A R T I C L E S
Serghiou et al.
46
Re and Zn have metallic radii within 2.6% of each other, they
3
Re Zn D024 structure by creating, for example, new polyhedra
have the same crystal structure (P6
gativities differ by about 15% and their conventional most stable
valencies are +6 or +7 and +2, respectively.
to this, we point at the anomalously large for an hcp metal c/a
ratio (1.87) for Zn at ambient pressure, attributed to a lowering
of band structure energy through lattice distortion, which may
bear on solid solution formation. However, while the anomalous
c/a ratio of Zn, at and above 20 GPa, becomes typical of
hexagonal structures (1.68), and the atomic radii of Re and Zn
3
/mmc), but their electrone-
but literally making the reaction between Re and Zn possible
and hence stabilizing the D024 structure. This then demonstrates
another property of nitrogen in conjunction with pressure, which
opens up further possibilities for engineering new materials.
Even when the formidable pressure variable does not alone lead
to novel architectures and chemistry, addition of a small amount
of nitrogen can make the new structures possible. Further, it
will be very interesting to refine, using ab initio calculations,
how the changes in the electronic structure of Zn, with the
normalization of its c/a ratio upon compression, make it more
compatible with Re, requiring now, at 20 GPa, just an injection
of nitrogen to promote alloying between the two. Finally, this
methodology can be used with other ions as well, including
hydrogen, deuterium, carbon, or oxygen, altogether thus mark-
edly enhancing our options for making targeted, normally
unformable new materials and simultaneously augmenting the
application domain further, for example, to hydrogen storage.
3
9,40
In addition
3
are within 4% at 20 GPa, and both Re and Zn retain the P6 /
8
,9,41-44
mmc structure,
the elements still do not react. This is
the case, even though their melting points are also converging
to within a factor of about 2 of each other at 20 GPa, rather
4
4
than 7 at ambient pressure. Thus, Re-Zn solid solutions at
ambient and at high pressure may not form because of the
differences in their electronegativity and their valency. The
noticeable difference in electronegativity is one factor that may
push the system in the direction of superlattice formation. This,
as discussed above, is because increased differences in elec-
tronegativity make it more favorable for atoms to strive to be
surrounded by unlike atoms, which is an important factor
Methods
The reactants were Re powder (325 mesh, 99.99% Alfa Aesar),
Zn powder (100 mesh, 99.9%, Alfa Aesar), and Zn N powder (99%
3
2
1
6
promoting superlattice rather than solid solution formation.
Alfa Aesar). For the ambient pressure experiments, samples were
heated in a tube furnace with a gas mix of 90% argon/10% hydrogen
flowing through. Four experiments were performed at ambient
pressure, two with mixtures of 33 mol %:66 mol % Re:Zn and
Indeed this scenario is in many respects favorable for superlattice
formation in the Re-Zn system. That is, if size ratios were too
different, then with the pronounced lattice distortions, the solvent
3 2
two with mixtures of 33 mol %:66 mol % Re:Zn N . These were
(
A) would not have been able to incorporate enough solute (B)
placed in boron nitride capsules and heated to 920 K within 10
min. The samples were held there for 5 min and cooled to room
temperature within 20 min. For the high pressure experiment at 5
to form a superlattice in the first place. On the other hand, with
the size ratios being very compatible (as they are for Re and
Zn), accommodation of solute is less of an issue and electronic
differences (electronegativity, valency) are handled by super-
3 2
GPa, an approximate mixture of 33 mol %:66 mol % Re:Zn N ,
was loaded in a capsule made of 0.025 µm thick rhenium foil. The
target pressure of 5 GPa was reached after 2 h. The target
temperature of 1500 K was attained in 20 min. The sample was
held there for 2 min and quenched within minutes. Three high
pressure experiments at 20 GPa were performed with approximate
1
6
lattice formation. We bear in mind that electronegativity and
valency values are nominal, and zinc in alloys, for example,
1
6
has also been assigned a valency of zero. Indeed, from the
atomic packing point of view, nitrogen through electron transfer
and change in the number of valence electrons may adjust the
effective valency of the metallic elements to a nominal
configuration, making superlattice formation favorable. Pressure
can also assist in shifting the balance in favor of the D024
superlattice structure because the D024 structure is only 1.2%
less dense than the stoichiometric component Re and Zn at
ambient conditions.
