Solid state NMR investigation of hydrous ruthenium oxide

Article abstract of DOI:10.1016/S0009-2614(00)01169-6

Amorphous hydrous ruthenium oxide (RuO₂·xH₂O) with different composition x has been studied using solid-state nuclear magnetic resonance (NMR) spectroscopy. The ²DNMR spectra at different temperatures illustrate that the water molecules undergo fast molecular motion even if the temperature is as low as 213 K. The static ¹HNMR spectra indicate the composition dependent proton-proton dipolar interaction. It is demonstrated that the mobility of the water molecules and their interaction with ruthenium oxides play an important role in the proton charge density. In conclusion, the competition between these two antithetical effects provides a mechanism for the proton charge storage of the RuO₂·xH₂O materials.

Full text of DOI:10.1016/S0009-2614(00)01169-6

24 November 2000
Chemical Physics Letters 331 (2000) 64±70
www.elsevier.nl/locate/cplett
Solid state NMR investigation of hydrous ruthenium oxide
Zhiru Ma a, Jim P. Zheng b, Riqiang Fu a,*,1
a
Center for Interdisciplinary Magnetic Resonance, National High Magnetic Field Laboratory, 1800 E. Paul Dirac Drive,
Tallahassee, FL 32310, USA
b
Department of Electrical and Computer Engineering, Florida A & M University and Florida State University,
Tallahassee, FL 32310, USA
Received 13 July 2000; in ®nal form 25 September 2000
Abstract
Amorphous hydrous ruthenium oxide (RuO2 Á xH2O) with di€erent composition x has been studied using solid-state
2
nuclear magnetic resonance (NMR) spectroscopy. The D NMR spectra at di€erent temperatures illustrate that the
water molecules undergo fast molecular motion even if the temperature is as low as 213 K. The static 1H NMR spectra
indicate the composition dependent proton±proton dipolar interaction. It is demonstrated that the mobility of the water
molecules and their interaction with ruthenium oxides play an important role in the proton charge density. In con-
clusion, the competition between these two antithetical e€ects provides a mechanism for the proton charge storage of
the RuO2 Á xH2O materials. Ó 2000 Elsevier Science B.V. All rights reserved.
1. Introduction
capacitance (i.e., maximum charge density), high
conductivity, good electrochemical reversibility, as
well as high power and high energy density.
However, this material has not been well charac-
terized due to its amorphous nature and compo-
sitional variability. Here, we utilize solid-state
nuclear magnetic resonance (NMR) spectroscopy,
a unique technique well suitable for dynamics
characterization in materials even in an amor-
phous phase [8], to study the hydrous ruthenium
oxide material, thus providing insights into the
implications of its structural water.
Anhydrous RuO2, whose crystalline structure is
rutile, is one of the best pseudocapacitive materials
and its energy storage mechanism is believed to be
the so-called fast faradaic reactions with ions.
Such reactions are typically limited to the surface
area due to its rutile structure. In contrast,
RuO2 Á xH2O is amorphous and the protons can
easily intercalate and di€use into the bulk of the
Hydrous ruthenium oxides (RuO2 Á xH2O) have
attracted great interest recently because of im-
portant technological applications in electrocatal-
ysis [1,2] and high-charge storage-capacity devices
(e.g., electrochemical capacitors) [3±7]. For exam-
ple, the presence of hydrous ruthenium oxide in
the nanoscale Pt±Ru electrocatalysts used in the
direct methanol fuel cell provide important con-
tributions to the electrocatalytic oxidation of
methanol due to its mixed electron±proton con-
ductivity [1]. Moreover, RuO2 Á xH2O is an excel-
lent electrode material for electrochemical
capacitors [3,4,6,7] because it exhibits high speci®c
* Corresponding author. Fax: +1-850-644-1366.
E-mail address: rfu@magnet.fsu.edu (R. Fu).
1
Also a member of the Department of Chemistry, Florida
State University, Tallahassee, FL 32306, USA.
