Interesting research into protein ground layers using Infrared s-SNOM tech.

[jamesadashley]

Has anyone here run across this research by Lisa Vaccari, Marco Malagodi et al?

It seems to be an interesting addition to the ever mounting coating research field.

I can find no reference to it here on MN

It seems to me (by no means a chemist) fairly worthy work.

It also tallies with my own use of a protein layer...which is nice.

 

https://pubs.acs.org/doi/full/10.1021/acs.analchem.2c02965

[JacksonMaberry]

Thanks for the link! My reading of the article is that, though a new technique was employed, the jury remains out as to whether, conclusively, proteins were used deliberately. This is in line with previous researches, such as those presented in the Brandmair and Greiner Stradivari Varnish book. 

Here's the last paragraph from the article you linked:

It is possible to conclude that, for the analyzed Stradivari’s violins, proteins are too diluted in the wooden matrix to clearly stand out from the background at the microscale, whereas the nanoresolved approach clearly highlights the local chemical composition, overcoming the dilution effects of the analytes of interest. The analysis of two samples cannot provide enough information to allow for a generalization on the construction techniques detected for the Toscano and San Lorenzo, but the obtained information allows us to demonstrate that infrared s-SNOM may be the proper strategy to tackle the long-debated question about the use of proteinaceous materials by the Master of the Cremonese golden age lutherie. The results presented in this work encourage further applications of the technique, here used for the first time to study musical instruments of inestimable value, for a detailed investigation of very small complex samples within the Cultural Heritage field.

 

The authors still refuse to assert that proteins were used on purpose, but suggest it is possible and that further expansions of their research method could help answer the question. And I hope that such research will be conducted!

In my own experimental work, approaching the question of Cremona ground from a practical direction (preparing and applying a huge variety of materials in different combinations, orders, methods, etc) and observing qualitatively, no protein alone or in combination with another material has yet given the desired optical result. However, I haven't given up on it and will continue to try them. 

Some of my friends who also conduct research in this space, such as @joerobson, @John Harte, @Advocatus Diaboli may have thoughts about this article and the related phenomena. 

[jamesadashley]

Thanks for the reply Jackson.

Yes, I was pleased that they didn't go for an out and out 'this was intentional' kind of approach to their conclusions. Although of course some of the publications who ran articles based on it went with the usual 'Stradivari varnish secret uncovered!!' angle. 

I have settled to some extent on a casein layer. Not because I try to kid myself that it looks perfectly Cremonese, more that it allows me to get instruments out of the door without, I hope, too much oil in the pores.

[Advocatus Diaboli]

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

[Skeezy Bojangle]

Hey good find! A very interesting article.

[jamesadashley]

35 minutes ago, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

Good point, I guess they had to choose non visible areas so under the tailpiece or fingerboard..

As Jackson points out there are many many possibilities as to what it could be. In that sense these types of results ask as many questions as they answer.

[jamesadashley]

34 minutes ago, Skeezy Bojangle said:

Hey good find! A very interesting article.

Thanks Skeezy, I was surprised to find something from a few years ago that hadn't been discussed on here before.

[JacksonMaberry]

1 hour ago, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

Great thoughts, thank you. 

[Davide Sora]

8 hours ago, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.

For this analysis, they had to do invasive sampling, which is already unusual. They got permission for the hidden areas of the top only; they did not get permission for maples.

[LCF]

6 hours ago, Davide Sora said:

For this analysis, they had to do invasive sampling, which is already unusual. They got permission for the hidden areas of the top only; they did not get permission for maples.

There are some complex paragraphs in that paper which take a long time to decode but they seem to be saying that any amount of protein detected was quite dilute. Isn't it more likely to find washed out glue from a centre joint in the location that the samples were taken from, either hide glue or casein? Or are those one piece tops? 

[John Harte]

On 2/19/2025 at 8:10 AM, JacksonMaberry said:

Thanks for the link! My reading of the article is that, though a new technique was employed, the jury remains out as to whether, conclusively, proteins were used deliberately. This is in line with previous researches, such as those presented in the Brandmair and Greiner Stradivari Varnish book. 

Here's the last paragraph from the article you linked:

It is possible to conclude that, for the analyzed Stradivari’s violins, proteins are too diluted in the wooden matrix to clearly stand out from the background at the microscale, whereas the nanoresolved approach clearly highlights the local chemical composition, overcoming the dilution effects of the analytes of interest. The analysis of two samples cannot provide enough information to allow for a generalization on the construction techniques detected for the Toscano and San Lorenzo, but the obtained information allows us to demonstrate that infrared s-SNOM may be the proper strategy to tackle the long-debated question about the use of proteinaceous materials by the Master of the Cremonese golden age lutherie. The results presented in this work encourage further applications of the technique, here used for the first time to study musical instruments of inestimable value, for a detailed investigation of very small complex samples within the Cultural Heritage field.