3 2 3 2
mixtures of 33 at%:66 at% Re:Zn, Re:Zn N , and pure Zn N
4
5
powders loaded in capsules made of 0.025 µm thick rhenium foil.
The target pressure of 20 GPa was reached after 4 h. The target
temperatures of either 1800 or 1900 K were attained in 20 min.
The 33 at%:66 at% Re:Zn and Re:Zn N mixtures were held at the
3 2
target temperature for 2 min and cooled to room temperature with
a cooling rate of 40 K/min within about 50 min, and the Zn N2
3
experiment was quenched within minutes. The samples from both
the ambient and high pressure experiments were recovered and
polished for electron microscopy measurements. The two experi-
ments with Zn N at 20 GPa both gave rise to the Re ZnN phase
3 2 3 x
Conclusions
In this work we have recovered a hexagonal Re
3
ZnN
x
nitride
whereas all the other six experiments gave no reaction between Re
and Zn. Description in great detail of the procedures used for both
the high pressure measurements and the processing procedures after
recovery are described elsewhere. High pressures and temperatures
were applied using a multinanvil apparatus (Hymag, 1000 ton
phase that is stabilized by the combined influence of pressure
and nitrogen. This adds a new dimension to the novel nitride
landscape prepared under extreme conditions of pressure and
temperature. In particular a D024-based nitride has not been
observed before. On the other hand binary A B D024 and related
3
D019 superlattice alloys exist in abundance. Since nitrogen is
situated interstitially, it is actually not altering the backbone
47
3
hydraulic press and a LaCrO heater). The recovered samples were
carbon coated for scanning electron microscopy (SEM) (Philips
XL30CP, with an energy dispersive X-ray analyzer [Oxford
instruments EDX detector - SiLi crystal with PGT spirit analysis
software]) for chemical analysis. The acceleration voltages used
were 10 kV and 20 kV in backscattering electron mode (BSE) for
chemical contrast. The lower voltage served to enhance detection
of the nitrogen light element by reducing the effect of X-ray
(
(
39) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. Chem. Inf. Comput.
Sci. 1996, 36, 746–749.
40) Varga, T. K.; Bello, C. PermaCharts: Periodic Table of the Elements,
Cheektowaga, NY, 2002.
48
absorption. Samples were also investigated with a Philips CM30,
(
(
41) Schulte, O.; Holzapfel, W. B. Phys. ReV. B 1996, 53, 569–580.
42) Steinle-Neumann, G.; Stixrude, L. Phys. ReV. B 2001, 63, 054103-
0
54108.
(46) Xia, S.; Bobev, S. J. Alloys Compd. 2007, 427, 67–72.
(47) Guillaume, C.; Morniroli, J. P.; Frost, D. J.; Serghiou, G. J. Phys.:
Condens. Matter 2006, 18, 8651–8660.
(48) Goldstein, J. I.; Newbury, D. E.; Patrick, E.; Joy, D. C.; Romig, A. D.;
Lyman, C. E.; Fioru, C.; Lifshin, E. Scanning Electron Microscopy
and X-ray Microanalysis, 2nd ed.; Plenum Press: New York, 1992.
(
(
(
43) Duffy, T. S.; Shen, G.; Heinz, D. L.; Shu, J.; Ma, Y.; Mao, H. K.;
Hemley, R. J.; Singh, A. K. Phys. ReV. B 1999, 60, 15063–15073.
44) Young, D. A. Phase Diagrams of the Elements, 1st ed.; University of
California Press: Berkeley, 1991.
45) Brewer, L. Science 1968, 161, 115–122.
1
5174 J. AM. CHEM. SOC. 9 VOL. 131, NO. 42, 2009