.

Angewandte
Communications

DOI: 10.1002/anie.201300680
Lithium Batteries

Lithium Polysulfidophosphates: A Family of Lithium-Conducting

Sulfur-Rich Compounds for Lithium–Sulfur Batteries**
Zhan Lin, Zengcai Liu, Wujun Fu, Nancy J. Dudney, and Chengdu Liang*

The conflict of the ever-increasing demand of energy of Li-S batteries is rooted in the PS phenomenon that is cause
consumption and the public concern over environmental by the use of liquid electrolyte. Furthermore, stable long-term
issues demands a sustainable energy future that requires cycling of metallic lithium anode in some organic liquid
innovations on harvesting and delivering clean energies. electrolytes such as carbonate-based electrolytes, is a long-
Batteries have been brought into the spotlight as the pivotal standing problem owing to the formation of an unstable solid
technology for the electricity-based energy infrastructure that electrolyte interphase and the dendritic growth of lithium
incorporates intermittent renewable energies. Although lith- deposition.[6]
ium ion (Li-ion) batteries have succeeded in powering Using solid electrolytes in Li-S batteries breaks the
personal electronics and small electric tools, they face the standard of conventional Li-S batteries. This game-changing
challenges of safety, cost, and energy density for the needs of shift from liquid to solid electrolytes eliminates the poly-
independent electrified transportation and large-scale energy sulfide shuttle and enables stable cycling of metallic lithium
storage.[1] As compared to the current Li-ion batteries, anodes. With the emergence of solid electrolytes with ionic
Lithium–sulfur (Li-S) batteries are a promising candidate conductivities comparable to that of liquid electrolytes,[7] all-
technology with advantages of abundant materials and solid-state Li-S batteries promise to be the next breakthrough
inherently high energy. However, achieving longevity of for electric energy storage.[8] As in conventional Li-S bat-
cycling in Li-S batteries is formidable because of the electrical teries, the cycling of all-solid-state Li-S batteries has to
insulating nature of elemental sulfur and the poor ionic overcome the poor electronic and ionic conductivities of the
conductivity of sulfur and its discharge products. After nearly sulfur cathode. Seamless transport of ions and transfer of
four decades of research, significant progress has been made electrons are basic requirements for batteries. Although the
to Li-S batteries: various carbon materials and conducting electronic conductivity of the sulfur cathode can be improved
polymers impart electronic conductivity to the sulfur cath- by add carbon to the cathode, the ionic conductivity of sulfur
ode,[2] and organic liquid electrolytes with good solubility to cannot be significantly improved by the simple mixing of solid
lithium polysulfides overcome the poor ionic conductivities of electrolytes with sulfur. There is a critical knowledge gap
the sulfur cathode.[3] Using a liquid electrolyte that dissolves concerning the ionic conductivity of the sulfur cathode at the
lithium polysulfides is a necessity in cycling conventional Li-S solid state. We report herein the first series of sulfur-rich
batteries: sulfur is actually cycling as a solution, also known as compounds that have high ionic conductivities and excellent
the catholyte.[4] The dissolved polysulfides move sulfur inside electrochemical reversibility during battery cycling.
the electrochemical cells, which leads to parasitic side Elemental sulfur (S8) reacts with lithium thiophosphate
reactions that have been well-documented as the polysulfide (Li3PS4) in tetrahydrofuran (THF) following the equation in
shuttle (PS).[5] The short cycle-life and low energy efficiency Scheme 1 a. The reaction occurs at the negatively charged
terminal sulfur atoms of PS43. Polysulfidophosphate anions

[*] Dr. Z. Lin, Dr. N. J. Dudney

Materials Science and Technology Division
Oak Ridge National Laboratory
Oak Ridge, TN 37831-6124 (USA)
Dr. Z. Liu, Dr. W. Fu, Dr. C. Liang
Center for Nanophase Materials Sciences
Oak Ridge National Laboratory
Oak Ridge, TN 37831-6493 (USA)
E-mail: [email protected]
[**] This research was sponsored by U.S. Department of Energy (DOE)/
Energy Efficiency and Renewable Energy (EERE) through Vehicle
Technologies Office. The investigation of the ionic conductivity of
these new compounds was supported by the Division of Materials
Science and Engineering, Office of Basic Energy Sciences, U.S.
Department of Energy (DOE). The synthesis and characterization Scheme 1. Chemical reactions of lithium polysulfidophosphates
was conducted at the Center for Nanophase Materials Sciences, (LPSPs). a) Reaction of sulfur with Li3PS4 yields Li3PS4+n,
which is sponsored at Oak Ridge National Laboratory by the n = (x + y + z)/3, at room temperature in THF. Formation of the SS
Division of Scientific User Facilities, U.S. DOE. bonds links S atoms to the core PS4 unit. b) The electrochemical
Supporting information for this article is available on the WWW reaction mechanism for charge and discharge of Li3PS4+n. Reversible
under http://dx.doi.org/10.1002/anie.201300680. scission and formation of SS bonds in LPSP.

