Skip to main content

Full text of "USPTO Patents Application 08479810"

See other formats


The resulting map (Fig. le) shows that the absorption feature 
has a mean W of 0.2 nm and stretches from roughly north to 
south across the entire emission line region, corresponding to 
a length of >30 kpc (the map presents only the area with good 
signal-to-noise ratio). Its spatial width is rather uncertain, 
because it is unresolved in the east-west direction (<1.5''). It 
cannot be narrower than 0.5" because otherwise even a 100% 
obscuration would be washed out by our beam into a relative 
depression of <25% (0.25 nm). So we assume a projected size 
of 10x10 kpc- for the absorber, with a deconvoluted equivalent 
width of about 0.5 nm. 

Such an absorber can cither consist of one or more clouds 
located well m front of 4C41.17 (if the blueshift of theabsorption 
is cosmological, the absorber sits at a comoving distance of 
5 Mpc). On the other hand, the velocity in the EELR itself is 
large enough to cover this blueshift; otherwise, we would be 
unable to detect the feature. TTius, a dense, partially ionized 
cloud at the edge of 4C41.17 could equally explain the absorp- 
tion. In this latter case more detailed observations are necessary 
for a physical interpretation. We therefore would like to pursue 
the former possibility of a physically separated absorber Such 
clouds— commonly known as Lyman-forest clouds— and their 
properties have been extensively studied in the absorption line 
spectra of high redshift quasars. 

O00^^.Tr"Ti tT T^"" "'^ ^ spectrum^^ of the quasar 
QOOOO-263 (2 = 4.11). the Lyman-forest of which covers the A 
range of our observation. We smoothed the original spectrum 
(resolution, 0.1 nm) to our instrumental resolution of 1 0 nm 
The comparison between smoothed and original spectrum 
^'^y ^^'sorption feature as deep as that observed in 
4C41. 7 typically consists of two or more narrow absorption 
hnes. We have to realize therefore, that our 'absorption cloud' 
IS hkely to be a superposition of several individual Lyman-forest 
clouds Nevertheless, we believe that the outline of the absorber 
in the W map (Fig. le) is most likely to be determined by one- 
smgle cloud which made the feature strong enough to become 
detectable, and we assign half of the measured equivalent width 
(0-25 nm to this cloud. Assuming a Dopplcr parameter 6 = 
35 km s and AAHyNH,.= 10-^ as typical for Lyman clouds of 
that depth (refs 1, 3). we find a column density N„,-- 10'^ cm'^ 
A cigar-shaped cloud of 40 kpc length and 10 kpc diameter 
would contain a total hydrogen mass of -3 x 10^ Mq 

What is the probability of detecting such an absorption feature 
in front of 4C4L17? Both the smoothed spectrum of QOOOO-263 
(ref. 14) and the standard dN(W,z)/dz relation*' yield --25 
features with H^ >0.4 nm on each line of sight and within one 
. un t at the observed wavelength. Considering the ^usefuP 
wavelength range of --1.2 nm (the blue half of the width of the 
fiT^cir^n features, and the area of 

the EELR inspeaed of ^20 arcsec^ the probability of detecting 
a cloud of a typical size of a few arcscc' is close to 1 

In conclusion, we believe that we have succeeded in obtaining 
he first direct observation of a Lyman absorption cloud. Either 
this cloud belongs directly to the mass concentration around 
4C41.17 or It IS a physically separated foreground object. In the 
Ito'tn T T"*"* rtP^^^^rM the population of Lyman-forest 
clouds known from the absorption spectra of quasars. In either 
case, our observations indicate that the relevant absorbers 

