Sign in to follow this  
Michael Szyper

Stradivari Varnish oil/resin ratio

Recommended Posts

2 hours ago, sospiri said:

 

What myth are you refering to?

In general, I am willing to offer highly experienced counterpoint to myths of any kind. But in this particular case, it is targeted to shellac myths.

Some people highly appreciate stuff like that, and others don't.

Share this post


Link to post
Share on other sites
7 hours ago, sospiri said:

John, could it be from skin cells? I'm sure that there must be some keratin in my oil grounds from rubbing.

There could well be.  What Echard has noted in various papers might support this.  BB's findings possibly suggest that there may be a little more involved.  This remains an area where the jury is still out.

Share this post


Link to post
Share on other sites
On 8/17/2020 at 12:01 AM, David Burgess said:

Oil without any resin at all will dry just fine, and will even eventually become hard and brittle, given enough time, or if exposed to enough sun or UV.

If you add add a siccative it will even dry fast. These siccatives (e.g. cobalt) simply work as catalysts for oxygen induced polymerisation. This is described here https://en.wikipedia.org/wiki/Drying_oil

In the UK gun stock makers used to varnish with just linseed oil.

I am however still not sure whether violin varnishes are air dried or light dried as all violin makers use light boxes. Probably a mix of both. Can be measured by weight change - hgher weicht means oxygen induced polymerisation.

Share this post


Link to post
Share on other sites
On 8/13/2020 at 5:36 PM, Michael Szyper said:

How can you prove that a layer consists mainly of substance x and not substance y if a method (EDX) is only capable of detecting substance x?

As I mentioned, what I wrote is only an interpretation - i.e. the same what Barlow did with his data. I would like to discuss the methods used in those papers and mainly their weaknesses, because only then we can interpret the data in a better way. I just find it intriguing that the data provided by the publications is not all that different, but the interpretations does.

I agree with you that FTIR alone is not sufficient to provide exact resin:oil ratios. Brandmair wrote that the FTIR spectra compared best in the case of 80:20 varnishes. This does not necessarily mean it is definitely a 80:20 mix. 

These are good points. In the polymerisation process not all fatty acids get cross linked. One of the 3 fatty acids from linseed oil, oleic acid, won't do much at all. This means there are fatty aicd molecules embedded in the polymer. This may be the reason for the optical properties of these varnishes.

With FTIR I would assume to see the polymer and the free fatty acids. With GCMS one will only see the free fatty acids, the polymer will neevr go up that GC column.

What Echard measured with GCMS is therefore probably the free fatty acid content of the polymer.

 

Share this post


Link to post
Share on other sites
On 8/12/2020 at 10:00 PM, Michael Szyper said:

Dear friends,

Finally I arrived at a stage where I would like to share and discuss my thoughts about the following publications:

  • B Brandmair, SP Greiner „Stradivari varnish“
  • J-P Echard, S Vaidelich et al; „The Nature of the Extraordinary Finish of Stradivari’s Instruments“; Angew. Chem. Int. Ed. 2010, 49, 197-201
  • CY Barlow and J Woodhouse; „Of old wood and varnish: Peering into the can of worms“ JCAS Vol 1 no. 4 (Series II) Nov 1989 pp2-9

First of all, I would like to explain the very basics of the used and discussed methods:

FTIR: Fourier transformation of infrared spectra. Objects are irradiated by blackbody radiators. The specific absorption spectrums are analyzed and help to identify especially OH, C=O and CH2/CH3 groups.

Scanning Electron Microscope (SEM): Electrons are fired at an object, a very sharp picture is achieved. Conductive coating and vacuum are needed.

SEM/EDX: The electrons fired at a sample induce electromagnetic radiation (x-rays) with specific spectra, depending on the atoms. The radiation spectrum is detected and analyzed by its peak distribution. This method CANNOT detect atoms with lower atomic number than sodium. That means, that for example carbon, hydrogen nor oxygen can be detected by this method. (This point will be important later on, since these are by far the main components of colophony and linseed oil)

 

Let’s start with the naked results without any interpretation.

1. Brandmair & Greiner

B&G compared the FTIR spectra of fresh applied varnish made by Stephan Peter Greiner to the Strad varnished. Greiner made several varnishes with different oil:resin ratios. The FTIR of Greiners 80:20 varnish (final and applied product) matched best to the Strad FTIR spectrum. 

UV and vis microscopy could not identify filler-type particles within the ground layer. SEM pictures were not provided in the book.

Side note: UV and Vis microscopy can not identify colloidal suspensions of very small particles, they would appear as a homogenic mass. Brandmair also commented about this issue in the pigment chapter.

 

2. Barlow and Woodhouse

IMG_0522.thumb.jpg.d4751b254520957339a7a8b01961821e.jpg

This is the only SEM picture provided in the publication, which can be attributed to Stradivari with sufficient confidence.

