The synthesis and characterization of LiFeAs and NaFeAs

Article abstract of DOI:10.1016/j.physc.2009.03.016

The newest homologous series of superconducting Fe-pnictides, LiFeAs (Li111) and NaFeAs (Na111) have been synthesized and investigated. Both crystallize with the layered tetragonal anti-PbFCl-type structure in P4/nmm space group. Polycrystalline samples a

Full text of DOI:10.1016/j.physc.2009.03.016

Physica C 469 (2009) 326–331
Contents lists available at ScienceDirect
Physica C
journal homepage: www.elsevier.com/locate/physc
The synthesis and characterization of LiFeAs and NaFeAs
a
a
d
a
d
a
C.W. Chu a,b,c, , F. Chen , M. Gooch , A.M. Guloy , B. Lorenz , B. Lv , K. Sasmal ,
*
Z.J. Tang d, J.H. Tapp d, Y.Y. Xue a
a Department of Physics and TCSUH, University of Houston, Houston, TX 77204-5002, USA
b Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, CA 94720, USA
c Hong Kong University of Science and Technology, Hong Kong, China
d Department of Chemistry and TCSUH, University of Houston, Houston, TX 77204-5002, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 19 March 2009
The newest homologous series of superconducting Fe-pnictides, LiFeAs (Li111) and NaFeAs (Na111) have
been synthesized and investigated. Both crystallize with the layered tetragonal anti-PbFCl-type structure
in P4/nmm space group. Polycrystalline samples and single crystals of Li111 and Na111 display supercon-
ducting transitions at ꢀ18 K and 12–25 K, respectively. No magnetic order has been found in either com-
pound, although a weak magnetic background is clearly in evidence. The origin of the carriers and the
stoichiometric compositions of Li111 and Na111 were explored.
PACS:
74.25.Fy
74.70.Dd
74.62.Fj
Ó 2009 Elsevier B.V. All rights reserved.
Keywords:
Pnictide superconductors
LiFeAs
NaFeAs
Magnetization
Transport measurements
1. Introduction
gen [3]. These consist of two layered substructures, namely, the
alternating TPn-layers and the RO-layers. Each TPn-layer is com-
Throughout the history of superconductivity, the discovery of a
new material system has often led to a paradigm shift in our
understanding of the phenomenon and has provided new clues
in the search for novel superconductors with a higher transition
temperature (Tc). Previous examples include the A15 superconduc-
tors, the rare-earth chalcogenide ternaries, the rare-earth rhodium
borides, the organic superconductors, the heavy Fermion super-
conductors and the cuprates, among others. The iron–pnictide sys-
tem discovered early last year by Hosono’s group [1] is the most
exciting example since the cuprates.
High temperature superconductors (HTSs) exhibit common fea-
tures, such as a layered structure, atomic substructures, strong
electron–electron interaction, instabilities and magnetic fluctua-
tions, that may play important roles in the occurrence of high tem-
perature superconductivity (HTS) [2]. There exists a large class of
equiatomic quaternary layered compounds ROTPn, that assume
the tetragonal ZrCuSiAs structure type, where R = rare-earth ele-
ment, O = oxygen, T = transition-metal element, and Pn = pnicto-
posed of a T-sheet sandwiched by two Pn-sheets with the T-atoms
tetrahedrally coordinated to four Pn-atoms. Each RO-layer consists
of an O-sheet sandwiched between two R-sheets. TPn-layers may
be expected to form the active block for the charge carriers to flow,
and the RO-layers constitute the charge reservoir blocks that inject
charge carriers into the active block to maintain the maximum
layer-integrity of the TPn-layers, as in the cuprate HTSs. Differ-
ences do exist between the ROTPn and the cuprates, e.g. in the for-
mer, T is tetrahedrally coordinated with four Pn-atoms, whereas in
the latter, Cu-atoms are coordinated by four O atoms to form a
square plane.
Hosono’s group [4] first discovered superconductivity in LaO-
FeP with
a
Tc ꢀ 4 K. By increasing the carrier concentration
through the partial replacement of O by F, its Tc was raised to
ꢀ9 K. Shortly thereafter, LaONiP was found to exhibit a Tc ꢀ 3 K.
However, immense excitement did not arise until the discovery
of superconductivity at 26 K in the F-doped LaOFeAs (La1111),
after the spin density wave (SDW) was suppressed by electron-
doping via the partial substitution of O by F in early 2008 [1].