0009-2614/00/$ - see front matter Ó 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 ( 0 0 ) 0 1 1 6 9 - 6

Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
65
material. As a result, the surface interaction area is
e€ectively increased resulting in a higher speci®c
capacitance. The maximum speci®c capacitance of
amorphous RuO2 Á xH2O obtained yet [6] was over
768 F/g, which is more than two times greater than
that from the RuO2 material. However, this is still
a largely qualitative theory to explain why the
amorphous ruthenium oxide is superior to the
crystalline material. In addition, the speci®c ca-
pacitance of RuO2 Á xH2O is also a€ected by the
hydrogen concentration [4]. Although a high water
content provides more protons available for the
intercalation, it results in a low speci®c capaci-
tance. A very lower concentration also gives rise to
a low speci®c capacitance. The maximum speci®c
capacitance is normally obtained in the range
x ˆ 0:5±0:7. Furthermore, the impurities such as
lithium and sodium in the hydrous ruthenium
oxide could dramatically a€ect the speci®c capac-
itance. For instance, 2% sodium in the amorphous
ruthenium oxide could cause a 20% reduction in
the speci®c capacitance [3,4], while the incorpora-
tion of Cr into the hydrous ruthenium oxide
slightly increases the speci®c capacitance [9]. Al-
though many studies of electrochemical properties
of the RuO2 Á xH2O material have been carried
out, the role of the structural water in the amor-
phous hydrous ruthenium oxide has not yet been
materials with di€erent water contents. The
2D NMR spectra at di€erent temperatures are
obtained to demonstrate the mobility of the water
molecules, while the static 1H NMR measure-
ments allow for characterization of the residual
proton±proton dipolar couplings. We will further
discuss the correlation between the mobility of the
water molecules and the proton charge storage,
based on the electrochemical measurements.
2. Materials and experiments
A sample of RuO2 Á 2:662H2O was purchased
from Alfa Aesarâ/Johnson Matthey Company and
was used without further puri®cation. A series of
ruthenium oxide RuO2 Á xH2O materials were
prepared through the annealing processes. The
original RuO2 Á 2:662H2O sample was weighted
and then placed in an oven where di€erent tem-
peratures (from 323 to 573 K) were set to allow the
sample to be annealed for about 24 h. Only the
water of the sample is lost during the annealing
process, thus the water content x can be calculated
through the weight change of the sample before
and after the annealing process. The x values ob-
tained at di€erent annealing temperatures are
shown in Table 1.
2
well understood. A D NMR observation was re-
The RuO2 Á 2:82D2O powder material was pre-
ported on the isotope-exchanged RuO2 Á xD2O and
DdRuO2 Á xD2O samples, indicating substantial
molecular mobility in the materials [10]. Mckeown
et al. [11] characterized the structures of the hy-
drous ruthenium oxides (RuO2 Á 2:32H2O;
RuO2 Á 0:29H2O and anhydrous RuO2) by a com-
bined use of thermogravimetric analysis (TGA),
X-ray di€raction, X-ray absorption near-edge
structure (XANES), and extended X-ray ®ne
structure (EXAFS) analysis. It was found that the
local structures of RuO2 Á 2:32H2O and RuO2 Á
0:29H2O hypothesised from the EXAFS data are
remarkably di€erent due to the presence of struc-
tural water. However, because the orientation of
water is not discernible from the EXAFS data, the
relation between the structural water and the
proton charge storage is not yet clear.
pared using a sol±gel process. The required
Table 1
Water content x of the RuO2 Á xH2O samples vs. annealing
temperature (K)
T (K)
x
298
323
348
373
381
389
398
408
418
428
438
448
473
523
573
2.662
2.271
1.616
1.110
1.012
0.922
0.847
0.769
0.682
0.638
0.586
0.541
0.461
0.356
0.280
In the present Letter, we use the solid-state
NMR technique to characterize RuO2 Á xH2O

66
Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
amount of anhydrous RuCl3 was dissolved in
99.8% deuterium oxide (D2O) to give a concen-
tration of about 0.3 M, and a LiOD solution with
a concentration of about 2.8 M was prepared by
reacting a 99.9% pure lithium metal with the D2O.