I colored the left side to discern between the different layers. They found a very basic porous layer (red). There are almost no particles penetrating the porous structure. Then there is a ca. 60 µm thick layer (green colored). It appears inhomogenic, with amorphous particles of different sizes between 1 µm up to around 10 µm. A third layer (blue colored) sits on top of that, it is around 2 µm thick and appears quite homogenic.

 

In the SEM/EDX different minerals were found. Please keep in mind, they can not provide information, if and how much of organic material is in the sample due to the method. So, the varnish/resin content within this particulate matter can not be provided.

 

3. Echard et al

IMG_0523.jpg.e271d73b76fd89e4a8e6756a6d771d6b.jpg

UV-Microscopy shows three different layers. The lowest one being the wood, covered with a green-whitish fluorescing layer and the top layer with yellow-salmon like fluorescence patterns.

Here are the according FTIR spectra of the three layers.

IMG_0524.thumb.jpg.a28e52aa443ceeb336e0947462d40482.jpg

The main difference lies in the OH-band, which is very weak in the lower varnish layer. Echard stated this to be the lower varnish layer, comprising of pure linseed oil.

My interpretation of the results

You all know the interpretation of the authors, so I left it out. First of all, let us begin with the SEM picture of Barlow/Woodhouse. They state that the bottom layer (red) is wood, the middle (green) being the particulate mineral ground, and the blue colored stratum should be the color varnish, which was in their opinion almost gone (but they stated about the same sample that it had a striking brownish color).

If you assume that the middle layer are the varnish layers, the thin top layer (blue) could be a restorers french polish. The middle stratum with around 50-70 µm does not appear way too thin to be a full varnish layer.

Barlow/Woodhouse did not perform an FTIR analysis on this layer, only SEM/EDX. Therefore, there is lacking evidence against the middle stratum being a varnish layer without a mineral filler.

 

What if the middle layer (green) would be the actual varnish, for example a cold-combined high-resin low-oil varnish. This would be rather a colloidal suspension of oil and amorphous colophony particles than a solid („firm pill stage“) varnish. You could easily obtain a SEM/EDX spectrum like the one Barlow measured. You still would not know if it is 0,1% Si and 99,9% organic material.

So, with the lacking FTIR, the Barlow/Woodhouse results could also fit my interpretation.

Echard et al. In contrast did an FTIR of the base layer stating that this would be pure drying oil. I do not disagree with their results, the FTIR fits (if the base varnish has not treated with a lye).

 

If I make a cold mixed varnish and put a drop on a paper, the colophony stays where it is, the oil soaks in the circumference of the paper. So the varnish disarranges - drying oil leaves the varnish and soaks into the wooetd. Et voila, here we have the pure drying oil as a base coat layer in the varnish. After the solvent evaporates, the dissolved colophony may form small amorphous particles surrounded by linseed oil.

 

And what about B&G? They did not found a mineral layer in the base coat and used FTIR AND SEM/EDX. On the other hand, uv- and vis-microscopy would be too low in resolution to show the varnish actually as it is - a suspension of colophony particles and dried linseed oil.

 

I have the lack of a modeled experiment to approve these claims. I would need to apply cold-combined varnish on a well prepared wood and let it polymerize. After a bit of aging, SEM, SEM/EDX and FTIR needs to be done on my prepared samples. If they would match, this could support my theory. Until then this is not more than a humble theory.

 

I would enjoy any comment about this.

Michael

Dear Michael

I think this generally makes sense. I think Woodhosue got an unlucky spot with some minerals underneath the varnish. Maybe Stradivari used some minerals for sanding and some was left.

I can't see why you think a lye would make a difference.

Echard also used GCMS to proof the presence of the rosin compounds. THe results are conclusive. I still wonder whether he picked up soem left-over monomers, but then he thought about this and used PyGC-MS. The pyrolysis is to break down polymers - a dirty process though.

 

 

Share this post


Link to post
Share on other sites
4 hours ago, uguntde said:

 I think Woodhosue got an unlucky spot with some minerals underneath the varnish.

Maybe Stradivari used some minerals for sanding and some was left.

I can't see why you think a lye would make a difference.

Echard also used GCMS to proof the presence of the rosin compounds. THe results are conclusive.

I still wonder whether he picked up soem left-over monomers, but then he thought about this and used PyGC-MS. The pyrolysis is to break down polymers - a dirty process though.

 

 

Not meaning to insult you any but your words make me realize why we will never find out how old cremonese varnish was made - too many unknowns or what ifs.

Share this post


Link to post
Share on other sites

What oil/resin ratio would the following be?  3 lbs. linseed oil, 8 oz. juniper gum and 4 oz. aloes hepatica.  Yes, that read pounds {lbs}.

From the 1550 or 1603 D. Alexii Pedemontani de Secrets Libre Septem. or Secrets of the Arts.