The relatively high Tc and the presence of a large concentration
of the magnetic Fe, which is antithetic to superconductivity, in
doped La1111 give us new hope in this novel material system to
help achieve higher Tc and to reveal the role of magnetism in HTS.
* Corresponding author. Address: Texas Center for Superconductivity, University
of Houston, 202 Houston Science Center, Houston, TX 77204-5002, USA. Tel.: +1 713
743 8222; fax: +1 713 743 8201.
E-mail address: cwchu@uh.edu (C.W. Chu).
0921-4534/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.physc.2009.03.016

C.W. Chu et al. / Physica C 469 (2009) 326–331
327
Electron-doped La1111 did not disappoint us. In the few weeks
following the report of Honsono et al. in February 2008, the Tc was
raised quickly to 55 K by replacing La with rare-earth elements (R)
of smaller radii in ROFeAs (R1111) [5] or by partial removal of O in
R1111 [6]. Pressure was also found to raise the Tc of La1111 at a
high rate of +1.5 K/GPa [7]. Optimism concerning enhancing Tc re-
turned after the long Tc-stagnation since 1993. Extensive theoreti-
cal and experimental studies have since been carried out [8]. In the
last six months, many layered Fe–pnitides have been found to be
superconducting. By studying these compounds, the less critical
role of O has been demonstrated, the importance of low dimen-
sionality confirmed, the significant role of magnetic fluctuations
in HTS critically reexamined, the electron-hole symmetry demon-
strated, the origin of SDW ascertained, the less critical characteris-
tic of layer-integrity (of the FeAs-layers) and thus the greater
flexibility in doping in Fe–pnictides affirmed, and the absence of
an insulator state shown. The layered Fe–pnictide superconductors
discovered to date can be grouped into four homologous series, i.e.
1111 (ROFeAs [1] and AeFFeAs [9] with Ae = alkaline-earth ele-
ment), 122 (AeFe2As2 [10] and AFe2As2 [11] with A = alkaline me-
tal), 111 (AFeAs) [12–14] and 011 (FeSe) [15]. Among the above
series, AFe2As2 [11] and AFeAs [12–14] become superconducting
apparently without external doping and other undoped nonsuper-
conducting ones can be made superconducting by the application
of pressures [16]. There exist clear differences between the maxi-
mum Tc’s of the different homologous series: ꢀ55 K for 1111,
ꢀ38 K for 122, ꢀ20 K for 111 and ꢀ13 K for 011.
Fig. 1. Tc vs. doping phase diagram of high temperature superconductors
(schematic).
were then synthesized [12] through solid-state reaction of high-
purity Li (ribbons, 99.9%) and FeAs precursors at high tempera-
tures. Stoichiometric amounts of the starting materials were sealed
in welded Nb tubes in an Ar atmosphere. All manipulations were
performed within a purified Ar-filled glove box, with total O2 and
H2O levels <0.1 ppm. The reaction charges were jacketed within
evacuated and sealed quartz containers, heated to 750 °C at a rate
of 1 °C/min and kept for 4 days at 750 °C. The reactant was then
slowly cooled to 150 °C at a rate of 0.05 °C/min and then air-
quenched to room temperature. Polycrystalline samples of Li111
so obtained are black with metallic luster and moderately sensitive
to moist air. Single crystals of size of a few tenths of a millimeter
were obtained by prolonged annealing at low temperature inside
the sealed Nb tube filled with purified Ar. Elemental analyses on
micro-single crystals and polycrystalline samples using an induc-
tively coupled plasma/mass spectrometer (ICPM) gave a Li:Fe:As
ratio of 1.00(3):1.03(2):1.00(1), representing a stoichiometric com-
position of LiFeAs. Phase purity and the lattice parameters of the
polycrystalline samples were investigated by using a Panalytical
X’pert diffractometer. The X-ray powder diffraction displayed no
detectable impurity phases. The structure of Li111 was determined
by XRD on a single crystal with dimensions of ꢀ0.28 Â 0.14 Â
0.02 mm3 mounted in a glass fiber under a stream of cold nitrogen
gas at À58 °C, using a Siemens SMART diffractometer. Li111 crys-
tallizes in a tetragonal structure (P4/nmm) of the anti-PbFCl-type
with a = 3.7914(7) Å and c = 6.364(2) Å, as shown in Fig. 2. Li111
features Fe2As2 layers based on edge-sharing FeAs4-tetrahedra or
derived from As capping of the Fe-square nets, above and below
each center of the Fe-squares. The Fe–As bond distance within
the layer is 2.4204(4) Å; the nearest Fe–Fe distance is 2.6809(4) Å.
