1.19.
The need for reliable data was recognised very early in the developmental science. Fig. 1.20
shows how the experiments have helped in establishing of theoretical postulates, and the experi-
ments were possible only with the developments in precision measurements. The logical consequen-
ces of preconceived views could be demonstrated by carrying out experiments. Thus precision
measurements form the background of advanced technology engineering.
Advancements in Reference Standards
Quantitative measurements require some sort of stable measurement standard if the results
of measurements in different places and at different times are to be usefully compared. Intimately
connected with the definition of units of measurements is the practical task of the physical
realization of these units, since a paper definition serves no useful purpose if it cannot be realized
with the required precision.
It was only since science reached the state where engineering began to rely on any significant
extent upon the discoveries of physics, that accurate standards of measurement became necessary.
Towards the middle of the nineteenth century the need for word-wide agreement on the principal
units of measurement was felt and this led to the signing in 1875 of the Convention. The founding
of the major national metrological laboratories took place soon after. In 1875 the only standards
deposited were those of mass and length. The latter very quickly required the establishment of
Fig. 1.20. The background of Advanced Technology Engineering.
accurate standards of temperature, since thermometers were needed to measure the thermal
expansion of the meter bars.
During the second half of the nineteenth century some of the leading physicists devoted
considerable time and effort to the problem of accurate measurement and the establishment of a
consistent system of units. James Clerk Maxwell in 1870 stated,: “If we wish to obtain standards
of length, time, and mass which shall be absolutely permanent, we must seek them not in the
dimensions or the motion of the mass of our planet, but in the wavelength, the period of vibration,
and the absolute mass of these imperishable unalterable, and perfectly similar molecules.” Indeed,
as physics progressed it became possible to envisage practical units based upon atomic phenomena.
The establishment of useful units based upon atomic and, in due course, quantum
phenomena, has not turned out to be easy. In order to establish an atomic or quantum-based base
unit at least three conditions must be satisfied. First, one requires a good theoretical understanding
of the quantitative behaviour of the atomic or quantum phenomenon in question and, second, a
practical way of realizing the phenomenon must exist that allows an explicit theoretical description
to be written down of the process employed, and third, a method is required of making a sufficiently
accurate link between the other base units and the proposed atomic or quantum-based unit. It is
often the third of these that turns out to be the most difficult. Nevertheless, it is in the attempt to
use the fundamental physical constants as the bases of measurement standards that the role of
precise measurement in physics became evident. In order to check that we have an adequate
theoretical understanding of atomic or quantum phenomena we need to be able to check quantitative
predictions made by theory. This can sometimes require the most accurate and precise measure-
ments possible, to distinguish between competing theories or interpretations of theory.
The bases of measurement standards, the fundamental physical constants are useful only
in so far as physical phenomena exist that can be described explicitly in terms of the fundamental
constants, and which can at the same time, be measured in terms of the macroscopic base units.
The extent to which this can be done is an indication of the extent to which modern atomic and
quantum physics can be used to produce much of today’s advanced technology. The practical
application of atomic and quantum physics depends upon our ability to make the precise measure-
ments that link these microscopic phenomena to the macroscopic world. An essential part of this is
the accurate measurement of the fundamental physical constants in terms of the base units.
Quantum Physics and Precision Measurements
Almost all high accurate measurements are dependent upon quantum mechanical properties
of system. Further there is no fundamental principle that forbids accurate classical measurements.
The application of advanced technology to classical measurements can lead to significant
improvements in accuracy. The gas constant, which together with the gravitational constant has
in the past always lagged far behind the other physical constants in the accuracy with which it is
known, has recently been measured using a spherical acoustic resonator with an accuracy approach-
ing 1 part in 106, and still further improvements seem possible. Some weighings are carried out as
part of an experiment designed to search for a fifth force, that have exhibited standard deviations
of about 5 parts in 1012 for series of measurements lasting a few days.
Such measurements outside the very limited immediate field of application are of interest
because one finds that when an atomic or quantum phenomenon is used as the basis of a particular
technology, there appears a demand for all aspects of that technology to increase in precision.
1.19.1. Nanometrology.
In order to connect monomode optical fibres together it is necessary
to have mechanical connectors that can mate with axial errors of fractions of a micrometer. Such
dimensional tolerances are characteristic of the demands now being placed upon length or dimen-
sion metrology by the optoelectronics industries.
