respectively, show somewhat broader peaks,
indicating a disordered pore structure. Even in mesoporous materials
with very low degrees of pore ordering, SAXS showed these
peaks.4,7,14 The broad high-angle X-ray diffraction ͑XRD͒ pattern
͑Fig. 1 inset, curve d͒ indicates that the tin phosphate consists of
either amorphous or possibly nanocrystalline mesoporous walls.
Curves b and c in Fig. 1 exhibit the SAXS patterns of the mesopo-
rous tin phosphate after discharging and charging to 0 and 1.5 V,
Here, we report a systematic observation of pore structure
changes in mesoporous tin phosphate during Li alloying/dealloying.
Experimental
Mesoporous tin phosphate was prepared by mixing 9.0 g SnF2
and 12 g H3PO4 followed by dissolution in distilled-deionized water
͑DDW͒. Twelve grams of sodium dodecyl sulfate ͑SDS͒ was dis-
solved in 40 mL of DDW, and this solution was added to a solution
of SnF2 and H3PO4. The mixture was stirred at 40°C for 1 h and
then maintained in an autoclave at 90°C for 5 days. After cooling to
room temperature, the precipitate was recovered by filtration,
washed with distilled water and ethanol, and vacuum-dried at 100°C
for 10 h. The powders were annealed at 500°C for 5 h, yielding a
mesoporous tin phosphate phase. The procedure for assembling a
coin-type half-cell with Li as the anode is described in detail
elsewhere.11,12 Generally, the area of the electrode was 2 cm2 and
contained 20 mg tin phosphate powder and 2 mg polyvinylidene
fluoride. A mixture of ethylene carbonate/diethylene carbonate ͑EC/
DEC͒ and 1 M LiPF6 salts was used as the electrolyte. After assem-
bling the cells, they were discharged and charged to 0 and 1.5 V,
respectively. The electrodes were detached from the cells, rinsed
several times with a dimethyl carbonate ͑DMC͒ solution, and then
the binder was removed in an N-methyl-2-pyrrolidone ͑NMP͒ solu-
tion. The solution was filtered and the sample was washed with
Figure 1. SAXS diffraction patterns of ͑a͒ as-prepared, ͑b͒ discharged, and
͑c͒ charged tin phosphates to 0 and 1.5 V, respectively. Insets of ͑d͒, ͑e͒ and
͑f͒ are wide-angle XRD patterns of ͑a͒, ͑b͒, and ͑c͒, respectively.
*
Electrochemical Society Active Member.
z E-mail: jpcho@kumoh.ac.kr
Downloaded on 2015-03-07 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1634
Journal of The Electrochemical Society, 153 ͑9͒ A1633-A1636 ͑2006͒
Figure 2. Nitrogen adsorption–desorption isotherms and pore-size distribu-
tion ͑inset͒ for as-prepared tin phosphate.
respectively. The discharged sample showed that the peak had
shifted to a higher 2␪ angle, indicating that the mesoperiodicity of
the annealed mesoporous tin phosphate contracted, corresponding to
a d-spacing of 4.5 0.2 nm. After charging to 1.5 V, the mesopore
array material expanded, showing
a
d-spacing value of
5.0 0.3 nm, which is quite similar to that of the as-prepared
sample. However, the d-spacing trend of the discharged and charged
samples is opposite that of the mesoporous/crystalline composite
consisting of tin phosphate reported by Kim et al.5 This finding may
be associated with the different nature of the porewall frame struc-
ture, and further investigation into the origin of this difference using
transmission electron spectroscopy is in progress.
Figure 3. ͑a͒ Potential profiles of mesoporous tin phosphates in a coin-type
half-cell as a function of cycle number between 0 and 1.5 V at a rate of
200 mA/g. ͑b͒ Differential plot of ͑a͒.
Figure 2 exhibits N2 adsorption and desorption isotherms of the
as-prepared sample. Barrett–Joyner–Halenda ͑BJH͒ analysis15 ͑in-
set͒ confirmed well-ordered mesoporous tin phosphate with a pore
size of 3 nm. Its BET specific surface area was 200 m2/g, which is
a quite reasonable value compared with other mesoporous oxides
such as TiO2, SnO2, tin phosphate, and iron phosphates with BET
surface areas of ϳ150–200 m2/g.13,16,17
teristic of framework-confined mesopores. BJH analysis of the dis-
charged and charged samples indicates pore sizes of 2 and 3 nm,
respectively ͑Fig. 4a and b insets͒. BET specific surface area of the
discharged and charged samples was 198 and 205 m2/g, respec-
tively. During lithium alloying ͑discharging͒, pore size was slightly
contracted by 0.5 nm, while it expanded slightly within the range of
1 nm. Hence, we believe that BET surface area did not change sig-
nificantly. Porewall size can be directly estimated from the differ-
ence between d-spacing and pore size values. Therefore, porewall
size of the discharged and charged samples can be deduced to be 2.5
and 2 nm, respectively. The results showed that the initial pore size
of 3 nm was slightly contracted to 2.5 nm, but the pore wall size of
2 nm was slightly expanded to 2 nm after discharging. Hence, the
volume change is negligible compared with that of SnO2 or Sn
particles which show 300% volume increase, resulting in particle
aggregation and pulverization. The broadened SAXS peak of the
charged sample shows partial destruction of the mesopore ordering
despite the maintenance and stability of the mesopore structure dur-
ing the discharge–charge cycle.
