Chemical ghosts

Roy Wogelius has no fear of ghosts. The University of Manchester reader in geochemical spectroscopy claims to have already seen dozens of them, and if his current investigations go according to plan then he expects to see many more. But these ghosts are not the sort of spooks that go bump in the night, Wogelius is quick to point out; instead, they are the previously unseen traces of ancient chemistries long hidden in the Earth’s geological and archaeological record.

Together with fellow ghostbusters Phil Manning, a University of Manchester colleague and palaeontologist, and physicist Uwe Bergmann at Stanford University, US, Wogelius’ research is concerned with looking for the remnants of fossil chemistries: the feather keratin of Archaeopteryx, for example – one of the world’s most famous fossils and the possible evolutionary ‘missing link’ between dinosaurs and birds – or the ink sacs of a 90m year old squid.

‘“Chemical Ghosts” refers to the residue that can only be seen using X-rays,’ Wogelius says. In the case of Archaeopteryx, for example, ‘I still recall watching the screen as we were mapping phosphorus and sulphur and realising that, actually, there was a trace of the original feather chemistry left behind, which nobody had been able to resolve for over 150 years [since the first specimen was discovered]’.

While the Archaeopteryx feathers had already been inferred from impressions in the surrounding fossil sediments, the discovery of this remnant keratin chemistry was the first evidence that residues of the actual feather chemistry remained (PNAS, 2010, 107, 9060).

Discoveries of chemical ghosts such as the Archaeopteryx feather keratin have been possible only in the past few years, Wogelius says, thanks to the superior sleuthing capabilities of new and more powerful analytical techniques, particularly X-ray spectroscopies. Just as conventional medical X-rays are able to see inside the human body to reveal its inner workings, so these new X-ray techniques – coupled to particle accelerators that dramatically increase their penetrating powers – are able to probe the hidden recesses of fossils to discern the tiniest traces of elements left over from these long forgotten chemistries.

And, unlike traditional methods of element analysis, he points out that they do so non-invasively – without recourse to damaging sampling procedures. Indeed, ‘for extremely valuable specimens such as Archaeopteryx, which can’t be sampled, SRS-XRF (Synchrotron Rapid Scanning X-ray Fluorescence) is a critically important advance in analytical chemistry.’

Developed by researchers at Stanford Synchrotron Radiation Lightsource (SSRL) in California, US, SRS-XRF involves bombarding the specimen with high energy X-rays and detecting the secondary or fluorescent X-ray emissions when ejected electrons return to their corresponding electronic orbitals – similar to traditional X-ray fluorescence spectroscopy. However, by coupling the technique to a particle accelerator, the resulting X-ray beams are about a million times brighter and so are capable of rapidly detecting and mapping trace concentrations of elements – down to the part per million level – much faster than is possible with conventional X-ray beams.

‘The high brightness of the beam means that we can resolve elements at very low concentrations, about an order of magnitude lower than conventional methods,’ Wogelius says, ‘so that not only can we do chemistry faster, but we can see things that typically can’t be seen via other methods.’

In the case of Archaeopteryx, it was the tell-tale traces of phosphorus and sulphur that initially revealed the presence of the residual feather keratin chemistry, the Manchester researchers reported. Mapping the distribution of phosphorus in the fossil showed that the element was enriched at points corresponding to the feather shafts – which were clearly not merely sedimentary ‘impressions’ as was first thought. A similar patterning was also observed with sulphur, which is also present in modern bird feathers.

‘A big advantage of SRS-XRF is that it allows us to map chemistry about 3000 times faster than conventional methods,’ Wogelius says. ‘This means that we can map large objects, like the Archaeopteryx, within a feasible time scale, say over four hours instead of 12,000 hours.’

Even more unexpected than the Archaeopteryx findings, meanwhile, more recent work on another fossil – the 125m-year old bird Confuciusornis sanctus – has turned up traces of yet another chemical spook related to feather coloration. ‘The result I am the most amazed by was where we showed that the copper in 125m year old bird feathers has the identical organic complexation that you would find in copper contained within existing pigment ,’ Wogelius says.

‘The fact that this pigment, eumelanin, is in my own hair and skin and can be successfully traced back to and identified within animal remains from deep geological time still fascinates me and keeps me excited at the prospect of finding further traces of chemical residue in the fossil record.’

C. sanctus is of great interest to palaeontologists because it is the first documented bird species with a fully derived beak, a characteristic feature of modern birds. This work appeared in the journal Science (2011, 333, 1622) and was selected as one of the top 10 scientific discoveries of 2011 by the magazine La Recherche.