Various physical properties of polycrystalline and single crystal
Li111 samples were determined with different techniques. They
While many of the questions raised [17] concerning the Fe–
pnictide superconductors soon after their discovery have been an-
swered, some remain unresolved. For example, the following three
have been asked: (1) How high can the Tc of 1111 (or the known
Fe–pnictides, more appropriately now) be raised? (2) How can
superconductivity be induced in FeAs-layered compounds without
external doping? (3) Is SDW a prerequisite for Fe–pnictides to be-
come superconducting? Among the four homologous series, 111 is
the newest, only discovered very recently. We therefore would like
to present our results on Li111 and Na111 and discuss whether
these two members can offer to help answer some questions about
Fe–pnictide superconductors, although some preliminary results of
Li111 have appeared elsewhere.
2. Experimental and discussion
To address whether the Tc of the 1111 series can be raised to be-
yond 26 K, soon after its discovery we examined the pressure influ-
ence on the Tc of electron-doped Sm1111 (SmO1ÀxFxFeAs) at
different x’s [18]. We found that pressure enhances the Tc of sam-
ples in the underdoped region where dTc/dx > 0 and suppresses the
Tc of samples in the over-doped region where dTc/dx < 0, similar to
the holed–doped cuprate HTSs (Fig. 1). Following the universal Tc–x
relationship proposed by Presland et al. [19] for the cuprates, we
therefore concluded that the Tc of the optimized R1111 compounds
should be in the 50’s K and independent of R. Therefore, we have
decided to search for new FeAs-layered compounds of different
structures to see whether Tc can be raised to above 50’s K. After
finding superconducting K122 and Cs122 with Tc = 3.7 K and
2.4 K [11], respectively, we succeeded in synthesizing and charac-
terizing Li111 [12] and Na111 and found them superconducting
without external doping with Tc = 18 K and 12–25 K, respectively.
Details will be discussed below:
include the low frequency (19 Hz) resistivity (q) of the polycrystal-
line samples using the standard four-lead technique with the
LR700 resistance bridge; the dc magnetization (M) and the dc mag-
netic susceptibility (v) using the Quantum Design SQUID magne-
tometer; the ac magnetic susceptibility (
vac) at 19 Hz by the
inductance technique using the Linear Research LR700 Inductance
Bridge; the specific heat (Cp) employing the relaxation technique
with the Quantum Design PPMS; and the thermoelectric power
(S) employing a high precision low frequency technique developed
by us. The hydrostatic pressure environment up 1.8 GPa was mea-
sured inductively using the clamp technique with the non-mag-
2.1. LiFeAs (Li111)
Pure FeAs precursors were first synthesized from the reaction of
high-purity Fe (pieces 99.999%) and As (lumps, 99.999%) in sealed
quartz containers at 600–800 °C. Polycrystalline LiFeAs samples

328
C.W. Chu et al. / Physica C 469 (2009) 326–331
insensitive to the interface resistance of grain boundaries, indeed
confirms a superconducting transition to zero at ꢀ18 K. It also
gives a negative value at room temperature, indicating an average
electron-behavior of the carriers. Upon application of a magnetic
field, the transition is shifted downward as expected. By taking
the temperatures where
q drops by 5% of the transition as Tc at dif-
ferent field, an upper critical field at zero temperature Hc2(0) > 80 T
is obtained according to the Ginzburg–Landau formula of
Hc2(t) = Hc2(0)[(1 À t2)/(1 + t2)], where t = T/Tc. The Cp(T) in ambient
and 7 T fields is shown in Fig. 4. A clear anomaly appears at ꢀ16 K,
indicating an unambiguous bulk superconductor. Under 7 T, the
anomaly is broadened, reduced and pushed to lower temperatures.
Analysis of the data between 10–20 K gives a Sommerfeld coeffi-
cient
The pressure effect on
c
= 22.5 mJ/molK2.
vac(T) of Li111 is shown in Fig. 5. The
superconducting transition is clearly shifted downward in parallel
by pressure at a rate of À1.56 K/GPa and the pressure-dependence
of the onset Tc is summarized in the same figure.