The broad peak near ϳ25° in curves e and f ͑discharged and
charged tin phosphate, respectively͒ in Fig. 1 is believed to stem
from the amorphous lithium phosphates matrix. The formation of
the amorphous lithium phosphates phase during lithium alloying/
dealloying was reported by Xiao et al.10 In addition, it has been
reported that the Li4.4Sn and Sn phases at 0 and 1.5 V, respectively,
showed major peaks between 20 and 32° in the XRD patterns.9,21
However, our XRD patterns did not show any peaks, indicating that
the Li4.4Sn and Sn phases did not grow into enough larger clusters to
be detectable by XRD.
Figure 3 shows cycling performances between 1.5 and 0 V of the
mesoporous tin phosphate in a coin-type half-cell at a rate of
200 mA/g. The first charge capacity was 610 mAh/g, with a large
irreversible capacity of 662 mAh/g, and capacity retention after 40
cycles was 93% of the first charge capacity. This capacity retention
is far superior to that of previously reported anode materials, such as
amorphous or crystalline-tin-containing oxides.18-20 Differential ca-
pacity plots ͑Fig. 3b͒ of the coin-type half-cell containing mesopo-
rous tin phosphate were constructed from the corresponding cycling
curves in Fig. 3b. The sharp peak at 1.2 V during the first discharge
is related to the reduction of the tin phosphate to tin and lithium
phosphate phases. The peaks at around 0.2 and 0.5 V are lithium
insertion/extraction reactions associated with lithium-tin alloy
phases of different compositions. As the differential capacity plots
are sensitive detectors of change in the voltage profiles from cycle to
cycle, the constancy of the plots is indicative of the good reversibil-
ity of the electrode material in repetitive charge and discharge op-
erations. However, maintaining broad peaks out to 40 cycles indi-
cates that tin nanoparticles are confined in the porewall frameworks
without growing into large clusters. This observation contrasts with
that of tin and tin oxide composite glasses reported by Courtney and
Dahn.18 Their dq/d͑mdV͒ was smooth for the first six cycles but
turned into strong sharp peaks after 15 cycles, indicating the forma-
tion of large tin clusters.
The major challenge in obtaining mesoporous oxides is to pre-
serve the mesoporous structure while the surfactant is being re-
moved during annealing. In order to avoid severe coarsening or
collapsing of the mesostructure, it is necessary to form a rigid, three-
dimensional, inorganic framework based on Sn-O-P bonding before
removing the surfactant. In our material, it is believed that the
In order to obtain more information on the pore size distribution
and BET surface area of the discharged and charged samples, N2
adsorption and desorption isotherm tests were carried out. Figure 4a
and b shows the N2 adsorption and desorption isotherms of the
discharged and charged tin phosphate, respectively, that are charac-
Downloaded on 2015-03-07 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑9͒ A1633-A1636 ͑2006͒
A1635
Figure 5. ͑a͒ Electron probe image of the nanocrystals in the charged tin
phosphate sample as indicated by the arrows. ͑b͒ NED pattern from the
nanocrystalline area containing amorphous lithium phosphate. ͑c͒ Radial in-
tensity profile of the NED pattern ͑Fig. 5b͒ obtained from the area of the
nanocrystals.
Figure 4. Nitrogen adsorption–desorption isotherms and pore-size distribu-
tion ͑inset͒ for ͑a͒ discharged and ͑b͒ charged tin phosphate to 0 and 1.5 V,
respectively. The sample was outgassed overnight at 150°C prior to analysis.
Acknowledgments
The authors thank Professor J. M. Zuo of the University of Illi-
nois, Urbana-Champaign, for NED analysis and for helpful discus-
sions. The experimental measurement of NED was carried out at the
Center for Microanalysis of Materials, University of Illinois, which
was partially supported by the U.S. Department of Energy, grant no.
DEFG02-91-ER45439. This work was supported by a Korea Re-
search Foundation grant funded by the Korean government
͑MOEHRD͒ ͓C00230 ͑R05-2004-000-10029-0͔͒.
Sn-O-P bonds are formed during the aging process through water
condensation. Once a condensed, 3D mesoporous network is
formed, its essential mesoporous features are likely to be preserved
after removing the surfactant at 500°C.