7460  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2013, 52, 7460 –7463
 Angewandte
Chemie

Figure 2. Conductivity of lithium polysulfidophosphates. a) Room-tem-

perature ionic conductivity as a function of n (SS bonds) in Li3PS4+n
(n = 2, 3, 5, 6, and 8) at 25 8C. b) Temperature dependency of ionic
conductivity of Li3PS4+5 (&, Ea = 0.37 eV) and Li2S (*, Ea = 0.74 eV).

log s ¼ 3:486  0:219 n ð1Þ

Figure 1. XRD patterns and Raman spectra to confirm the formation of

Temperature dependence of ionic conductivity was inves-
lithium polysulfidophosphates. a) XRD patterns of S, Li3PS4 and
Li3PS4+n (n = 5) with/without THF; b) Raman spectra of S, Li3PS4, and
tigated for Li3PS4+5 (Figure 2 b). This compound has a room-
Li3PS4+n (n = 3, 5, and 6); c), d) Raman shift between the wavelengths temperature ionic conductivity of 3.0  105 S cm1 at 25 8C.
of c) 125–175 cm1 and d) 460–490 cm1. The activation energy was derived from Arrhenius equation:

s ¼ A exp ðEa =kTÞ ð2Þ

are formed through the formation of SS single bonds. The Where s is the ionic conductivity, T is the absolute
reaction products are lithium polysulfidophosphates (LPSP), temperature, k is the Boltzmann constant, A is the pre-
which have the general formula Li3PS4+n (0 < n < 9). The n in exponential factor, and Ea is the activation energy. When log s
the formula represents the number of SS bonds in the LPSP is plotted against 1/T, a straight line is expected with a slope of
molecule. X-ray diffraction (XRD) and Raman spectra in Ea/k and an intercept of log A. The calculated activation
Figure 1 verify the formation of this new family of sulfur-rich energy from Figure 2 b is 0.37 eV, which is comparable to that
compounds. Comparing the XRD patterns of sulfur and of the lithium super ionic conductor (LiSICON).[12] The ionic
Li3PS4·3 THF in Figure 1 a, a new crystalline structure of the conductivity of Li3PS4+5 is increased by a factor of 10 when the
sulfur compound Li3PS4+n·x THF (n = 5) is apparent in the temperature is raised to 60 8C. As a comparison, we measured
reaction. After the removal of THF at 80 8C under vacuum, the ionic conductivity of Li2S, the discharge product of
the parent compound Li3PS4 becomes amorphous while the elemental sulfur, at the same conditions. As shown in
Li3PS4+n (n = 5) is in crystalline form. Raman spectra (Fig- Figure 2 b, the ionic conductivity of LPSP is 8 orders of
ure 1 b) of Li3PS4+n (n = 3, 5, and 6) show the symmetric magnitude higher than that of Li2S (ca. 1013 S cm1 at 25 8C).
stretching of the PS bond at 418 cm1 and the stretching Excellent cyclability was demonstrated using the com-
vibrations of the PS bond of 184 cm1 regardless of number pound Li3PS4+5 as the cathode material at room temperature.
of sulfur atoms attached to the PS43 anion.[9] With striking The cathode shows an initial discharge capacity of
contrast to Li3PS4, the peaks centered at 155, 218/476, and 1272 mAh g1 (based on the incorporated sulfur content) or
248/437 cml appear for the LPSP compounds. These peaks 599 mAh g1 (based on the compound; Figure 3 a) with 100 %
are signatures of the E2, A1, and E3 symmetry species of the of coulombic efficiency after a few initial cycles (Supporting
SS bond and confirm the formation of polysulfidophosphate Information, Figure S1; unless otherwise noted, the capacities
anions.[10] With a decrease in sulfur atoms in LPSP com- hereafter are normalized to the sulfur content within the
pounds, the SS bond peaks at 155 and 476 cm1 shift to lower compound; for convenience, the normalized capacities based
wavelengths (Figure 1 c,d). The downshifting of these peaks on the sulfur compound are also shown in figures). The
results from the vibrational effect of the PS bond in core PS4 cathode capacity stabilizes at 700 mA h g1 after 300 cycles.
unit and the increase in the sulfur chain length at the Even better cycling performance was observed at 60 8C. The
terminals.[11] initial capacity was over 1400 mA h g1. A high capacity of
These LPSP compounds are lithium-ion conductors with 1200 mA h g1 was maintained after 300 cycles. This observa-
room-temperature ionic conductivity in the range of 104 to tion confirms that the LPSP has an extremely stable cycling
106 S cm1 depending on the number of sulfur atoms. The performance. The improved cycling performance is attributed
correlation of ionic conductivity to the number of additional to the increased ionic conductivity of both the solid electro-
sulfur atoms incorporated into the parent compound Li3PS4 is lyte and the sulfur-rich cathode at 60 8C. Increased temper-
shown in Figure 2 a. The log scale of the ionic conductivity ature may favor the electrochemical reactions as well. Polar-
(log s) of LPSP follows a linear relationship with the number ization of the cell was large at room temperature and
of sulfur atoms (n) described in Equation (1): significantly reduced at 60 8C. Energy efficiency was

