вход по аккаунту


Comment on Scanning tunneling microscopy with atomic resolution on ReS2 single crystals grown by vapor phase transport.

код для вставкиСкачать
Ann. Phvsik 2 (1993) 155-759
der Physik
0 Johann Ambrosius Barth 1993
Comment on:
Scanning tunneling microscopy with atomic resolution on ReS, single crystals grown
by vapor phase transport
K. Friemelt, S. Akari, M.-Ch. Lux-Steiner, T. Schill, E. Bucher, and K. Dransfeld
(Universitiit Konstanz)
Ann. Physik 1 (1992) 248-253
D.R. Lloyd
Chemistry Department, Trinity College, Dublin 2, Ireland
Received 13 August 1993, accepted 16 September -1993
The scanning tunneling microscope produces data of the greatest importance, but the
exact significance of the images is still unclear; we do not know “what we are seeing”.
The paper of Friemelt et al. [l] suggests that in the particular case of the layered
chalcogenide ReS2 the atoms “seen” are not those of the uppermost layer of chalcogen, but those of the underlying transition metal atom layer. This casts doubt on the
validity of simple interpretations of images, and, if confirmed, would be very relevant,
for example, to studies of adsorbed molecules on metal surfaces [2].
In the related compounds MoS2 and MoSe2 the top of the valence band has mainly
metal character; the mainly chalcogen-localised bands are about 1 eV more tightly
bound [3,4], so the higher energy of the metal levels could compensate for their greater
distance from the tip, especially since the occupied metal levels are df (do)in character,
with their main lobes perpendicular to the layer plane. Since the translation symmetries
of the metal and chalcogen layers are identical for MoS2- or MoTe2-type compounds,
Friemelt et al. have chosen to study ReS2, where there is a distortion of the metal sublattice. The metal atoms form planar 4-clusters with a rhomboid form, and so there is
a doubling of the metal sub-lattice periodicity. This has also been described as double
chains of metal atoms, which have undergone a Peierls-type distortion, giving chains
of tetramers, and a band gap [ 5 ] . ReSe2 is isostructural [6, 71, so it is of considerable
interest that STM and AFM experiments have also been carried out on ReSe2 [8]. Unfortunately, this work comes to an opposite conclusion from that in Ref. [l], i.e. the
authors conclude that their STM images of the valence band represent chalcogen atoms,
not Re atoms.
Ann. Physik 2 (1993)
Refs. [l] and [8] both use tip biases of about +1 V in air, though they use different
scan modes. They agree that there is a lattice periodicity doubling in the valence band
STM experiments, which is much less clear or absent in AFM images, though the fine
structure of the doubling is not identical. Images in [8] are shown at higher magnification than for [l], and seem to show more detail; in particular, a (1 x 1) periodicity can
be seen, presumably from individual atoms, this is modulated very strongly in two of
the directions of the pseudo-hexagonal surface, but not so strongly in the third, i.e. there
is a row structure present; these rows are of the brightest images, separated by less bright
rows. Ref. [l] also reports a row structure, which has a sharper contrast than that in [8],
but was apparently not always seen. The image in Fig. 3 of [l] shows no evidence of
any features between the bright rows, unlike Fig. 7 of [8], but has a very clear step structure along these rows.
In [l] the lattice doubling is interpreted as evidence that the images are due to emission from the metal tetramers, since the spacings of the chalcogen atoms along the a
and b directions are only very slightly different from that of the ideal hexagonal structure. The stepping along the bright rows is interpreted as being due to the slight zig-zag
of the edges of the tetramers along the a direction, and there is good agreement between
the image and the stepping calculated from the structure. Some images show a double
structure in the spots, which is separated by the length of an edge in the tetramer, and
this is believed to be responsible for the shorter distance, the “tread”, of a step in the
row. It is suggested that the double structures represent pairs of metal orbitals on adjacent atoms, and that, for some reason connected with the details of the band structure,
only two of each set of four metal atoms are detected.