J^a' dT' '''V' ^'-"g-ted shape 

hke^a cigar or a sheet seen almost edge-on in the case of 


9. K in £>xf> or Ga<«, FcnmUan fn^ C S. et m 39-56 (K-uwer. DorCr^ch. 

10. IWy. S. Astrophys. I 333, 161-167 (1988) 

14. Webb. J. K. et al. ESOMssseneer SI, 15-18 {1988) 

15. MurO«^ H S, fi^tea* a W. Pettini. M. & Blades. 1 C Astros. X 309. 19-32 (1S86J, 

Received 23 September 199?: accepted 19 January 1993 


Superconducth^Hy at 94 K 
in HgBa2Cu04,.fi 

S. N. Piitmn*t. E. V. Antipov* O. Chmalssemt 
& M. Mareziott 

Department, Moscow State University 
119899 Moscow. Russia 

t LatXH-atoire de CristallograpWe CNFJS-UF. BP 166 
38042 Grenoble Cedex 09, France 

% AT&T Bell Latioratorles. Murray Kll, New Jersey 07974, USA 

Following the dfacoveTr' of high-transition^eniperatBre (hlgh- 

related compoands have been discovered which have layers of CuO, 
TJ^,^T^ requirement for sape«o«ductivlty: the highest 
t«Mltlon temperatures so far hare been found for thallium- 

Sa RCuT""^- "^'^ '""«="'y;b*«ri->8 compound 
HgBajRCu,0«^, (Hg-lZU) was synthesized* (where R is a rare- 
fy!-- T^^-^'i*" ' '0 thaUium-bearing 

r « «S K (ref. 2). But 

HI ^fte of Its resemblance to Tl-1212, Hg-1212 was found not to 
be superconducting. Here we report the synthesis of the related 
comp^ind HgBa C„0.., (Hg-I201). with only one CuO, tyer 
|«Jr unit ceil, and show that ft is superconducting below 94 K Its 
s ructure » similar to that of TI-1201 (which has a of <10 K)\ 
but us transition temperature is considerably higher. The availabil- 

1°, IT""' "'8," ^' ""'y " «»8'e;etal oxide (HgO) 
ayer may i^e importaot for technological applications, as it seimi 

conducting properties in a magnetic field* 

«„]?h -!'""f ^^'^ P'^P"'*' "^''^ '"«'on between 

A^H i,,"'^'"'^ °^ Ba,CuO,« and yellow HgO (98% 
sam^t'vni rr^-J^' "u^''*"'"' Ba.CuO,.. was obtained by the 
same ype of reaction between BaO, (95% purity. Aldrich) and 
CuO (NormaPur. Prolabo) at 930 °C in oxygen, according to 
the procedure described by De Ueuw « a/A fhe powders were 
ground in an agate mortar and placed in silica tubes. All these 
operations were carried out in a dry box. After evacua^on the 

^nd '^V'"%".' ''''' - described in 

rer. 3, and heated for 5 h to reach -800 «C. The samoles were 
then ^cooled in the furnace, reaching room temSre Tfter 

bv^%^v""'"°w °^ ^^^^ "8Ba,CuO,., was revealed 

by X-ray powder analysis, performed with a Guinier-Haes 
focusing camera and Fe Ka radiation (1.93730 A). Finely po^ 

standard ° ^ '^^"^^ ^ ^'^ ^'^ « i"«rnal 

rtandard. The intensities of the reflections were evaluated by 

Ten w tb l2f '"""""^ ^"^ o« a tetragonal 

i^^nln ? ^y^eraatic absences were observed 

group PA/mmm. The c parameter corresponded to the value 
calculated from the formula c = 9.5 + 3 2(n-n similar m.w 

ooHhfs 2V t^ ^^•^--.C"..0., Jh'Lli^gor S s° W 
took this as a strong indication that the powder pattern corres- 
fK.mied ,0 that of the firs, member of the kB^.^JZoZZ, 

tttrb I AVAlLAbLt COPY 

NATURE • VOL 362 • 18 MARCH 1993 

Betters to nature 

TABLE 1 Crystaltographlc data for HgBaaCuO*.^, 

Positional, therrnal and occupancy parameters 

50 100 150 200 250 

RG. 1 AC magnetic susceptibility x (b) and normalized resistivity (ft) as a 
function of temperature for HgBa2Cu04+4. 