 

The authors still refuse to assert that proteins were used on purpose, but suggest it is possible and that further expansions of their research method could help answer the question. And I hope that such research will be conducted!

In my own experimental work, approaching the question of Cremona ground from a practical direction (preparing and applying a huge variety of materials in different combinations, orders, methods, etc) and observing qualitatively, no protein alone or in combination with another material has yet given the desired optical result. However, I haven't given up on it and will continue to try them. 

Some of my friends who also conduct research in this space, such as @joerobson, @John Harte, @Advocatus Diaboli may have thoughts about this article and the related phenomena. 

Jackson, your comments and quote provide a good summary of the Arvedi Lab group findings, at least as they relate to Stradivari micro sample material. They, and others like Echard and Brandmair, are yet to conclusively identify the presence of a discrete protein based size/layer, such as a glue size forms, in any Stradivari micro sample material.

FWIW, the Arvedi Lab group also have maple micro sample material taken from the Toscano.

[Dr. Mark]

Sounds like the Toscano will be naked as a jaybird soon with nothing even compelling, much less conclusive, to show for it.

[John Harte]

15 hours ago, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

Yes a very vague possibility but even very dilute glue sizes will show up as a discrete presence in micro sample material at the magnification seen in their micro sample images. I'm certainly not seeing any presence in the Toscano UV images in the Supporting Information. These provide particularly clear detail.

[John Harte]

31 minutes ago, LCF said:

There are some complex paragraphs in that paper which take a long time to decode but they seem to be saying that any amount of protein detected was quite dilute. Isn't it more likely to find washed out glue from a centre joint in the location that the samples were taken from, either hide glue or casein? Or are those one piece tops? 

Both instruments feature 2 piece tops.  It seems from Figure. 1 that both micro samples were taken from areas slightly to the right of the centre joint.

[JacksonMaberry]

Thank you as always John, I never feel fully confident in my readings of such papers until I've run them by you first!

So for our topic creator, it seems this doesn't provide the support they might have hoped for their use of protein size

[John Harte]

16 minutes ago, JacksonMaberry said:

So for our topic creator, it seems this doesn't provide the support they might have hoped for their use of protein size

I agree. As you have said, the jury remains out.

I don't doubt a protein presence and, like you, I hope that research will continue in this area. There are several possibilities that are yet to be considered but I suspect that other researchers may need to get involved and that may take some time...

[Advocatus Diaboli]

8 hours ago, John Harte said:

Yes a very vague possibility but even very dilute glue sizes will show up as a discrete presence in micro sample material at the magnification seen in their micro sample images. I'm certainly not seeing any presence in the Toscano UV images in the Supporting Information. These provide particularly clear detail.

Quite right.  Just playing Devil’s Advocate here.  
 

I think proteins showing up as a result of cleaning up glue from say purfling inlay is much more likely.  Protein presence in lake pigment particles is also something that I think needs to be studied more as well. 

[Andreas Preuss]

For me new publications especially on varnish often leave a ‘show off’ taste behind. Its more about what new technology is capable to do. Nobody asks if a new examination method (which was certainly not invented to examine Cremonese varnish layers in the first place) is really useful. 

I wished researchers would examine also comparable test samples where we know exactly what was applied in which amount and what method. So in the above mentioned research I wished someone had taken a piece of wood add deliberarltely a layer of different protein materials and also make the test of washing out the brush in hot water and go over the wood (thinking that you’d apply only hot water). 
 

Just finding ‘something’ isn’t a big help to clarify ancient varnish methods. My rough guess is that if we would just use known examination methods and focus more on comparing with test samples we produce ourselves we’d cut off most of the speculations which are floating around. 

[jamesadashley]

11 hours ago, JacksonMaberry said:

So for our topic creator, it seems this doesn't provide the support they might have hoped for their use of protein size

That was supposed to be a bit tongue in cheek really, i've tried all sorts as a first layer over the years. Glue, gesso, casein, pop, liquin, Shellac, the list goes on. I keep going back to casein as it is quick and easy and works for me. 

I'm sure many of us would love to find that what works for us turns out to be the real deal but now in my 50's I'm not convinced we'll find out till I've had to hang up my gouge or plaster of paris'd my clogs.

[John Harte]

3 hours ago, Advocatus Diaboli said:

Quite right.  Just playing Devil’s Advocate here.  
 

