C O MMU N I C A T I O N S
This raises the question of the selection mechanisms that determine
which modification is actually formed. By employing differently
orientated substrates we have ruled out any kind of epitaxial
induction from the substrate surface to be presented. Most probably
the selection occurs in the first step of the crystallization process,
1
7
i.e., nucleation. According to Ostwald’s, or Volmer’s, rule, the
nuclei of lower density polymorphs reach their critical sizes the
quickest at supersaturated conditions. Nucleation-controlled solid-
1
8,19
sate reactions have been studied in detail by Johnson et al.
To
summarize, by dispersing educts for a solid-state reaction at an
atomic level the transport distances can be reduced drastically. By
using such starting mixtures solid-state reactions can be run at
temperatures as amazingly low as room temperature, or even lower.
This opens the door for the synthesis of thermally labile or meta-
stable solids, and can be regarded as the experimental counterpart
to computational approaches to rational design of solid-state syn-
thesis.
Figure 1. X-ray powder diffraction pattern of a Ag/O/N deposit; bottom,
measured directly after deposition at 77 K; top, measured after heating to
9
2
98 K. Line pattern of silver nitrate (HT-polymorph).
Experimental Section. (i) Formation of silver nitrate: Ag
(99,99% Heraeus, Hanau) was evaporated from a molybdenum
crucible by using an electron beam gun at a deposition rate of 0.3
nm/min. The oxygen and nitrogen (purity 5.0; Westfalen, M u¨ nster)
mass flows were controlled to 1.0 sccm O
.3:1). The gas mixture was streamed into the preparation chamber
through a microwave plasma source (tectra, Frankfurt, 2.45 GHz,
0 mA). A sapphire substrate (random, both sides optical polished;
2 2
and 0.3 sccm N (ratio:
3
5
TBL-Kelpin, Neuhausen) was cooled to liquid nitrogen temperature.
The deposition process was run with a process pressure of 4 ×
Figure 2. X-ray powder diffraction pattern of a Ag/O deposit measured at
98 K. Line pattern of silver(I,III) oxide.10
-
5
10
mbar for 5.5 h, and resulted in a layer thickness of ca. 100
2
nm. (ii) Formation of silver(I,III) oxide: Ag (0.45 nm/min) and
oxygen (1.3 sccm) were deposited onto a sapphire substrate (details
see above) at 298 K for 3 h. The obtained dark-colored samples
have a layer thickness of ca. 100 nm. (iii) All X-ray powder
diffraction patterns were measured on a θ/θ D8-advance powder
diffractometer (Bruker AXS, Karlsruhe) with a parallel beam (G o¨ bel
mirror, Cu KR) under vacuum (5 × 10 mbar, incidence angle 2
to 10°, step 0.01°). The program Topas (Bruker AXS) was used
for refinement of lattice constants.
nitrogen is oxidized to the pentavalent state, which otherwise cannot
be directly achieved, is amazing.
However, the most striking detail among the observations is the
unprecedented low thermal activation needed to transform the solid
educt mixtures into well crystalline product phases. The mild
conditions allow for the synthesis of, e.g., a compound as labile as
AgO in an all-solid state reaction, which so far had only been
-
7
accessible through precipitation1 or electrocrystallization
1,12
10,13
from
cooled aqueous solutions. Thus, we feel that by the two studies
presented it has been convincingly demonstrated that, by employing
the setup presented, all-solid state reactions can be performed at
temperatures as low as those usually applied in solution chemistry.
The main reason behind the remarkable readiness by which the
reactions proceed appears to be the extremely finely dispersed educt
mixtures used. By this, the transport lengths have been reduced to
atomic dimensions, which has indeed been the underlying concept
References
(1) DiSalvo, F. J. Science 1990, 247, 649.
(2) Jansen, M. Nordrhein-Westf a¨ lische Akademie der Wissenschaften 1996,
N 420.
(
(
(
(
3) Sch o¨ n, J. C.; Jansen, M. Angew. Chem. 1996, 108, 1358; Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1286.
4) Corey, E. J. Angew. Chem. 1991, 103, 469; Angew. Chem., Int. Ed. Engl.
1991, 30, 455.
5) Nicolaou, K. C.; Jim Li Angew. Chem. 2001, 113, 4394; Angew. Chem.,
Int. Ed. Engl. 2001, 40, 4264.
of our study. This is also the only way to understand the difference
6) Diffusion coefficient of O at 1600 K in MgO is 7.5 × 10-15, in Al
2-
2 3
O
from the results of Johnson and co-workers1
4,15
who, in a similar
-16
2
it is 8.5 × 10
cm /s.
approach, have deposited the educt materials layer by layer.
However, judging by the thicknesses applied (some nanometers),
the boarder to macroscopic dimensions seems to have been reached,
and rather high temperatures (450-800 K) have been required to
achieve the reactions as desired.
(7) Sch o¨ n, J. C.; Jansen, M. Z. Kristallogr. 2001, 216, 307.
(
8) Sch o¨ n, J. C.; Jansen, M. Z. Kristallogr. 2001, 216, 361.
(9) Meyer, P.; Rimsky, A.; Chevalier, R. Acta Crystallogr. Sect. B 1976, 32,
1143.
(
(
10) Jansen, M.; Fischer, P. J. Less-Comm. Met. 1988, 137, 123.
11) Galiba, H.; Csanyi, L. J.; Szabo, Z. G. Z. Anorg. Allg. Chem. 1956, 287,
152.
In addition to the kinetic arguments mentioned, there is a ther-
modynamic aspect that contributes to the enhanced reactivity of
the co-deposits used in our experiments. Since we do not start from
the elements in their states at standard conditions the heat of reaction
is increased by the energies of atomization, which are physically
provided prior to the chemical reaction. This generally raises the
driving force and, in particular, facilitates the access to metastable
compounds.16
(12) Selbin, J.; Usategui, M. J. Nucl. Chem. 1961, 20, 91.
13) Scatturin, V.; Bellon, P. L.; Salkind, A. J. J. Elektrochem. Soc. 1961,
08, 819.
(14) Johnson, D. C. Curr. Opin. Solid. State Mater. Sci. 1998, 3, 159.
(
1
(
15) Hughes, T. A.; Kevan, S. D.; Cox, D. E.; Johnson, D. C. J. Am. Chem.
Soc. 2000, 122, 8910.
(
16) Fischer, D.; Jansen, M. Angew. Chem. 2002, in press.
(17) Hollemann, A. F.; Wiberg, E. In Lehrbuch der Anorganischen Chemie;
81st-90th ed.; Walter de Gruyter: Berlin, 1976; p 309.
(18) Fister, L.; Novet, T.; Grant, C. A.; Johnson, D. C. AdV. Synth. React.
Solids 1994, 2, 155.
(19) Novet, T.; Johnson, D. C.; Fister, L. AdV. Chem. Ser. 1995, 245, 425.
It appears to be a paradox that through a low-temperature syn-
thesis the high-temperature polymorph of AgNO
3
has been obtained.
JA017845A
J. AM. CHEM. SOC.
9
VOL. 124, NO. 14, 2002 3489