Prior to the discovery of superconductivity in Li111, LDF calcu-
lations with virtual crystal approximation were carried out and it
was shown by Singh [20] that the compound possesses a small Fer-
mi surface with hole-cylinders at the zone-center and electron-cyl-
inders at the zone corners, as well as high density of states N(E)
below the Fermi surface EF, similar to La1111. He predicted that
Li111 should have an antiferromagnetic ground state and that
N(EF), which comes mainly from the Fe 3d electrons, is insensitive
to doping or Li-content. Indeed, our XRD discussed above shows
that Li111 has a structure closely related to the undoped La1111.
Without doping, Li111 has Fe+2-ions like La1111, which is not
superconducting and undergoes a SDW transition at 135 K as evi-
Fig. 2. Crystal structure of LiFeAs.
denced by a large
q drop [1]. In contrast to the above predictions,
we found Li111 to be superconducting without evidence of a mag-
netic transition. This was also observed by other independent
groups at about the same time [12–14]. Mössbauer measurements
of Li111 have not found any indication of magnetic order at low
temperatures [21]. According to later LDA calculations by Nebrasov
et al. [22], one may attribute the absence of SDW transition or the
weaker magnetic effect in Li111 to its smaller N(EF) resulting from
the tighter As-tetrahedral coordination. Jishi and Alyahhyei [23]
recently calculated the electron–phonon coupling constant of
Li111 and found it to be too small to account for the relatively high
Tc of the compound. Both the XRD and the powder neutron diffrac-
tion (PND) data showed that Li111 should have the stoichiometric
composition ratio of 1:1:1. Its presence of superconductivity and
absence of SDW without external pressure or doping remain puz-
zling based on the valence count of Li111. Previous investigation
netic Be–Cu high pressure cell, where the pressure was measured
by a superconducting Pb-manometer placed next to the sample.
Typical temperature dependences of
q(T) and
v(T) of polycrys-
talline Li111 samples are shown in Fig. 3. The
q(T) decreases with
temperature decrease as a metal but with a large negative curva-
ture, indicating the presence of strong electron–electron interac-
tions in the compound. It drops rapidly to close to zero,
indicative of a superconducting transition, with a Tc ꢀ 18 K. This
is further evidenced by the appearance of a diamagnetic shift in
v
(T) at ꢀ18 K, corresponding to ꢀ100% of a bulk superconductor.
The small but non-zero residual is attributed to the nonconduct-
q
ing grain surfaces resulting from the chemical sensitivity of the
polycrystalline sample to moist air. S(T), which is known to be
Fig. 3. Resistivity
q(T) and volume susceptibility
v
(T) (inset) of LiFeAs.
Fig. 4. Heat capacity Cp(T) of LiFeAs at zero magnetic field and at H = 7 T.

C.W. Chu et al. / Physica C 469 (2009) 326–331
329
tural with Li111, crystallizing in the anti-PbFCl-type as shown in
Fig. 6. This is in agreement with a previous report [13]. The
Na111 structure also features Fe2As2 layers based on edge-sharing
FeAs4-tetrahedra. The Fe–As bond distance within the layer is
2.457(1) Å; the nearest Fe–Fe distance is 2.819(1) Å. The critical
As–Fe–As bond angles are 108.45(2)° and 109.99(2)°, comparable
to the As–Fe–As angles of 103.11(2)° and 112.74(1)° in Li111.
The
q(T) is shown in Fig. 7 for Na111 with nominal x = 0.9 (for
reasons to be evident later). On cooling from 300 K,
q(T) decreases
almost linearly with temperature decrease, deviates from this lin-
earity below ꢀ100 K, decreases more rapidly below ꢀ50 K and fi-
nally drops to zero at 12 K, slightly higher than the Tc previously
reported [13]. The transition appears to be rather broad. When a
field up to 7 T is applied, the overall transition is shifted progres-
sively almost in parallel to lower temperature and the temperature
where q(T) becomes zero decreases as expected of a superconduc-
ting transition (Fig. 7). The onset Tc is rather difficult to define from
Fig. 7, although the data gives a value much higher than 12 K.
