Looking back on the early days of scanning tunneling microscopy
(STM), for the development of which Gerd Binnig and I were
awarded the Nobel Prize for Physics in 1986, we find a few
aspects common to many ventures into an unknown world. The
original goal was to learn about the local structural, electronic,
and growth properties of very thin insulating layers, in particular
at tunnel junctions. The term "local" meant on the
scale of the inhomogeneities of these properties, which were
believed to be no larger than a few nanometers in size, a
scale that was entirely inaccessible with existing techniques.
Electron tunneling appeared to be a promising approach, provided
it could be done locally. This led in a natural, non-premediated
way to the local probe method now known as scanning tunneling
microscopy. Electron tunneling already contained two of the
four major technical elements of a local probe method: a strongly
distant-dependent interaction and, inherently necessary, close
proximity of probe and object. One tunneling electrode in
the form of a sharp conducting tip would provide the third
element, the local probe. Metal tips with a radius of curvature
of about 20 nm to bring the resolution to the desired level
were already in use as field emitters and in field ion microscopy.
These three elements determine the resolution. The fourth
and final element was the stable positioning of the probe
with respect to the object with an accuracy better than the
desired resolution and within the practical range of the interaction.
We expected to achieve this in a vibration-protected environment
with piezo drives made from commercially available material.
Although the development of STM appears straightforward in
retrospective, it required nevertheless ideas and efforts;
we had mistakes to correct and, in particular, had to deal
with many unknowns. We were given the time and opportunity
to do both. For instance, replacing the well defined field
emission tip with a ground-metal tip simplified matters and
this could - and did - help achieve atomic resolution since,
unless specially prepared, most tips end up with an apex of
one atom, thus becoming an atomic size probe. However, it
was by no means clear that such a tip would be mechanically
stable. Indeed, it was usually unstable in the early days
of STM; nowadays there are nearly as many recipes for stable
tips as there are scientists using them. The same goes for
the piezo drives. With atomic resolution in sight, the tip
position had to be controlled within a fraction of an angstrom,
not just of a nanometer. Only the experiments themselves showed
afterwards that the response of the piezo to an applied material
was continuous and reasonably linear down to at least the
picometer level. It is remarkable how often success is the
reward for trying the unknown.
In the beginning, the anticipated resolution matched only
that of scanning electron microscopy as far as structural
properties were concerned. But STM offered something else.
A local tunneling experiment, e.g. tunneling spectroscopy,
contains a wealth of information and provides local electronic
and chemical properties, in addition to the structural ones.
This and the conceptual simplicity of the approach were sufficient
incentive, and by the time we had finished our first successful
experiment, resolution was nearly at an atomic level. This
is reminiscent of electron microscopy, which at the beginning
offered lower resolution than optical microscopy, was more
complicated and even destroyed most of the samples in the
imaging process. We might say that what is different constitutes
progress, and not so much what is "better".
STM was not developed from one of the already existing local
probe methods or ideas about them nor within the community
of microscopists nor in other circles with the appropriate
competence. No technologically new component or new material
was necessary, no new physical insight was required nor did
an additional theoretical basis have to be established, yet
somehow the believe prevailed in those communities that "it"
could not be done. Vacuum tunneling apparently crossed many
a mind but was dismissed as infeasible. The topografiner came
closest. Stylus profilometry did not go beyond carefully shaped,
smoothed, and well defined sensing tips with a radius of curvature
of about a micron and, therefore, stayed in the micron resolution
range. Instead, applying a hammer blow to such a tip and taking
a splinter of it, accompanied by a few novel ideas could have
brought about atomic force microscopy (AFM) with nanometer
resolution. In the case of STM, we heard many objections,
including those citing the uncertainty principle, strangely
enough, even after the STM had worked. We might learn from
this that an occasional change in the field of interest can
bring new opportunities for science as well as for scientists.
The initial reservations, if not to say skepticism, with which
scanning tunneling microscopy was met ranged from healthy
scientific caution to a competition-inspired defensiveness,
which has yet to be overcome in some communities. new approaches,
however, should not be regarded as competitors of existing
ones; they complement them or replace them, and each one has
its merits in its own time. Defensiveness does not prevent
progress, at most it merely slows it down.
Nevertheless, what initially appeared to be rather exotic
is now seen largely for what it is, namely a class of new
methods, called local probe methods, for working on the nanometer
scale. By "work" I mean observing a single, individually
selected nano-object, measuring and understanding its properties,
manipulating it, modifying it, and ultimately observing and
controlling its possible functions and related processes.
The local probes are so to speak the "finger tips"
of the nanoworld and are now regarded as a key to the new
coming period, the nanometer age. This period will affect
technology and society no less than did today's microtechnology.
A modest but extraordinary venture can indeed have big, extraordinary
consequences.
The Rise of Local Probe Methods 