The LiOD solution was added slowly into the
RuCl3 solution until the pH value of the mixture
reached 7.0. During the mixing, the RuCl3 solution
was stirred by a magnetic stirring bar. At the pH
value of 7.0, the stirring was stopped, and black
powders precipitated in a clear LiCl solution. The
powders were separated from the LiCl solution by
a ®lter and washed three times in D2O.
scans were used for signal averaging with a recycle
delay of 2 s.
3. Results and discussion
In order to illustrate the mobility of the water
molecules in the RuO2 Á H2O material, the
2D NMR spectra of amorphous RuO2 Á 2:82D2O
were obtained at di€erent temperatures, as shown
in Fig. 1. At room temperature (i.e. 300 K), a
narrow peak with a linewidth of 3 kHz at the half-
height is obtained in the center of the spectrum.
Such a narrow peak re¯ects fast motional aver-
aging, indicating fast mobility for the deuterons in
the bulk of the material [12,13]. When the tem-
perature is 213 K, the peak becomes broader
(ꢀ3:5 kHz at the half-height), implying that the
mobility of the deuterons slows down slightly.
When the temperature reaches 183 K, two reso-
nance signals are superimposed in the spectrum.
The peak in the center is further broadened, im-
plying that the mobility of the deuterons decreases,
while a broad peak (ꢀ200 kHz broad) starts to
appear, indicating that some of the deuterons be-
come rigid. At 173 K, the peak in the center be-
comes broader and smaller, and the broad peak
The speci®c capacitance of the RuO2 Á xH2O
material was measured by cyclic voltammetry
(CV) experiments, in which a three-electrode con-
®guration was employed. A saturated colomel
electrode (SCE) was used as a reference electrode,
a Pt plate as a counter electrode, and the
RuO2 Á xH2O electrode as the working electrode.
The RuO2 Á xH2O electrode was made from
RuO2 Á xH2O powder mixed with about 5% Te¯on
binder and then pressed on the Pt mesh. During
the CV experiments, the voltage between the
working and the reference electrodes was linearly
swept at a constant scan rate of s, at the same time,
the current i ¯owing from the working electrode to
the counter one was measured. The capacitance of
the working electrode was calculated by the
equation cp ˆ i=s. The speci®c capacitance is de-
®ned as the capacitance per unit weight of the
electrode material.
2
The D NMR experiments were carried out on
a CMX-400 NMR spectrometer (B0 ˆ 9:4 T) op-
2
erating at a D Larmor frequency of 61.5 MHz,
equipped with a home-built 2D wideline NMR
probe. A standard quadruple echo sequence (i.e.
90°x±s±90°y±s-acquisition) was used to record the
2D NMR spectra. The echo time s used in the
experiments was 30 ls and the 90° pulse length
was 3 ls. For each experiment, 128 scans were
accumulated with a recycle time of 1 s. All tem-
peratures were controlled within Æ1 K. The static
1H NMR measurements were performed on a
Bruker DMX-300 NMR spectrometer (B0 ˆ 7 T)
with a 1H Larmor frequency of 300 MHz. The
temperatures were controlled within Æ0:1 K by a
Bruker BVT2000 unit. For each experiment, 8
Fig. 1. Static 2D NMR spectra of amorphous RuO2 Á
xD2O ꢁx ˆ 2:82† recorded on a CMX 400 NMR spectrometer
at di€erent temperatures.

Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
67
gains in intensity, which means that most of the
deuterons are rigid.
result from the loss of the water. In fact, the mo-
bility of the hydrogen is inversely proportional to
the interaction (i.e., the hydrogen bonding) be-
tween the water molecules and the RuO2 lattices.
Highly mobile water molecules, which allow more
protons to participate in the energy charge storage,
exhibit a weak hydrogen bonding with the RuO2,
while rigid water molecules present a strong hy-
drogen bonding with the lattices. Therefore, it can
be reasonably concluded from Figs. 1 and 2 that
the capacitance is associated with the competition
between these two antithetical e€ects resulting
from the mobility of the water molecules and their
interaction with the RuO2 lattices. The mobility of
the water molecules tends to enable the protons to
intercalate into the bulk of the material, while their
interaction with ruthenium oxides provides a
mechanism for retaining the protons with the lat-
tice ultimately resulting in high proton charge
density.
