Is the fine structure constant changing?
An international team of astrophysicists from the University of New South
Wales and others from Britain and the USA have raised the possibility that
the fine structure constant a might have changed
slightly as the universe got older. This is the conclusion they have reached
from spectroscopic studies of gas clouds that lie between us and distant
The value of a(alpha) is equal to
e2/hc, where e is the value of the charge on an electron (or a
proton), h (more correctly, 'h-cross', the letter h with a line through it)
is one way of stating Planck's constant*, and c is the speed
of light. For this constant to be changing, one of the values would also
need to change, and most of the comparatively few media accounts tended to
follow the lead of the New York Times, which suggested that it was the speed
of light which was changing, but physicists say this is the least likely
of the three 'constant' values to change.
But first, a quick look at the technical details: the researchers examined
ancient light being re-emitted from ancient (and more distant) gas clouds,
and compared what they saw with light being re-emitted from nearby and more
recent gas clouds. They then corrected their data to allow for the redshifts
caused by the movement of the gas clouds, and examined the spacing of doublets
of absorption lines in the spectra of a variety of different atoms in the
dust clouds. These gave a steady result: a consistent shift in alpha with
increasing redshift, especially where z>1.
The news broke almost two weeks before the official date of the print publication
in Physical Review letters, but raised remarkably little interest, in part,
perhaps, because a number of physicists suggested that there might be other
explanations (which they left carefully unspecified). The authors identify
13 potential systematic errors that could have produced the result, and indicate
that most of the possible false causes may be ruled out. The only two that
cannot be ruled out should have produced smaller values of alpha, not the
larger values that are being detected.
For example, one possible systematic error, caused by a wrong alignment of
the spectrograph slit would, if it had occurred, have reduced the observed
effect, not increased it, and they comment that:
''We summarise and stress the following important points regarding potential
systematic effects: (i) a thorough investigation reveals no systematic effect
which can produce the results we report. (Furthermore, applying either of
the 2 significant corrections would enhance the significance of our
Putting it another way, the value of alpha is very close to 1/137, and the
measured change is about 1 part in 100,000, while there remains a chance
of around 1 part in 10,000 that the observed effect is a statistical fluke.
That means the probability that the effect is real is 99.99%, fairly good
So whatever they have captured here, it would appear to be significant, and
that probably means that one or more of the values that physicists regard
as constant does in fact change overtime. So how do people go about looking
into the past, to assess what the value of the fine structure constant should
Quasars shine with the sort of intensity you expect from a whole galaxy,
but they appear to be 'tiny' - about small enough to fit inside the orbit
of Mercury, and that makes them an extremely small and bright point-source
of light, when they are seen from any distance. The other key feature of
a quasar is that they have a continuous spectrum, emitting at all
When you shine a continuous spectrum through a cloud of gas, some wavelengths
are absorbed and as Robert Bunsen and Gustav Kirchhoff knew almost a century
and a half ago, the wavelengths that are absorbed are exactly the same
wavelengths that would be emitted if the same atoms were 'excited', given
extra energy. The effect is slightly varied if the gas is whizzing away from
us, due to the effects of redshift.
Now the lines are also divided up, with small variations either side, and
the extent of this shifting is proportional to the fine structure constant.
(Note that this is a deliberately non-mathematical account of a complex issue.)
The redshift of radiation depends on how old it is, so we can determine the
age of each quasar, and we can also look at the extent of the shift away
from the central line. So if the extent of the shifting differs with age,
this would be a good reason to argue that the fine structure constant was
changing. It may not be a sufficient reason, but it is certainly a good
The sample looked at the quasars covering a period from 23% of the age of
the universe up to 87% of the age of the universe, using measurements made
on the Keck telescope's HIRES spectrograph at Mauna Kea on Hawaii, with
observations being made on frequencies associated with iron, zinc, magnesium,
chromium and other metals in the clouds, and the results for all the measurements
showed a consistent trend.
The fine structure constant is the fundamental constant of electromagnetism,
and may be thought of as a measure of the inherent strength of the
electromagnetic force. In the past, others have explored the possibility
of what are known as secular changes in fundamental constants, but the previous
cases have all been explained away by other means - which is perhaps part
of the reason why physicists have not been carrying on too much about the
result so far.
So what does it mean if the value of alpha has been shifting slowly with
time? The fine structure constant explains how electromagnetic forces hold
atoms together. There are certain stable orbitals around atoms, and there
is a very precisely known amount of energy that an electron needs to absorb
when it jumps from one orbital to one with higher energy, and the same amount
of energy is emitted when the electron drops back down again. Now it seems
that, in the past, this amount of energy may have been slightly different.
This means that physicists would need to go back and revise the 'standard
model', if, as the evidence seems to suggest the value of alpha has increased
by even a tiny amount. The people who will gain most from this are those
who hold that
theory is valid, because this theory holds that there are many more
dimensions than time and the three dimensions of space that we can see, and
if these are real, they could account for the value of alpha increasing.
It is lucky for us that the increase, if it is real, is small, because if
the value of alpha had increased too much, carbon atoms could not be stable,
and life forms like us could not exist.
*Note: h-cross or h-bar is actually Planck's constant divided
by twice the value of pi. The value of the fine structure constant alpha
is normally given in units which need not concern us here, because if h-cross
is changing, then so is h-bar.
Key names: John K. Webb, Michael Murphy, V. V. Flambaum and V. A. Dzuba of
the university of NSW, John Barrow (Cambridge), Chris Churchill (Penn State),J.
X. Prochaska (Carnegie Observatories) and A. M. Wolfe (UCSD). It is worth
noting here that Michael Murphy is just six months into his Phd studies,
and already has his first publication in the world's leading specialist
Our source for this is Peter Reece, also at the same stage in his Phd studies,
and second author in a paper which appeared in Nature for August23, a story
described in The glowing photovoltaic cell, this month. Peter and
Michael have known each other since they were 12, when they were at the same
school, St. Columba's High School at Springwood, in the Blue Mountains to
the west of Sydney. Is there something in the mountain air, or is it special
teaching? All we can ascertain from the school is that both students were
always keen on their mathematics and science, and there was an enthusiastic
To see an aftermath of this story,see
Is the speed of light
changing?, August 2002.
©WebsterWorld Pty Ltd/contributors 2002