Angew. Chem. Int. Ed. 2013, 52, 7460 –7463  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 7461
 .
Angewandte
Communications

Figure 4. Raman spectra to unravel the electrochemical cycling mech-

anism. Raman spectra of the electrode materials consisting of carbon
and Li3PS4+n (n = 5) before cycling (a), at the end of first discharge
cycle (b), and at the end of first charge cycle (c). ^ peaks from the S
Figure 3. Electrochemical evaluation of Li3PS4+5 as the cathode for all- S bond in Li3PS4+n, ! peaks from the PS bond in Li3PS4, I the peak
solid-state Li-S batteries. a) The cycling performance at the rate of C/ of Li2S.
10 at room temperature and 60 8C (pink/red charge, black/blue di-
scharge); b) representative voltage profile at room temperature (red
charge, black discharge); c) representative voltage profile at 60 8C; and and yields Li2S and Li3PS4. When the cathode was charged
d) the rate performance at 60 8C from C/10 to 2C.
again, SS bond features reappeared with the disappearance
of Li2S and Li3PS4 at the end of the charge cycle. The Raman
spectra of the pristine material and the charged cathode are
improved from 59.7 % at room temperature to 82.8 % at 60 8C very much the same. These Raman spectra confirmed that the
(Figure 3 b,c). A good rate performance at 60 8C is shown in electrochemical insertion and removal of lithium ion in the
Figure 3 d. The cell has a capacity over 1200 mA h g1 at C/10 cathode is a reversible reaction of Li3PS4+5 to a mixture of
and shows a reversible capacity of 735 mA h g1 at 2C after 50 Li2S and Li3PS4. The electrochemical scission and formation
cycles at various rates. Further cycling at a low rate of C/10 of SS bonds are consistent with the reports of cycling
brings it back to a reversible capacity of 1200 mA h g1. A organosulfur compounds with the thiol (SH)/disulfide (SS)
control experiment using Li3PS4 and carbon material at 60 8C redox couple.[13]
showed negligible capacity contribution from carbon (Sup- In summary, elemental sulfur reacts with Li3PS4. The
porting Information, Figure S2). In a conventional Li-S reaction adds sulfur atoms to the charged terminal S atoms of
battery, the dissolution and migration of sulfur species cause PS43 anion by forming SS bonds. The addition of sulfur to
the cathode structure to deteriorate which leads to a short PS43 yields a new family of sulfur-rich compounds. These
cycle-life. Conversely, the integrity of the LPSP cathode compounds have a lithium-ion conductivity in the range of
structure was preserved after 300 deep cycles at 60 8C. SEM 105 to 106 S cm1 that is comparable to that of conventional
images and elemental maps of the pristine LPSP cathode and Li-ion cathode materials, such as lithium metal oxides and
the LSPS cathode after 300 cycles are compared in the phosphates. Reversible electrochemical reactions occur
Supporting Information, Figure S3. No distinct difference was through the breaking and forming SS bonds in LPSP
observed in these images. compounds when they are applied as the cathode materials
The voltage profile of the LPSP cathode does not show for all-solid-state Li-S batteries. These materials are excellent
two voltage plateaus that are normally observed in conven- cathode materials for such Li-S batteries with an impressive
tional Li-S batteries. The difference in voltage profiles of cyclability. The ionic conductivity of the sulfur cathode is of
LPSP and conventional S indicates the electrochemical paramount importance in the cycling of these Li-S batteries.
reaction of LPSP follows a different reaction pathway. The However, this topic has been rarely studied for conventional
speculated reaction mechanism is depicted in Scheme 1 b; the Li-S batteries. This research bridges the knowledge gap
reversible scission and formation of SS bonds at the terminal between convention Li-S batteries and the new trend of all-
S atoms accounts for the discharge/charge of LPSP. To solid-state Li-S batteries.
unravel the charge/discharge mechanism of these sulfur-rich
compounds, Raman spectra (Figure 4) were taken on the
pristine material and at the end of each of the charge and Experimental Section
discharge cycles. The peaks centred at 155, 218/476, and Li3PS4 was prepared by a procedure described previously.[14] The
180 cml of the pristine Li3PS4+5 are attributed to the SS sulfur compounds Li3PS4+n (n = 2,3,5,6, and 8) were synthesized
bond. After the first discharge cycle, the characteristic peaks through solution reactions of Li3PS4 and elemental sulfur in
tetrahydrofuran (THF) under magnetic stirring without exposure to
of the SS bond at 155, 218/476, and 180 cml had vanished.
air or moisture. After the solids were fully dissolved, the solution was
New strong peaks at 375 and 418 cml and a weak peak at dried under vacuum for 2 h at 80 8C. The recovered solids were
283 cml appeared as evidence of the formation of Li2S and lithium polysulfidophosphates. The cathode slurry was prepared by
Li3PS4. The discharge process breaks the SS bond in Li3PS4+5 dispersing Li3PS4+5 (60 wt %), WVA-1500 carbon (MeadWestvaco