The
v(T) of a single crystal of Na111 shown in Fig. 8 exhibits a
Fig. 5. Tc vs. pressure of LiFeAs. Inset: vac vs. T at different pressures.
broad superconducting transition with a Tc ꢀ 18 K and a large hys-
teresis, implying a large flux pinning force and possibly some mag-
netic background. Whether this magnetic background is intrinsic
to the sample is not clear at this time. It is interesting to note, how-
by Juza and Langer [24] showed that LiFeAs could form only in its
Li-rich and Fe-deficient forms. If this is true, the dilemma concern-
ing superconductivity and magnetism can then be resolved. Such a
scenario may not be impossible, given the uncertainty of XRD and
PND in determining light elements like Li. The negative S detected
at room temperature indicates that the carriers in Li111 behave as
electrons on the average, suggesting that the sample we have
investigated may be Li-rich.
ever, that the field-cooled and zero-field-cooled v(T)’s converge on
warming at ꢀ25 K, representing a higher onset Tc near this temper-
ature. The signal size of the zero-field-cooled diamagnetic shift at
4 K gives a superconducting volume <10% of a bulk superconduc-
tor. Unfortunately, other single crystals of Na111 exhibit only a
much smaller superconducting signal, if at all. Thus, we decided
It has been demonstrated that for hole-doped cuprates and Fe–
pnictides, dTc/dP is positive in the underdoped region where dTc/dx
is positive, while dTc/dP is negative in the over-doped region where
dTc/dx is negative (Fig. 1). This suggests that pressure produces the
same effect as hole-addition or electron-reduction [17]. The ob-
served negative dTc/dP in Li111 implies that the compound may
be in the hole over-doped region or the electron underdoped re-
gion, consistent with the Li-rich state of the sample suggested.
to examine the Na content (x)-dependence of the
v(T) of NaxFeAs.
The magnetization of the polycrystalline samples with
0.5 6 x 6 1.0 sealed either in thin quartz tubes or vacuum-grease
coated gel-caps was measured at 10 Oe. Small but noticeable mag-
netic background appears, which has the lowest value of
4
p
v
ꢁ 2 Â 10À3 for x P 0.9 and rises up to 0.1 for smaller x or after
aging, including exposure to air. This background, fortunately, is
only weakly dependent on the temperature between 20 and
50 K. Therefore, 4
the transition. The zero-field-cooled susceptibility
ZFC,30 K) and the field-cooled one 4 FC,30 K) appear
p
(
v
À
v30 K) is used to extract information about
2.2. NaFeAs (Na111)
4
p
(vZFC
À
v
p
(
vFC v
À
Polycrystalline and single crystal NaFeAs samples with nominal
composition NaxFeAs with 0.5 6 x 6 1 were synthesized by the so-
lid-state reaction technique employed for Li111 [12], but with a
small (5%) excess of Na. The Na111 polycrystalline samples are
black with metallic luster. Post annealing around 300–400 °C, both
inside the Ar-filled Nb tubes and under continuous vacuum, were
also carried out for some synthesized samples. Plate-like single
crystals with dimensions of a few tenths of a millimeter were also
obtained after prolonged heat treatment at low temperature. All
Na111 samples so obtained are extremely sensitive to air. Careful
steps in handling the samples were taken to minimize the sample
exposure to air. Both powder and single-crystal XRD analyses were
to converge at temperature up to 40 K for the x = 0.5 sample, sug-
performed, and the resistivity (
q), the dc magnetization (M), the dc
magnetic susceptibility ( ) and the thermoelectric power were
v
measured, by employing the same techniques used for Li111, de-
scribed above.
The powder XRD data show that samples with nominal compo-
sitions NaxFeAs, x = 1.0 and 0.9 are single-phase and all diffraction
peaks can be indexed to a tetragonal unit cell with the P4/nmm
space group. X-ray diffraction refinements were performed on data
obtained from
a
single crystal with dimensions of
ꢀ0.08 Â 0.06 Â 0.04 mm mounted in a glass fiber under a stream
of cold nitrogen gas at À57 °C. The XRD data were collected using
a Siemens SMART diffractometer, using the Mo Ka1 radiation. Final
cell parameters are: a = 3.9866(16) Å, c = 7.094(4) Å, V = 112.74
(9) Å3, Z = 2. Structure refinements indicate that Na111 is isotruc-
Fig. 6. X-ray diffraction spectrum of a NaFeAs (x = 1.0) polycrystalline sample.

