FTIR and the Shroud

Alan Adler (‘Further Spectroscopic Investigations of Samples of the Shroud of Turin,’ The Orphaned Manuscript, 2002), and Robert Villarreal (‘Analytical Results on Threads Taken from the Raes Sampling Area (Corner) of the Shroud,’ Presentation at the Ohio Conference, 2008) made FTIR spectra of numerous extracts from the Shroud, and derived conclusions from their spectra, which they thought supported the authenticity of the Shroud. In my opinion they were mistaken.

Adler’s paper is particularly thorough. “Five fibers representing non-image areas in the vicinity of the feet, the waist, and the head of the frontal image were collected from their designated sample tapes. Similarly, four waterstain fibers from the head and knee areas, four scorch fibers from the knee area, two serum coated fibers from the edge of the lance wound, two image fibers from the finger area, two backing cloth fibers from the area adjacent to the radiodate sample area, and two blood globs (particles unattached to fibers) from the lance wound area were also isolated for investigation.” He was also supplied with three threads from the riserva part of the radiocarbon sample. “Five fibers were taken from each of these samples for comparison with those collected from the sticky tapes.”

From these 26 fibres, 34 FTIR spectra are published, together with 8 spectra derived from various simulations and controls.

The spectra derived from FTIR analysis show peaks and troughs of infrared absorbance or radiation at different wavelengths, each corresponding to a particular molecular bond. Typical absorbance spectra for flax fibres look like this:

(From: ‘The Application of FTIR Microspectroscopy in a Non-invasive and Non-destructive way
to the Study and Conservation of Mineralised Excavated Textiles,’
Christina Margariti, Heritage Science 7, 2019)

The wavenumbers referenced are related to specific bonds as follows (from the same paper):

Note that peaks for the same bond may shift somewhat depending both on the material being tested and the method of testing. Adler describes that he collected Transmission Spectra.

And this diagram, but note of reflectance spectra, from the same paper, compares flax to other cellulosic fibres:

Here are Alan Adler’s spectra from non-image fibres, made entirely of undamaged flax, with such wavenumber references as I can determine – the x-axis is far from precise:

All these peaks are indicative of cellulose, which is what linen, and cotton, are mostly made of.

Now let’s compare these spectra with the “Standards” presented by Robert Villarreal at the Ohio Conference in 2008:

“Very different,” Villarreal comments, and so they are: no wonder he decided that all the fibres he rested were cotton. However, there is no possibility that his ‘standard’ for linen was in fact any such thing. He did not give a source for his standards, and presumably nobody else knew much about the FTIR spectra of textiles, which is why nobody realised at the time that his demonstration that the radiocarbon area samples contained cotton was actually meaningless. This is further discussed in a previous blogpost (‘Raes Ruminations’).

Moving on to Adler’s spectra from the image and backing-cloth fibres:

The essential difference between these spectra and those of the non-image fibres is the 1000 – 1200 wavenumber range, which in the image spectra is enhanced, and in the backing cloth spectra reduced. Rather to our surprise, the first resembles the spectrum of cotton – see these spectra from Chinkap Chunga et al., ‘Characterization of Cotton Fabric Scouring by FT-IR ATR Spectroscopy,’ Carbohydrate Polymers 58, 2004:

Why this might be is a matter of speculation. The chief chemical difference between linen and cotton is that cotton is almost pure cellulose (95% or so), whereas linen has more hemicellulose and lignin. If these decomposed, perhaps as part of the image making process, then perhaps the linen spectrum would indeed resemble cotton, as above.

The spectrum of the backing cloth resembles that of madder-dyed textile, as suggested by this spectrum from M. Zarkogianni et al., ‘Identification and Quantitative Determination of Madder
by High Performance Liquid Chromatography: Application to Historical Textiles,’ Journal of Liquid Chromatography & Related Technologies, 2009:

Although Adler claims that “the backing cloth pattern is readily distinguishable from the other patterns,” it actually matches the water-stain spectra quite well, which again suggests that colourant is suppressing the 1000-1200 wavenumber peaks.

Adler detects a “progressive oxidation-type pattern” from non-image, through water-stain, image, and scorch to “finally” radiocarbon. See the composite diagram below. This is what he says:

“The position (given in cm^-1 )and relative intensity of the peaks in the carboxylic acid salt region (1650 – 1540) [yellow] and conjugated ketone region (1680 – 1640) [blue] show an apparent progressive oxidation-type pattern with the non-image (1593, 1643) the weakest, then water stained (broad 1697, week 1640), then image (strong 1694, 1645 shoulder), then scorch (1591, broad 1645), and finally the radiocarbon pattern (1590, 1643, both strong).”

I’m afraid this is weak. I can, and have, selected spectra that just about correspond to this description (see below, left), but I can, and have, selected other spectra from the same groups which do not fit it at all (see below, right). The gradual increase in size of the bump at about 1600 cm^-1 is largely an illusion.

Adler goes on to note that bonds typical of proteins are conspicuously absent. “There is no evidence of the typical amide pattern (1695 – 1630) associated with proteins.” On the other hand 1695 – 1630 is a range pretty well coincident with the “conjugated ketone region (1680 – 1640) [blue], so at this level of precision it can’t be said that this has been demonstrated.

Proteins that could occur on the Shroud include blood and a paint medium such as tempera. A typical spectrum for blood looks like this:

(from: ‘Health Economic Evaluation of a Serum-Based Blood Test for Brain Tumour Diagnosis: Exploration of Two Clinical Scenarios,’ Ewan Grey et al., BMJ Open 8(5), 2018)

And here is Alan Adler’s spectrum of a sample of dried blood:

The two spectra above seem to me to match quite well, demonstrating the accuracy and precision of Adler’s work. But here are his three spectra from two ‘globs’ of blood from the Shroud.

Comparing the Shroud with the control sample, Adler says: “The spectral pattern of the blood globs from the sample tape is in good agreement with the spectral features of the various blood controls.” Not to me, it isn’t.