Fig. 3 shows the speci®c capacitance as a
function of water content x. Crystalline structures
of RuO2 Á xH2O were characterized using an X-ray
di€ractometer. No di€raction peak was obtained
for materials annealed at temperatures lower than
428 K (i.e., x ˆ 0:638) and the di€raction peaks
corresponding to anhydrous RuO2 were observed
when the temperature was up to 438 K (x ˆ 0:586)
(spectra not shown). These peaks became
narrower and gained in intensity with further
A three-cell electrochemical capacitor was made
with amorphous RuO2 Á xH2O powders (x ˆ 0:769)
as the electrode and 39 wt% sulfuric acid solution
as the electrolyte. The capacitor was charged and
discharged under constant current at di€erent
ambient temperatures. Fig. 2 shows the capaci-
tance as a function of the temperature for the ca-
pacitor. As in Fig. 2, the maximum capacitance
appears at about 255 K. When the temperature
decreases, the measured capacitance steadily be-
comes lower, which is consistent with the decrease
in the mobility of the deuterons at lower temper-
atures as indicated in Fig. 1. It was found experi-
mentally that the feature of the CV curves of the
commercially synthesized RuO2 Á xH2O and the
sol±gel synthesized RuO2 Á xD2O is very similar,
although the latter possesses about 15% lower
speci®c capacitance (unpublished results). It is
anticipated that the capacitance could become
very small when the temperature reaches 173 K,
where most of the deuterons in the bulk material
become rigid and do not participate in energy
charge storage. Surprisingly, the measured capac-
itance is low at higher temperature (e.g., >340 K),
where the hydrogen in the bulk are supposed to be
highly mobile. It is believed that the water content
should not be a€ected when the operating tem-
perature is lower than the annealing temperature
(e.g. 408 k for x ˆ 0:769) at which the sample was
prepared. Thus, the reduced capacitance does not
Fig. 3. Speci®c capacitance of the amorphous RuO2 Á xH2O
samples as a function of water content x. The potential in the
CV experiments was scanned in the range 1±0 V at a scan rate
of 5 mV/s.
Fig. 2. Capacitance of a three-cell electrochemical capacitor
made with the RuO2 Á xH2O ꢁx ˆ 0:769† powder as a function
of temperature.

68
Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
increases in the annealing temperatures. These re-
sults indicate that the material is in an amorphous
phase at the lower annealing temperatures
(<428 K). As shown in Fig. 3, the speci®c capac-
itance of the RuO2 Á xH2O material increases
gradually as the water content decreases from
x ˆ 2:662 with the maximum value of 730 F/g
appearing at x ˆ 0:847. After the appearance of
anhydrous RuO2 peaks (x ˆ 0:586), the speci®c
capacitance decreases rapidly. This observation is
consistent with the previous measurement [4].
Obviously, the water content of the RuO2 Á xH2O
material is structurally important and plays an
active role in energy charge storage.
lar motion. However, the residual proton±proton
dipolar coupling is still strong enough so that the
resulting linewidths here are broader than that of
the randomly oriented water absorbed in polyi-
mide ®lm [15], implying that the water molecules in
the RuO2 Á xH2O interact weakly with the ruthe-
nium oxide. Furthermore, it can be noticed from
Fig. 4 that the 1H NMR resonance is slightly
shifted by the amount of water content, indicating
that the local structures of the RuO2 Á xH2O may
depend on the water content [11]. Importantly, the
1H linewidth is strongly dependent upon the water
content, as shown in Fig. 5. At the water content
x ˆ 2:662, the 1H linewidth observed at room
Fig. 4 shows the static 1H NMR spectra of
amorphous RuO2 Á xH2O samples at di€erent wa-
ter contents. Di€erent from a rigid sample where
temperature (i.e., 300 K) is 3.4 kHz. The H line-
1
width becomes broader with less water and reaches
its maximum of about 7.8 kHz in the vicinity of
x ˆ 0:7. When the water content becomes very
small (e.g., x ˆ 0:280), where some crystalline
1
the typical H linewidth at the half-height is sev-
eral 10-kHz [14], all spectra here exhibit a rela-
tively narrow resonance (< 8 kHz linewidth at
half-height), again indicating that the structural
water in the RuO2 Á xH2O undergoes fast molecu-
1
particles are formed, the H linewidth is narrow
again. A similar linewidth versus water content
behaviour can also be observed at 260 K (also
shown in Fig. 5). However, the 1H linewidth at 260
K is broader than that at 300 K, which indicates
the lower mobility of the water molecules and the
stronger hydrogen bonding with the ruthenium
oxides at the lower temperature. Such a stronger
interaction indeed results in the higher capaci-
tance, as shown in Fig. 2.