330
C.W. Chu et al. / Physica C 469 (2009) 326–331
a
b
Fig. 7. Resistance R(T) of a Na0.9FeAs sample in zero field. Inset:
q(T) at different
fields. From left to right: 7, 5, 3, 1 and 0 T.
c
Fig. 9. Diamagnetic shifts vs. T for NaxFeAs with (a) x = 1.0, (b) x = 0.9, and (c)
x = 0.5, respectively. The open and filled symbols are the field-cooled and zero-field-
cooled susceptibilities, respectively.
of XRD (Fig. 6). The a-axis shrinks and the c-axis remains constant
with a decrease in x. For samples with x < 0.9, noticeable impurity
phases appear while the majority phase retains the Na111 struc-
ture and the lattice parameters of the majority phase vary negligi-
bly with x. By examining the amount of phases present in the
samples and by balancing the masses of the constituents, the ac-
tual Na-value of the Na111-phase in the samples with x < 0.9 is
close to 0.86 for the x = 0.8 and 0.7 samples. For the x = 0.5 sample,
the presence of a large amount of impurity phases make the anal-
ysis less reliable.
As described above, magnetic measurements carried out at 10
Oe show small but noticeable magnetic background which has
Fig. 8. Volume susceptibility
and circles: MZFC
v(T) of a NaFeAs (x = 1.0) single crystal. Triangles: MFC
the lowest value of 4
pv
ꢁ 2.10À3 at y P 0.9 and rises up to 0.1 as
.
x decreases or after air exposure. This background, fortunately, de-
pends only weakly on temperature between 20 and 50 K. The al-
most same 4
p
(
vFC
À
v
FC,30 K) and 4
p
(
vZFC
À ZFC,30 K) at x ꢀ 0.9
v
gesting a possible onset Tc at this temperature, consistent with the
suggestion from the (T) results in Fig. 7. As shown in Fig. 9, the
diamagnetic transition is rather broad, and the exact onset Tc re-
mains difficult to determine. The magnitude of the diamagnetic
shift depends on x and demonstrates a 100% superconducting vol-
ume fraction only for the x = 0.9 sample (Fig. 9b). The rather small
indicate that samples of this x have relatively weak pinning. The
isothermal M–H loops around 4 K ꢂ Tc support the interpretation
that the flux pinning is rather weak over the x-range so the de-
q
duced 4
p
(
v v30 K) in the field-cooling can be taken as a measure
À
of the relative superconducting volume fraction in the samples.
They are shown in Fig. 9. The sharpness of the x-dependence of
4
p
(
vZFC vZFC,30 K) at the nominal x = 1.0, on the other hand, indi-
À
4
p
(
vFC vFC,30 K) on x and its large value at x = 0.9 suggests that
À
cates that bulk superconductivity may not occur there (Fig. 9a).
In order to determine the exact x for the superconducting
Na111-phase, we analyze the XRD and magnetization data and
determine the x-dependences of Tc and the superconducting vol-
ume fraction of Na111 of different x’s. Detailed analyses will be
published elsewhere. Here we shall report the preliminary results.
Samples with x = 1.0 and 0.9 are pure phase of Na111 with the
tetragonal structure of the anti-PbFCl-type, within the resolution
bulk superconductivity takes place in Na-deficient Na111-phase
with a narrow x-range.
To understand the observed superconductivity in the stoichi-
ometric Na111, we further examined the ‘‘aging” effect on these
samples. An aging effect on the samples with x = 1.0 is clearly evi-
dent in Fig. 10. Both the Tc and the superconducting signal grow
when the sample is aged, which occurs even in a vacuum-grease

C.W. Chu et al. / Physica C 469 (2009) 326–331
331
ered tetragonal anti-PbFCl-type structure with a space group of P4/
nmm. The lattice parameters (a,c) are (3.7914 Å, 6.364 Å) for Li111
and (3.9866 Å, 7.094 Å) for Na111, respectively. In contrast to con-
ventional wisdom, they are found to be superconducting with a
Tc ꢀ 18 K for Li111 and 12–18 K for Na111 without external dop-
ing, although superconductivity occurs in Na-deficient samples
over a very narrow x-range close to 0.9. The onset Tc of Na111
may be even as high as 25 K. Carriers in both compounds exhibit
an electron-behavior.
The compounds of Li111 and Na111 add a new homologous
111 series to the superconducting Fe–pnictide system. Together
with the previously known 1111, 122, and 011 series, 111 may
provide additional insights to the role of magnetism in the
occurrence of HTS, given its lack of it, at least in terms of mag-
netic ordering. The four homologous series do offer basic ideas
concerning the formation of compounds with more complex
structures for higher Tc. Unfortunately, the similarities between
them overwhelm the difference. Given the large number of lay-
ered transition-metal–pnictides, the system as a whole still holds
promise for higher Tc. As to the two new members of the 111
series, two important issues warrant further careful investiga-
tion, i.e. the exact compositional ranges and the origin of the
magnetic background in these compounds, although preliminary
investigations have been done.