Scanning electron microscopy using a JEOL SM 840A equip- 
ped with an energy-dispersive spectroscopy (EDS) attachment 
revealed that the sample was well crystallized with particle sizes 
of several micrometres. EDS analysis of several well crystallized, 
flat and oriented grains was performed. Beside Hg, Ba, Cu and 
O, no other element was detected in the spectra. The average 
metal ratio found for eight grains was Hg:Ba:Cu = 
28(1) : 47(2) : 25(1), where the numbers between parentheses are 
the standard deviations. Determination of the oxygen content 
by EDS analysis was not possible, so it was estimated by struc- 
tural analysis and iodo metric titration. The cation stoichiometry 
is in qualitatively good agreement with the proposed formula 
of the new compound. 

Alternating-current magnetic susceptibility measurements 
between 4.2 and 120 K, done without any additional oxygen 
treatment, showed that HgBaaCuO^+s samples undergo a transi- 
tion from paramagnetic to diamagnetic with an onset as high 
as 94 K (Fig. 1 a, where the susceptibility is in electromagnetic 
units g~'). The estimated magnetic susceptibility at 4.2 K. corres- 
ponds to >50% of the ideal diamagnetic values. 

The resistivity was measured between 4.2 and 250 K by the 
four-probe technique. The sample was a pressed pellet which 
was annealed in oxygen for 2 h. The temperature dependence 
of the normalized resistivity, shown in Fig. 1, exhibits a sharp 
drop at r^, but the transition is broad and it reaches the value 
of zero resistance only at 35 K. This behaviour indicates that 
the sample is not homogeneous. 

To determine the structure of HgBa2Cu04+8, X-ray powder 
data were collected by a 8/2$ STADI Pdiffractometer in trans- 
mission mode. The experimental conditions were as follows: 26 
range =6-115° (0.02° steps) with fixed counting time 60s and 
a rotating sample. An absorption correction was applied and 
the sample thickness was calculated from the primary beam 
absorption (ftR = 17, where /a is absorption coefficient and R 
is thickness). The structural refinements were done by the Riet- 
veld method. The initial positional parameters were deduced 
from a structural model containing the sequence (Hg)(BaO)- 
(Cu02)(BaO)(Hg). After convergence (intensity discrepancy 
factor, K] = 0.039), a Fourier difference map revealed that the 
position at (^.i,0) of the Hg layer was partially occupied. 
During the final cycle of refinement, the occupancy factor of a 
third oxygen atom placed in this position was varied together 
with the positional and thermal parameters for all atoms (except 
for the thermal parameter of 0(3) which was kept fixed at 
1.0 A^). The final intensity (RJ and profile {Rp) discrepancy 
factors based on 84 reflections were K, = 0.0367 and Kp = 0.1 16, 
with a GOF (goodness of fit) = 0.33. 

The final positional and thermal parameters together with the 
relevant interatomic distances are given in Table 1. Observed, 
calculated and difference diffraction patterns are shown in 
Fig. 2. A schematic representation of the structure is shown in 
Fig. 3. Preliminary structural refinements based on powder 
neutron diffraction data support the presence of oxygen in the 
0(3) position with an occupancy factor slightly larger than that 
found by X-ray powder diffraction data. The neutron data also 

NATURE • VOL 362 • 18 MARCH 1993 


















































Selected Interatomic distances (A) 

Hg-0(2)(x2) 1.95(2) Cu-0(l) {x4) 1.940(1) 
Hg-0(3)* 2.742(1) Cu-0(2) (x2) Z79(2) 

8a-0a)(x4) 2.730(1) 
Ba-0(2)(x4) 2.880(5) 
Ba-0(3)* 2.831(1) 

Data obtained using monochrofnatized CuKa^ radiation (A » 1.54056 A), giving a« 
337 766(4) A. c = 9.5073(1) K 
* Partidiry occupied sites. 

confirm the large value for the mercuiy thermal factors. As in 
the case of the X-ray data, the anisotropic model shows a very 
slight difference between B„ = Bji and Bjj, the thermal factors 
along X, y and z respectively. 