The increase in the density of packing of circuit elements on silicon wafers has resulted in
the dimensions of individual elements falling to tens of nanometers. In order to make such devices
it is now necessary to be able to identity and reproduce the position that objects should have on a
silicon wafer to this accuracy. Thus on a part of a silicon wafer roughly 10 cm square it is necessary
to be able to locate and identify any point with a precision of 10 or 12 nm. The requirement in any
particular production line may indeed be only for relative precision, but as soon as the process is
transferred elsewhere, either to another factory or even to another part of the same factory, the
position of a point on the wafer must be expressed in some unit. While in the short term a wholly
local unit can be maintained, based upon the properties of a particular machine, in the long term
it is necessary to measure in meters or nanometers. National laboratories are now finding that it
is by no means a simple exercise to measure and identify the position of points on a 10 cm square
with an accuracy of 10 nm or better. This is the level of accuracy and precision required of those
national laboratories that are concerned with the testing of the commercial master grids that are
used in the fabrication of masks for large-scale manufacture of integrated circuits.
Another example of extreme precision in manufacturing for the optoelectronic industries is
the production of small aspheric lenses whose performance is diffraction limited, but whose
dimensions are small. The production of the master mould for such lenses is possible with
high-precision machining. Diamond machining, elastic-emission machining, and other processes
now can produce surfaces smooth at the nanometer level. Intimately linked to the manufacture of
objects having such properties is the measurement and testing of them. New optical devices now
exist to measure surface profiles and at least one of them can observe large-scale surface structure
at the level of tens of picometers.
The need for rapid inspection of complex objects manufactured to high precision by the new
generations of numerically controlled machines has led to the development of coordinate-measuring
machines. These machines are designed to measure the spatial coordinates of any point within a
volume of up to 1 m3 with an accuracy of about 1 um. The calibration of such machines and the
design of test objects for them is now a major activity of many national laboratories. The new
“nanometrology” will require the development of new, more accurate techniques for surface profile,
roundness, and diameter measurement. In these fatter domains, existing optical or mechanical-
contact methods for identifying the position of a surface are unlikely to be adequate. In order to
identify the position of the surface of an object at the nanometer level it will be necessary first to
define more carefully what it is meant by a surface and second to apply new methods, perhaps based
upon the new tunnelling devices, to indicate contact.
Even if we can define what we mean by a surface and have devices that can detect its
proximity, it is still necessary to be able to servo-control tools, work pieces and different parts of
the device to maintain the required relative positions of all the component parts. The existing
displacement detectors will have to be pushed to their limit. Capacitance devices are, of course,
sensitive detectors of position but are difficult to make linear over a long range, and greater care
is needed to avoid forces being applied to the object being displaced. Silicon cell detectors are not
very sensitive at the level of fractions of a nanometer and, in any case, it can be difficult to supply
enough light to obtain a sufficient signal-to-noise ratio without developing too much heat. Tunneling
devices appear to be the most sensitive but have an infinitesimal range. Silicon X-ray inter-
ferometers are perhaps equally sensitive but are very difficult to adjust. Thus in the areas of length
metrology, the advanced technology will require the ultimate in precision measurement. It will
certainly have a great influence on the practical realization of the definition of the meter and the
activities of the national laboratories in this field.
In order to detect relative movements, between different parts of an optical interferometer,
of parts in 1018 to 1021, great advances have been made.and still need to be made in servo-control,
position detection, and in ways of agumenting the signal-to-noise ratio in interferometers.
Why precise measurements and greater accuracy
We need to be able to make precise measurements for two reasons:
— to test fundamental physical theories of matter and the universe;
— to provide the basis for the application of these tested physical theories to practical life
through advanced technology.
We should continually strive for greater accuracy because advanced technology by its very
nature is continually searching for greater efficiency and improved performance in all of its
products. This inevitably results in continually increasing demands being placed on all of the factors
contributing to efficiency and performance—one of these factors is measurement capability.
Needless to say, improved measurement accuracy leads to improved efficiency in manufac-
turing high-technology products. It must be obvious that being able to make a measurement to a
given accuracy in only one tenth of the time will improve the efficiency of the process that relies
upon that measurement. Such an increase in speed can only come about, however, by a greatly
increased knowledge of the factors that previously limited the accuracy of the measurement, for
example, improvement in signal-to-noise ratio that would probably have been acquired in the
measurement system.
For international trade to be feasible, world-wide uniformity in measurement standards is
essential. International trade in high-technology products, communication, and navigation net-
works, the exchange of scientific information, as well as a multitude of pure and applied scientific
and technological projects carried out on an international basis, are all highly dependent upon
traceability of measurements of national, and hence, to international measurement standards.