Fig. 10.
v
vs. T for an aged NaFeAs (x = 1.0) sample. The top and bottom figures are
respectively. Circles:
for the zero-field-cooled and field-cooled values of
immediately measured; triangles: 8 h later; and squares: 66 h later.
v,
Acknowledgments
This work is supported in part by the T.L.L. Temple Foundation,
the John J. and Rebecca Moores Endowment, the State of Texas
through TCSUH, the US Air Force Office of Scientific Research,
and the LBNL through the US Department of Energy. A.M.G. and
B.L. acknowledge the support from the NSF (CHE-0616805) and
the Robert A. Welch Foundation.
References
[1] Y. Kamihara, T. Watanabe, M. Hirano, H. Hosono, J. Am. Chem. Soc. 130 (2008)
3296.
[2] C.W. Chu, AAPPS 18 (2008) 9.
[3] P. Quebe et al., J. Alloys Compd. 302 (2000) 70.
[4] Y. Kamihara, H. Hiramatsu, M. Hirano, R. Kawamura, Hi. Yanagi, T. Kamiya, H.
Hosono, J. Am. Chem. Soc. 128 (2006) 10012;
T. Watanabe et al., Inorg. Chem. 46 (2007) 7719.
[5] X.H. Chen et al., Nature 453 (2008) 761;
Z.A. Ren et al., Europhys. Lett. 82 (2008) 57002;
G.F. Chen et al., Phys. Rev. Lett. 100 (2008) 247002.
[6] Y. Yang et al., Supercond. Sci. Tech. 21 (2008) 082001;
Z.A. Ren, Europhys. Lett. 83 (2008) 17002.
[7] W. Lu et al., New J. Phys. 10 (2008) 063026.
[8] See for example: R.A. Jishi, H.K. Alyahyaei, arxiv:/0812.1215v, and references
therein.
[9] M. Tegel et al., Europhys. Lett. 84 (2008) 67007.
[10] M. Rotter et al., Phys. Rev. Lett. 101 (2008) 107006.
[11] K. Sasmal et al., Phys. Rev. Lett. 101 (2008) 107007.
[12] J.H. Tapp et al., Phys. Rev. B 78 (2008) 060505(R);
M.J. Pitcher et al., Chem. Comm. 2008 (2008) 5918.
[13] D.R. Parker et al., arxiv:/0810.3214, 2008.
[14] X.C. Wang et al., Solid State Commun. 148 (2008) 538.
[15] F.C. Hsu et al., Proc. Natl. Acad. Sci., U.S.A. 105 (2008) 14262.
[16] P. Alireza et al., J. Phys.: Condens. Matter 21 (2009) 012208;
M.S. Torikachvili et al., Phys. Rev. Lett. 101 (2008) 057006;
H. Okada et al., J. Phys. Soc. Jpn. 77 (2008) 113712.
[17] See for example: C.W. Chu et al., J. Phys. Soc. Jpn. 77 (2008) C-72.
[18] B. Lorenz et al., Phys. Rev. B 78 (2008) 012505.
[19] M.R. Presland et al., Physica C176 (1991) 95.
[20] D.J. Singh, Phys. Rev. B 78 (2008) 094511.
Fig. 11. Thermoelectric power S(T) of a NaFeAs (x = 1.0) sample.
sealed gel-cup. The effect can be understood in terms of the loss of
Na when it reacts with whatever small amount of O2 or H2O in the
sample environment. All these provide another piece of evidence
that superconductivity occurs in Na-deficient Na111. However,
this is not consistent with the electron-doping characteristic of
the carriers of the superconducting samples as indicated by the
negative S(T) shown in Fig. 11.
3. Conclusion
[21] I. Felner, private communication.
[22] I.A. Nekrasov et al., JETP Lett. 88 (2008) 543.
[23] R.A. Jishi, H.M. Alyahyaei, arxiv:/0812.1215, 2008.
[24] R. Juza, K. Langer, Z. Anorg. Allg. Chem. 361 (1968) 58.
We have synthesized and characterized both polycrystalline
and small single-crystal Li111 and Na111. Both crystallize in a lay-