7462 www.angewandte.org  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2013, 52, 7460 –7463
 Angewandte
Chemie

Corporation, 30 wt %), and PVC binder (10 wt %) in THF. No side Carbon 2008, 46, 229 – 235; e) J. L. Wang, J. Yang, C. R. Wan, K.
reaction occurred to the binder during the preparation of electrode Du, J. Y. Xie, N. X. Xu, Adv. Funct. Mater. 2003, 13, 487 – 492;
because of the rapid evaporation of solvent.[15] A solid electrolyte f) L. F. Xiao, Y. L. Cao, J. Xiao, B. Schwenzer, M. H. Engelhard,
pellet of Li3PS4 with a lithium metal anode was prepared by cold- L. V. Saraf, Z. M. Nie, G. J. Exarhos, J. Liu, Adv. Mater. 2012, 24,
pressing nanoporous b-Li3PS4 (120 mg) on top of a thin lithium foil 1176 – 1181; g) Y. Yang, G. H. Yu, J. J. Cha, H. Wu, M.
(50 mm thick) under 240 MPa pressure in a 13 mm die. The slurry was Vosgueritchian, Y. Yao, Z. A. Bao, Y. Cui, ACS Nano 2011, 5,
then coated on the other side of the pellet and dried under vacuum at 9187 – 9193.
80 8C. The loading of the cathode materials is 0.25–0.60 mg cm2, [3] a) J. Shim, K. A. Striebel, E. J. Cairns, J. Electrochem. Soc. 2002,
which is equivalent to 0.15–0.36 mAh cm2. The final thickness of the 149, A1321 – A1325; b) J. Gao, M. A. Lowe, Y. Kiya, H. D.
pellets was 0.5 mm. Abruna, J. Phys. Chem. C 2011, 115, 25132 – 25137; c) Y. Jung, S.
The ionic conductivity of LPSP compounds was conducted in the Kim, B. S. Kim, D. H. Han, S. M. Park, J. Kwak, Int. J. Electro-
frequency range of 10 MHz to 1 Hz with an amplitude of 10 mV by chem. Sci. 2008, 3, 566 – 577; d) J. H. Shin, E. J. Cairns, J. Power
using a frequency response analyzer (Solartron 1260). Swagelok cells Sources 2008, 177, 537 – 545.
were used to evaluate the cycling performance. Carbon-coated [4] a) S. S. Zhang, J. A. Read, J. Power Sources 2012, 200, 77 – 82;
aluminum foil (samples kindly provided by Exopack Advanced b) S. S. Zhang, D. T. Tran, J. Power Sources 2012, 211, 169 – 172;
Coating) was used as the current collector. Charge and discharge c) Z. Lin, Z. Liu, W. Fu, N. J. Dudney, C. Liang, Adv. Funct.
cycles were carried out using a Maccor 4000 series battery tester at Mater. 2013, 23, 1064 – 1069.
a current density of 0.015 mA cm2 (C/10) between the cut-off [5] a) Y. V. Mikhaylik, J. R. Akridge, J. Electrochem. Soc. 2004, 151,
potentials of 1.5–2.8 V vs. Li/Li+. The current densities of 0.03 (C/ A1969 – A1976; b) S. S. Zhang, Electrochim. Acta 2012, 70, 344 –
5), 0.075 (C/2.5), 0.15 (C), and 0.2 (2C) mA cm2 were applied to 348.
measure the rate performance of the LPSP cathode. The calculation [6] D. Fauteux, R. Koksbang, J. Appl. Electrochem. 1993, 23, 1 – 10.
of specific charge/discharge capacities was based on the mass of sulfur [7] N. Kamaya, K. Homma, Y. Yamakawa, M. Hirayama, R. Kanno,
and the compound, respectively. M. Yonemura, T. Kamiyama, Y. Kato, S. Hama, K. Kawamoto,
The morphologies of the cathode before and after cycling were A. Mitsui, Nat. Mater. 2011, 10, 682 – 686.
examined using a field emission scanning electron microscope (FE- [8] a) T. Kobayashi, Y. Imade, D. Shishihara, K. Homma, M. Nagao,
SEM; Zeiss Merlin) at 15 kV. The elemental maps of carbon, sulfur, R. Watanabe, T. Yokoi, A. Yamada, R. Kanno, T. Tatsumi, J.
and phosphorus were taken using the energy-dispersive spectroscope Power Sources 2008, 182, 621 – 625; b) F. Mizuno, S. Hama, A.
of the FE-SEM. X-ray diffraction (XRD) analysis was performed Hayashi, K. Tadanaga, T. Minami, M. Tatsumisago, Chem. Lett.
using a PANalytical X’pert PRO 2-circle X-ray diffractometer with 2002, 1244 – 1245; c) J. Hassoun, B. Scrosati, Adv. Mater. 2010,
a CuKa source (l  1.5418 ). Raman spectroscopy was recorded 22, 5198 – 5201; d) Z. Lin, Z. Liu, N. J. Dudney, C. Liang, ACS
from 600 to 100 cm1 on a Renishaw Confocal MicroRaman Nano 2013, 7, 2829 – 2833.
spectrometer at room temperature. A green laser with wavelength [9] a) F. Mizuno, A. Hayashi, K. Tadanaga, M. Tatsumisago, Adv.
of 532 nm was used for Raman excitation. Mater. 2005, 17, 918 – 921; b) A. Mller, N. Mohan, P. Cristoph, I.
Tossidis, M. Drager, Spectrochim. Acta Part A 1973, 29, 1345 –
Received: January 25, 2013 1356.
Revised: March 6, 2013 [10] a) D. W. Scott, F. H. Kruse, J. P. Mccullough, J. Mol. Spectrosc.
Published online: June 4, 2013 1964, 13, 313 – 320; b) A. T. Ward, J. Phys. Chem. 1968, 72, 4133 –
4139.