I think proteins showing up as a result of cleaning up glue from say purfling inlay is much more likely.  Protein presence in lake pigment particles is also something that I think needs to be studied more as well. 

As I suspected

The two samples considered in this study are some distance from purfling, so unlikely in this case.  Having said this, repairs or glue lines near to sampled sites could well explain at least some protein presence.

As you and I have discussed in the past, there are a number of yet to be considered areas that could potentially prove interesting. The problem is gaining access to and convincing researchers to reconsider prior interpretations of data. Once something has been published this becomes unlikely.

[Dr. Mark]

5 hours ago, Andreas Preuss said:

For me new publications especially on varnish often leave a ‘show off’ taste behind.

Not that that would not be bad in an otherwise good paper, but the content of this one is pretty much 'We have this cool technology with a neat acronym.  We examined varnish samples from -->STRADIVARI<-- violins to find a protein layer.  We found some proteins.

[John Harte]

On 2/20/2025 at 9:47 PM, Andreas Preuss said:

For me new publications especially on varnish often leave a ‘show off’ taste behind. Its more about what new technology is capable to do. Nobody asks if a new examination method (which was certainly not invented to examine Cremonese varnish layers in the first place) is really useful. 

I wished researchers would examine also comparable test samples where we know exactly what was applied in which amount and what method. So in the above mentioned research I wished someone had taken a piece of wood add deliberarltely a layer of different protein materials and also make the test of washing out the brush in hot water and go over the wood (thinking that you’d apply only hot water). 
 

Just finding ‘something’ isn’t a big help to clarify ancient varnish methods. My rough guess is that if we would just use known examination methods and focus more on comparing with test samples we produce ourselves we’d cut off most of the speculations which are floating around. 

Andreas you raise some important issues. This is good.

I agree that publications outlining the use of new technology/analytical techniques can often appear as you describe. However becoming aware of and then applying such technology is a necessary first step in discovering what any new technology is actually capable of. I think that the technology used in this study could prove useful if the researchers involved were to broaden the scope of their inquiry. At present their view of what might potentially be involved is somewhat limited and it is this that has, in my opinion, limited what they have been able to detect. What you think you are seeing is often a function of the frame of reference that you are using. Background reading, historical knowledge, knowledge of the range of possible materials involved, practical experience in working with these materials etc., are important contributors to developing potentially better and ultimately more useful lines of investigation. In this regard, Brandmair's findings regarding Stradivari's varnish clearly benefited from her collaboration with Greiner.

I am with you in your wish that “researchers would examine also comparable test samples”. Your mentioned experiment would have been a good start. In the past researchers were more inclined to attempt to replicate what they thought was involved as a practical means of testing the validity of their interpretation of research data. Unfortunately it now seems that using earlier published interpretations of data, often not tested, is adequate.

It is possible for you to carry out your own tests using a relatively cheap microscope. Your mentioned experiment would be easy to do. More complex varnish systems are more difficult to make sense of and will require you to build up a body of knowledge of what can contribute to what you might be seeing. The below photo is a very clumsy low resolution attempt to compare features in a sample of my own against those in an image published in the research discussed in this thread.

[David Beard]

On 2/18/2025 at 10:28 PM, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

In the of a more typical historical woodworking shop this would be the easy and natural likely path. IMO.

[David Beard]

The trend in recent years seems to be toward new methods that allow confirming materials in very localized ways, saying not just what was found but also precisely where it was found.

This study isn't telling us exactly where, but it is saying it's looking at very small localized spots.

I applaud all such efforts.  Classical finishes are not at all homogenous.  

Because of this, 'finding' tells us more than 'not finding'.   

For some reason, a number of violin folk seem committed to denying any meaningful role for proteins in classical finishes.  I don't understand why people are committed to such a bias??

Protein use, particularly sizing with simple glue, was a very common and normal process employed in a great range of historical arts.  Woodwork also used glue sizing historically.

I truly don't understand the inclination to deny this common thing in the case of old Cremona violin making.

[David Burgess]

On 2/19/2025 at 1:28 AM, Advocatus Diaboli said:

I'm curious why they only analyzed samples from the tops.  Glue sizing before scraping or abrading could be a possibility?

 

2 hours ago, David Beard said:

In the of a more typical historical woodworking shop this would be the easy and natural likely path. IMO.

Which would be the more natural likely path? Analyzing only samples from the top, or glue sizing before scraping or abrading?

Glue sizing prior to final scraping wouldn't tend to produce the textured look on the top that we associate with the best-preserved old Cremonese instruments.