HgBa2Cu04+a has a structure related to that of Hg-I212 
(ref. 3). lu lattice parameters correspond to four-layered packing 
along the c-axis of a unit cell: a^a^^^ , c^la^^.^ (where a^^, 
is the parameter of the perovskite subcell) and its structure 
contains the sequence (Cu02)(BaO)(Hg06)(BaO)(Cu02). The 
Cu cations are octahedrally coordinated, while the coordination 
of the other cations depends upon the value of 6. This, as 
obtained from powder X-ray data, is 0.10(3). An important 
consequence is that most of the Hg cations have two oxygen 
atoms near them in a 'dumb-bell* configuration, an appropriate 
coordination for Hg^* cations. Because 5 is small and different 
from zero (within about three standard deviations) X-ray powder 
data alone are insufficient to determine which sites of the rock- 
salt positions in the HgO layer are occupied and how they affect 
the Hg coordination. The extra oxygen atoms are needed in 
order to increase the average oxidation number of the Cu and 
to create the concentration of holes necessary for superconduc- 
tivity, lodomctric titration performed with a large excess of KI 
leads to 16% of Cu^*, corresponding to 5 = 0.08. 

Similarly, the structure of HgAajRCujOfi+a (the second mem- 
ber of the HgBa2R„_iCu„02„+2+a series) can be described as 
six-layered blocks made of rock-salt and perovs kite-type struc- 
tures. In the structure of Hg-1212 the layer sequence is: 

rock-salt perovskite rock-salt 

The Cu02 monolayer in Hg-1201 has been replaced by the 
(Cu02)(R)(Cu02) block. As a consequence the Cu cations are 

RG. 2 Observed [a), calculated ib) and difference ic) powder patterns after 
Rietveld refinement for Hgpa2Cu04.^5. 




Rece«d 18 Oeamber 1992: Koottd 12 Fehiuar* 199a 

1 Bedhorz. I. G. & Muner. K. A. Z HVS. B64. 189-193 Me5). 
2. I«»n. S. S. P «. ai 1^""^ f^^l ^i, 1399-1307 (1991). 
I Sa^T^V^.r^"^ r«^*"S.l«3-ia6U991, 

I ^.t'^!^"c"-.'a 1':^. c. s^»^ H O... P. 1 

19-24 (19901. 
Cnnd on SupaconductMily. 

FIR 3 Structure of HgBajCuO,«. Tlie large, nriedium and small circles 
LsentT4 Hg^O ato*,;^. respectively. The Cu ator^s are u^* 
.^SUa. N^tE^hat the partially occupied oxygen 0(3) site on the Hg 
layer is represented by a partially filled circle. 

Dvramidally coordinated. The coordination of the Ba and Hg 
rations in Hg-12l2 is similar to that of the same cations in 
He- 1201 The R cations are surrounded by 8 oxygen atoins 
ai^ngcd as a prism. The valence of the Cu cations depends 
upor> the value of 5 and the valence of the R "'•"["j^'fj^^ 
same Cu valence or hole concentration as m Hg-»201 >s needed 
to induce the superconducting state in Hg-12l2, then the R 
cations should be 2+ and should be appreciably greater 
ftan 5„„, . Forthe previously reported Hg-1212. R was a mixture 
of Eu and Ca, and 5 was not precisely determined . It is possible 
that 5 was not large enough to compensate for the higher valence 
of the R cations and to transfer the needed extra charges to 

CuO-> layers. . _ ^ „ 

As" stated above, the structural arrangment of HgBa,CuU«+, 
is similar to that of TlBa^CuOj-a. except for the oxygen 
stoichiometry of the HgO. and T!0,.» layers tcspertively. For 
the former. 5 is very small and this depletion is possible because 
the dumb-bell coordination is appropriate for the Hg " ions. 
For the latter, the TIO,.. layer is only slightly oxygen dep eted. 
creating the appropriate coordination for the thallium cations 
resulting in either a distorted octahedron or a five-coordinated 
DoWhedron. These different requirements for attainmg the 
optimal concentration of holes are due to the different preferred 
coordination geometries of the Tl'* and Hg cations. 