.
Keywords: batteries · ionic conductivity · lithium ·
sulfidophosphates · sulfur
[11]
[12]
R. Steudel, Z. Naturforsch. B 1975, 30, 281 – 282.
G. Y. Adachi, N. Imanaka, H. Aono, Adv. Mater. 1996, 8, 127 –
135.
[13] a) Y. Nuli, Z. P. Guo, H. K. Liu, J. Yang, Electrochem. Commun.
2007, 9, 1913 – 1917; b) Y. J. Li, H. Zhan, L. B. Kong, C. M. Zhan,
[1] a) M. Armand, J. M. Tarascon, Nature 2008, 451, 652 – 657; Y. H. Zhou, Electrochem. Commun. 2007, 9, 1217 – 1221.
b) L. W. Ji, Z. Lin, M. Alcoutlabi, X. W. Zhang, Energy Environ. [14] Z. Liu, W. Fu, E. A. Payzant, X. Yu, Z. Wu, N. J. Dudney, J.
Sci. 2011, 4, 2682 – 2699. Kiggans, K. Hong, A. J. Rondinone, C. Liang, J. Am. Chem. Soc.
[2] a) X. L. Ji, K. T. Lee, L. F. Nazar, Nat. Mater. 2009, 8, 500 – 506; 2013, 135, 975 – 978.
b) G. Y. Zheng, Y. Yang, J. J. Cha, S. S. Hong, Y. Cui, Nano Lett. [15] G. X. Xu, L. Qi, B. T. Yu, L. Wen, Chin. J. Polym. Sci. 2006, 24 –
2011, 11, 4462 – 4467; c) J. C. Guo, Y. H. Xu, C. S. Wang, Nano 3, 307 – 313.
Lett. 2011, 11, 4288 – 4294; d) J. Wang, S. Y. Chew, Z. W. Zhao, S.
Ashraf, D. Wexler, J. Chen, S. H. Ng, S. L. Chou, H. K. Liu,

Angew. Chem. Int. Ed. 2013, 52, 7460 –7463  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 7463