Fig. 5. Linewidths of the 1H NMR spectra of the hydrous
RuO2 Á xH2O samples at di€erent temperatures as a function of
the water content x. All the linewidths were measured at the
half-height of the corresponding maximum peak intensity. The
circles and the squares indicate the 1H linewidths at 300 and 260
K, respectively.
Fig. 4. Static 1H NMR spectra of amorphous RuO2 Á xH2O
samples recorded on a Bruker DMX 300 NMR spectrometer at
room temperature (300 K).

Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
69
An important point to be noted from Figs. 3
and 5 is that the maximum speci®c capacitance
appeared in the vicinity of x ˆ 0:7, where the rel-
atively broader 1H linewidth is observed. Although
there are more protons at x ˆ 2:662 available for
the intercalation, the water molecules are over-
whelmed in the bulk of the material and thus be-
come nearly randomly oriented resulting in narrow
1H NMR linewidth [15]. Such mobile water mol-
ecules exhibit a very limited interaction with the
ruthenium oxide and consequently the speci®c
capacitance becomes low. At a lower water con-
tent (e.g., x ˆ 0:280), on the other hand, the water
molecules are limited to the surface of the crys-
talline particles and thus the surface interaction
areas are e€ectively reduced, resulting in the re-
duction in the speci®c capacitance. In contrast, in
molecules and the lattice so as to increase the
speci®c capacitance [9].
4. Conclusion
We have demonstrated here that solid-state
1H=2D NMR spectroscopy is a very powerful
technique for the characterization of amorphous
RuO2 Á xH2O electrode material with di€erent
water contents. To our knowledge, this is the ®rst
comprehensive NMR investigation of the amor-
phous RuO2 Á xH2O materials, although NMR
spectroscopy has been widely used in ionic con-
ductivity materials [16±18]. The variable-tempera-
ture 2D NMR spectra indicated that the water
molecules undergo fast molecular motion even if
the temperature is as low as 213 K, while the
1H NMR spectra showed that the residual pro-
ton±proton dipolar interaction is composition de-
pendent. The latter implies that the interaction
(i.e., the hydrogen bonding) between the water
molecules and the ruthenium oxides varies with the
water content. Based on the electrochemical mea-
surements, it is concluded from the NMR mea-
surements that the speci®c capacitance (i.e., the
maximum charge density) is associated with the
competition between the two e€ects resulting from
the mobility of the structural water molecules and
their interaction with the ruthenium oxides. The
mobility tends to allow the protons to intercalate
and di€use easily into the bulk of the material,
while the interaction provides a necessary force to
retain the protons with the ruthenium oxides so as
to increase proton charge density. The conclusion
suggests that the speci®c capacitance could be
optimized through the control of the two anti-
thetical e€ects from the mobility and the interac-
tion, either by the temperature or by the
incorporation of impurities. The impurity e€ect is
currently under investigation in our laboratory.
Therefore, this fundamental study describes a
quantitative basis for understanding the mecha-
nism of the proton charge storage of amorphous
hydrous ruthenium oxides, and would provide a
guideline for future development of electrode ma-
terials and also for the formulation of analogous
materials.