To determine the presence of eumelanin, the researchers turned to an alternative X-ray technique –absorption spectroscopy – to reveal information about the local chemical coordination environment of specific elements in the pigment. ‘We homed in on copper as a biomarker for eumelanin pigment and showed that the Cu was organically chelated and in a configuration almost identical to the Cu complexation within fresh eumelanin pigment,’ Wogelius explains. ‘This type of information concerning chemical bonding is absolutely critical to the interpretation of the maps we produce.’

The same technique can be used in parallel with fluorescence in order to map specific oxidation states of an element, say organic sulphur versus inorganic sulphur, he adds.

‘Because we are using a synchrotron, once we get the sample under the beam many other possibilities besides mapping are created. In particular, because we can control the incident beam energy, we can do X-ray absorption spectroscopy on points of interest that we uncover by mapping.’

One area where synchrotron coupled X-rays are proving particularly valuable, Wogelius says, is for the analysis of soft tissues, such as feathers, skin, hair, leaves, and the remains of invertebrates, which are often absent or obscured from the fossil record owing to their propensity to degrade. ‘Any information on soft tissue is useful, since the preservation potential is so low relative to bone and shell,’ he says, pointing out that ‘SRS-XRF is currently the only way to feasibly map large specimens with soft-tissue preservation’. While ultraviolet light has become popular and can often reveal the outlines of soft tissue residue, he points out that it can’t do element specific mapping.

Besides these intrinsic fossil chemistries, meanwhile, palaeontologists are also interested in finding out about their other chemical history – what elements may have been added or taken away during the passing of geological time. The surrounding sedimentary matrix is particularly important in this respect, Wogelius explains. ‘In particular, we look for enrichment in metals along fractures. We also look for aureoles of elemental loss or enrichment – SRS-XRF is a real advantage in this regard because we can map quite large areas and look for evidence of element mobility.’

The technique is also useful for revealing evidence of early conservation work to ‘patch up’ or ‘repair’ fossils, he continues. In the case of Archaeopteryx, for example, the group was able to highlight the bone ‘repair’ work carried out by early curators with a bromine-rich epoxy compound and identified ‘numerous fingerprints’ around the edge of the fossil as revealed by chlorine residues from sweat.

In the hunt for chemical ghosts, meanwhile, Wogelius says that opportunities to search for more hidden details within the fossil record abound. The group first became interested in using synchrotron radiation to study fossils back in 2007, and has since clocked up hundreds of hours of beamtime, though getting research time at a synchrotron is like getting a grant proposal funded, Wogelius admits: ‘Roughly I’d say it is equal to about £10,000 a day.’

He likens that the discovery of these fossil ghosts – specifically the eumelanin pigment – to the transition from radio to black and white TV. ‘Eumelanin pigment is dark brown to black. It is the most important pigment in the animal kingdom and is also present in eyes (except blue), hair, skin, squid ink and in many more tissue types.’ While we may never be able to restore fossil chemistries to the extent that we can see the original creatures in full colour, he adds that: ‘there is no question that we will resolve other pigments.’

Previously unseen works by Archimedes and Cherubini revealed

The first ‘high profile’ application of SRS-XRF was to reveal the writings of a 3^rd century physics’ BC treatise by Archimedes. Overwriting of a book by Archimedes – known as a palimpsest – had completely obscured the original text, the only known copy of his physics’ treatise The Method.

Despite the similar chemical compositions of the two superimposed inks, SRS-XRF imaging distinguished the underlying writing owing to subtle compositional differences combined with a 90 degree difference in orientation of the two texts, Wogelius explains.

Meanwhile, other more recent work by Wogelius and colleagues is also expected to lead to the first performance in full of the opera Médée by the Italian composer Luigi Cherubini. For centuries, the full score of Médée has remained hidden from view – overpainted at some point in its history so as to obscure completely the aria ‘Du trouble affreux qui me d’evore’ from the third act.

Fortunately, the heavy black overpaint was mostly carbon, which is relatively transparent to X-rays. While previous chemical analysis had revealed the presence of several trace metals in the Cherubini ink, it was impossible to decipher the original score.

The key to revealing the hidden notations, Wogelius says, lay in mapping the elements present in the original iron-gall ink. Iron produces a high energy X-ray emission signal at 6.4keV; however, a significant amount of potassium (K) is also present, which emits at a much lower 3.3keV.

At this lower energy, the paper absorbs sufficient radiation so text can be distinguished on both the front and reverse. By combining these K maps with the Fe maps to enhance clarity, the researchers were able to decode notation on both sides of the paper.

Cath O’Driscoll is deputy editor of Chemistry & Industry