In order to confirm the formation of Sn nanoparticles in the
charged sample to 1.5 V, coherent nanoarea electron diffraction
͑NED͒ techniques were applied.22 The high angular resolution and
the small probe size of the NED allowed us to directly determine the
structure of very small nanoparticles. Figure 5a shows the electron
probe image of the nanocrystals in the charged tin phosphate as
indicated by the arrows. The electron probe was measured to be
50 nm in diameter, which was sufficient to obtain a high intensity of
NED patterns from the area of nanocrystals. Figure 5b shows two
additional bright diffraction rings ͑marked 1 and 2 in the right figure
of Fig. 5b͒ from small nanoparticles. Radial intensity distribution
analysis of the NED pattern was performed to measure the atomic
distance of rings 1 and 2. Figure 5c shows the radial intensity profile
of the NED pattern obtained from the area of nanocrystals. The
reciprocal distances of peaks 1 and 2 measured from the profile were
2.67 nm−1 ͑3.75 Å͒ and 5.13 nm−1 ͑1.95 Å͒, which corresponded to
the ͑111͒ and ͑311͒ reflections of ␣-Sn ͑face-centered-cubic͒, re-
spectively. The Sn nanocrystals are electrochemically formed from
the amorphous lithium phosphate framework during Li dealloying,
and the porewall structure is effectively conserved without collaps-
ing during the discharge/charge cycle.
Kumoh National Institute of Technology assisted in meeting the publica-
tion costs of this article.
References
1. J. El Haskouri, R. S. Cabrera, M. Bertran-Polter, D. Beltranpoter, M. D. Marcos,
and P. Amoros, Chem. Mater., 11, 1446 ͑1999͒.
2. C. Serre, C. Magnier, M. Hervieu, F. Taulelle, and G. Ferey, Chem. Mater., 14, 180
͑2002͒.
3. C. Serre, A. Auroux, A. Gervasini, M. Hervieu, and G. Ferey, Angew. Chem., Int.
Ed. Engl., 35, 541 ͑1996͒; Angew. Chem., Int. Ed., 41, 1594 ͑2004͒.
4. J. Y. Ying, C. P. Mehnert, and M. S. Wong, Angew. Chem., Int. Ed., 38, 56 ͑1999͒.
5. E. Kim, D. Son, T.-G. Kim, J. Cho, B. Park, K. S. Ryu, and S. H. Chang, Angew.
Chem., Int. Ed., 43, 5987 ͑2004͒.
6. E. Kim, M. G. Kim, and J. Cho, Electrochem. Solid-State Lett., 8, A452 ͑2005͒.
7. F. Chen and M. Liu, Chem. Commun. (Cambridge), 1999, 1829.
8. I. A. Courtney and J. R. Dahn, J. Electrochem. Soc., 144, 2943 ͑1997͒.
9. I. A. Courtney and J. R. Dahn, J. Electrochem. Soc., 144, 2045 ͑1997͒.
10. Y. W. Xiao, J. Y. Lee, A. S. Yu, and Z. L. Liu, J. Electrochem. Soc., 146, 3623
͑1999͒.
11. J. Cho, Y. J. Kim, T.-J. Kim, and B. Park, Angew. Chem., Int. Ed., 40, 3367 ͑2001͒.
12. J. Cho, Y.-W. Kim, B. Kim, J.-G. Lee, and B. Park, Angew. Chem., Int. Ed., 42,
1618 ͑2003͒.
13. N. K. Mal, S. Ichikawa, and M. Fujiwara, Chem. Commun. (Cambridge), 2002,
112.
14. S. Cabrera, J. E. Haskouri, J. Alamo, A. Beltran, D. Beltran, S. Mendioroz, M. D.
Marcos, and P. Amoros, Adv. Mater. (Weinheim, Ger.), 1, 379 ͑1999͒.
15. P. A. Webb and C. Orr, Analytical Methods in Fine Particle Technology, Micromer-
itics Instruments Corp., Norcross, GA ͑1997͒.
Conclusion
We observed reversible changes in the pore and porewall of tin
phosphate during lithium alloying/dealloying, within a range of
Ͻ1 nm. Even after charging to 1.5 V, mesopore order was well
preserved, and ϳ2 nm sized tin particles were grown in the pore-
wall framework.
16. X. Guo, W. Ding, X. Wang, and Q. Yan, Chem. Commun. (Cambridge), 2001, 709.
17. P. Yang, D. Zhao, D. I. Margolese, B. F. Chmelka, and G. D. Stucky, Nature
Downloaded on 2015-03-07 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1636
Journal of The Electrochemical Society, 153 ͑9͒ A1633-A1636 ͑2006͒
(London), 396, 152 ͑1998͒.
18. I. A. Courtney, W. R. Mckinnon, and J. R. Dahn, J. Electrochem. Soc., 146, 59
͑1999͒.
20. M. Behm and J. T. S. Irvine, Electrochim. Acta, 47, 1727 ͑2002͒.
21. C. Kim, M. Noh, M. Choi, J. Cho, and B. Park, Chem. Mater., 17, 3297 ͑2005͒.
22. J. M. Zuo, M. Gao, J. Tao, B. Q. Li, R. Twesten, and I. Petrov, Microsc. Res. Tech.,
64, 347 ͑2004͒.
19. K. Wan, S. F. Y. Li, Z. Gau, and K. S. Siow, J. Power Sources, 75, 9 ͑1998͒.
Downloaded on 2015-03-07 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).