1
the vicinity of x ˆ 0:7, the H NMR linewidth is
broader, indicating relatively less proton mobility
and a stronger residual proton±proton dipolar
interaction. In other words, the water molecules
exhibit a relatively stronger interaction with the
ruthenium oxides, probably forming intermolecu-
lar hydrogen bonds with the ruthenium oxides and
thus resulting in the coexistence of the
Ru2‡; Ru3‡; and Ru4‡ valency states, which could
dramatically change the local structure of
RuO2 Á xH2O [11]. Such
a highly disordered
structure allows the protons to di€use easily from
one site to the another in the bulk of the material
while preserving the relatively strong hydrogen
bond between the water molecules and the ruthe-
nium oxides. Thus, the resulting capacitance is
greater in the vicinity of x ˆ 0:7. Therefore, it is
demonstrated once again that such competition
between the two antithetical e€ects resulted from
the mobility of the water molecules' and their in-
teraction with the ruthenium oxides provides a
mechanism for the energy charge storage of the
RuO2 Á xH2O materials. As a result, the speci®c
capacitance can be optimized through the control
of the mobility versus the interaction. In fact, by
lowering the temperature to 260 K the mobility of
the water molecules slows down and their inter-
action with the ruthenium oxides thus becomes
stronger (cf. Fig. 5) leading to the greater capaci-
tance, as shown in Fig. 2. The impurities may also
strengthen the interaction between the water

70
Z. Ma et al. / Chemical Physics Letters 331 (2000) 64±70
[6] J.P. Zheng, T.R. Jow, J. Power Sources 62 (1996) 155.
Acknowledgements
[7] T.R. Jow, J.P. Zheng, J. Electrochem. Soc. 145 (1998) 49.
[8] K. Schmidt-Rohr, H.W. Spiess, Multidimensional solid-
state NMR and polymers, Academic Press, San Diego,
1994.
The authors are indebted to Dr Gang Cao of
the CM/T facility at the NHMFL for allowing us
to access the X-ray di€ractometer. This research
was supported by the Program Enhancement
Grant (PEG) (Project# 550240537) of the Florida
State University Cornerstone Programs (1999) and
the work was largely performed at the National
High Magnetic Field Laboratory supported by the
National Science Foundation Cooperative Agree-
ment DMR-9527035 and the State of Florida.
[9] Y.U. Jeong, A. Manthiram, Electrochem. Solid State Lett.
3 (2000) 205.
[10] T.R. Jow, J.P. Zheng, X. Zhang, Y. Dai, S.G. Greenbaum,
Electrochem. Soc. Proc. 97±24 (1998) 442.
[11] D.A. Mckeown, P.L. Hagans, L.P.L. Carette, A.E. Russell,
K.E. Swider, D.R. Rolison, J. Phys. Chem. B 103 (1999)
4825.
[12] H.W. Spiess, Adv. Polym. Sci. 66 (1985) 24.
[13] O. Isfort, B. Geil, F. Fujara, J. Magn. Reson. 130 (1998)
45.
[14] B.C. Gerstein, C.R. Dybowski, Transient Techniques in
NMR of Solids: An Introduction to Theory and Practice,
Academic Press, London, 1985.
References
[15] S.Z. Li, R.S. Chen, S.G. Greenbaum, J. Polym. Sci.:
Polym. Phys. 33 (1995) 403.
[1] K.E. Swider, C.I. Merzbacher, P.L. Hagans, D.R. Rolison,
Chem. Mater. 9 (1997) 1248.
[16] Y. Wang, J. Sakamoto, C.K. Huang, S. Surampudi, S.G.
Greenbaum, Solid State Ionics 110 (1998) 167.
[17] Y. Dai, Y. Wang, V. Eshkenazi, E. Peled, S.G. Green-
baum, J. Electrochem. Soc. 145 (1998) 1179.
[18] F. Croce, S.D. Brown, S.G. Greenbaum, S.M. Slane, M.
Salomon, Chem. Mater. 5 (1993) 1268.
[2] D.R. Rolison, P.L. Hagans, K.E. Swider, J.W. Long,
Langmuir 15 (1999) 774.
[3] J.P. Zheng, P.J. Cygan, T.R. Jow, J. Electrochem. Soc. 142
(1995) 2699.
[4] J.P. Zheng, T.R. Jow, J. Electrochem. Soc. 142 (1995) L6.
[5] J.P. Zheng, T.R. Jow, Q.X. Jia, X.D. Wu, J. Electrochem.
Soc. 143 (1996) 1068.