The first member of the latter series (Tl-1201) has been repor- 
ted and found to become superconducting at <10 K {ref. 4) by 
doping the Ba sites with La this value can be increased to 52 K 
(ref 7) The second member of the mono-Tl senes becomes 
superconducting at 85 K (ref. 2). Ttiis increase U a general rule 
for the first few members of this series of compounds. If this 
behaviour holds for the Hg-series, the second member could 
reach values for T, as high as those of the thallium. 

The possible advantages for technical applicauons of 
HgBa,CuO,... in analogy with one-Tl-layer matenals. would 
be due to the relatively short distance between CuOj layers. 
This might lead to lower anisotropy of the superconductmg 
properties and to higher flux-melting temperatures than those 
of two-TlO-layer superconductors . 

Dependence of aggregate 
morphology on structure 
of dimeric surfactants 

R. Zana* & Y. Talmon 

Department of Chemical Engineering, Technion-lsrael Institute of 
Technology. Haifa 32000. Israel 

Surfactant molecnles in water form organized asserabhes of 
rarious shapes, such as miceUes and bllayer lamellae, -hirfi are 
of Interest as analogues of biological structures, as m«>del systenu 
for studying complex phase behaviottr and because of their techno- 
Scal Li^rtance. for example to the food and paint mdus tr.^ 
Tte polar head groups are usually arranged randomly at the 
^a^ of these Leiblies. We have studied the effect on the 
microstrucfure of these assemblies of imposing constraints ou the 
head-group spacing. We investigate the structures formed by 
•double-headed' sarfactents in which two quaternan^ ""rrH 
species (C„H^.,NnCH3)2) are linked at the lerel of the he«d 
-VTups by a hydrocarbon spacer (C.H,.). Here we report the 
microstmctures formed by these dimeric surfactants with m = 12 
and s = 2, 3 or 4 in aqueous solntion, by rapidly cooling the micellar 
solutions and investigating the vitrified structures withtram»m=»'»" 
electron microscopy. The surfactants with a short spacer (» - 2, 3) 
fom. long, thread-like and entangled micelles even at low con- 
centrations, whereas the corresponding monomeric ammonlom sur- 
factants can form only spherical micelles. The dimeric surfactants 
Ilirt " =4 form sphemidal micelles. Thus short spacers (which 
impose reduced head-group separation) appear to promote lower 
spontaneous cnrrature in the assemblies. This approach may afford 
a new way to control amphiphile self-aggr^ation. 

Conventional surfactant niolecules g^=nerally compnse wo 
distinct parts that are incompatible with each other, one polar 
head and either one or two alkyl chains. These mo ecules tend 
to self-associate in water, where they produce micellar solution* 
in the dilute range, and lyotropic mesophases at higher con- 
centrations. Whatever the structure, the surfactant polar head, 
are located at the interface between the hydrocarbon and water 
regions. Their relative positions and distances are <»«e™'"«^ 
mainly by their electrostatic interactions, and also by the pacKing 
requirements of the disordered alkyl chains-\ In caesium or 
rubidium soaps at low temperature in the presence of water to 
example the head groups form well developed he«gonal or 
rectangular crystalline arrays^ Generally, however, they are 
arranged randomly, and little is known of their packinggeometry 
or the width of their spacing distribution. , , , 

To investigate the effect of a perturbation of the local arrange- 
ment of polar heads on the micellar and mesomorphic properties 

Mo OTresponde™ «-U0 1» ««r.s«d In H=i.a. 0« le«^ """^ 
Sadran (CNRSI. 6 rue Ban^-vM. 670S3 StrasBort CeOa. France. 



NATURE • Va 362 • 18 MARCH 1993