The tetramer, with its planar, rhomboid geometry, is easy to model as an isolated
unit, but there is a difficulty with the interpretation proposed in [l]. The STM emission,
on the model proposed, would probably be from the highest lying of the four occupied
combinations of d t orbitals of the cluster, or from the top of a band made up from
these. In all four combinations of d : wavefunctions of the tetramer there are usually
predominating pairs of atom wave functions. However, the C2 axes of the cluster force
these pairs to lie along diagonals of the rhombus. In contrast, the pair images of [l]
are aligned along edges of the rhombus; this conclusion is inevitable from the step structure referred to above. On point symmetry grounds alone therefore, these stepped images cannot represent d$ combinations. As a check, model calculations have been carried out using Hiickel theory, and p r functions, which in the vacuum half-space have
the same symmetry as d:. Of the two highest lying molecular orbitals of the cluster,
one is essentially the pseudo-diatomic antibonding combination across the short
diagonal, i.e. is almost completely localised on these two atoms, while the other is more
delocalised, but has major components on the long diagonal.
From this argument it seems that the interpretation of the images proposed in [l] is
a little unlikely, though interactions between the tetramers along the chains in the b
direction might produce a modulation of the top of the band in this direction. Further
doubt is cast by a comparison with the interpretations proposed by the authors of [8]
for their images. They note that the chalcogen layers are buckled, so that rows of Se(S)
atoms are alternately high and low with respect to the mean chalcogen plane. These
rows run along b, and because of the alternation of underlying metal sites along the
chains, there are four crystallographicallydistinguishable selenium atoms in each unit
cell. Band structure calculations show that these four atoms have very different densities
of chalcogen pr orbitals in the energy region of a few volts below the valence band
edge IS]. It is shown that the main features of the image can be reproduced very well
Comment from D.R.Lloyd
by a calculated density pattern over the top 0.5 eV of the valence band, and the bright
rows in the image are therefore assigned to Se rows, rather than to metal rows as in [I].
The two interpretations are clearly not compatible with each other. However, while
the most consistent interpretation of the images of [I] and [8] could be that they are
both due mainly to the uppermost chalcogen atoms, and not to the underlying metal
atoms, this is by no means certain. In particular this provides no explanation of the
stepping which is so prominent in some of the images of [I]. There is a possible experimental method of distinguishing between the two interpretations, which, conceptually at least, is remarkably simple. In Ref. [I] the bright rows, because of the clear
stepping, are assigned to the a direction. In contrast, the authors of Ref. [8]assign their
bright rows to rows of Se, which run along b. In neither case is a determination of these
axes reported. Thus the experiment which is needed is a crystallographic determination
of the axes a and b on the STMsampIe. Ideally this would be done for both ReS2 and
ReSez and should allow a clear distinction to be made.
I thank DF.D.A. Morton-Blake for carrying out the calculation, and I thank him, Dr. I.T.McGovern,
and Dr. C. J. Cardin for useful discussions.
[l] K. Friemelt, S. Akari, M.-Ch. Lux-Steiner, T. Schill, E. Bucher, K. Dransfeld, Ann. Physik 1 (1992)
[2] F.P. Netzer, M.G. Ramsey, Critical reviews in solid state and materials sciences 17 (1992) 397
[3] K. Fives, I.T.McGovern, R. McGrath, R. Cimino, G. Hughes, A. McKinley, G. Thornton, J. Phys.:
Condens. Matter 4 (1992) 5639, and references therein
[4] R. Coehoorn. Thesis, University of Groningen 1985
[5] E. Canadell, A. Lebeuze, M.A.E. Khalifa, R. Chevrel, M.-H. Whangbo, J. Am. Chem. SOC.111
(1989) 3778
[6] N. W. Alcock, A. Kjekshus, Acta Chem. Scand. 19 (1965) 79
[7] J.C. Wildervanck, F. Jellinek, J. Less-Common Metals 24 (1971) 73
[ti] B.A. Parkinson, J. Ren, M.-H. Whangbo, J. Am. Chem. SOC.113 (1991) 7833
Без категории
Размер файла
202 Кб
crystals, resolution, vapor, phase, res2, tunneling, transport, microscopy, single, scanning, atomic, comments, grow
Пожаловаться на содержимое документа