Share this post


Link to post
Share on other sites
1 hour ago, uncle duke said:

What oil/resin ratio would the following be?  3 lbs. linseed oil, 8 oz. juniper gum and 4 oz. aloes hepatica.  Yes, that read pounds {lbs}.

From the 1550 or 1603 D. Alexii Pedemontani de Secrets Libre Septem. or Secrets of the Arts.

Assuming that the piedmontese pound of the 16th century also consisted of 16 ounces, and assuming that the aloe is considered a resin component, that shakes out to 12 oz resin and 48 oz oil, or a 1:4 resin:oil ratio. A long oil varnish.

Share this post


Link to post
Share on other sites
1 hour ago, uncle duke said:

What oil/resin ratio would the following be?  3 lbs. linseed oil, 8 oz. juniper gum and 4 oz. aloes hepatica.  Yes, that read pounds {lbs}.

From the 1550 or 1603 D. Alexii Pedemontani de Secrets Libre Septem. or Secrets of the Arts.

So breaking down the recipe into thirds I get 79.6 grams of resin/gum, 16 ounces of linseed oil and 1.33 oz. of aloe hepatica.  Seems kind of gooey to me. 

It is mentioned that essences to be used with linseed oil are lavender, rosemary and turpentine.  I notice the two words and turpentine - are these three essences all incorporated with each other or used individually or just use one of the three?  They are not mentioned in the above Pedmontese ingredient list.

Also mentioned elsewhere is the use of alum - imo, for the makers who use alum for their varnish making don't lose hope - your way could be right too.

Share this post


Link to post
Share on other sites
On 8/18/2020 at 10:11 AM, uncle duke said:

What oil/resin ratio would the following be?  3 lbs. linseed oil, 8 oz. juniper gum and 4 oz. aloes hepatica.  Yes, that read pounds {lbs}.

From the 1550 or 1603 D. Alexii Pedemontani de Secrets Libre Septem. or Secrets of the Arts.

Juniper gum has a water soluble/sugar component,  some oil and a small amount of the resinic acids necessary for a film forming varnish.  I would personally avoid it.

Share this post


Link to post
Share on other sites
7 hours ago, joerobson said:

Juniper gum has a water soluble/sugar component,  some oil and a small amount of the resinic acids necessary for a film forming varnish.  I would personally avoid it.

So it sounds like what you’re saying is it would be perfect for a ground varnish. 

Share this post


Link to post
Share on other sites

It's hard to jugde old varnish and old recipes by modern analysis and modern terminology.

Many years ago I read some of the manuscripts and tried to identify the measures and ingredients and found out that no one was really sure what exactly are the resins and oils. I recall that juniper resin was possibly sandarac,  but could be of any tree or shrub similar to common juniper growing around mediterranean. I tried to tap common juniper (juniperus communis) and it yelded barely few drops, you would have to tap hundred of trees and wait long time to get any usable amount from real juniper. There are some similar looking trees of other species Ive seen in Italy that ooze droplets of nice amber colored resin without asking I collected some of it but never got to testing it.

Looking at the Barlow SEM picture  (particulate) I immediately thought of old crumbly  plastics... varnish that is 300 years old may be similar to old plastics and piece broken off will possibly crumble into particles at the break. Modern varnish (even of the same composition) may break or cut cleanly just because it is fresh. Just my thoughts...

Few months ago after discussion aboout linseed oil drying soft (or not drying at all) and soaking into wood I took out old bottle of linseed oil I cooked with lead acetate almost 20 years ago (I cooked lead filings in vinegar and then added the white powder into oil and cooked like I read in some old manuscript recipes) The oil in the bottle is still iquid with some lead white precipitated on the bottom of bottle and after not too thin application on bare maple (painted on with finger) it dried overnight in my basement ,without any UV or sunlight, to nice thin satin varnish layer that would make pretty decent looking ground. Not too hard but not soft either. I tried to gouge the area and found no visible deep penetration into maple surface (though I didn't apply it to endgrain). Oil that I treated with modern (manganese) siccative didn't dry as fast. There are many myths about linseed oil on internet and in some varnish books as well.

Share this post


Link to post
Share on other sites
16 hours ago, joerobson said:

Juniper gum has a water soluble/sugar component,  some oil and a small amount of the resinic acids necessary for a film forming varnish.  I would personally avoid it.

Wouldn't the water get boiled out, and the sugars reduced to simple carbon during the cook process?

Share this post


Link to post
Share on other sites

Dried linseed oil will fluoresce after it yellows over time. Will this explain the fluorescence observed from the under coat beneath the colored varnish layer? I know that a lean clear varnish fluoresces very brightly. Will a thin linseed oil film do this too? Any research into this?

Share this post


Link to post
Share on other sites
11 hours ago, Michael_Molnar said:

Dried linseed oil will fluoresce after it yellows over time. Will this explain the fluorescence observed from the under coat beneath the colored varnish layer? I know that a lean clear varnish fluorescence very brightly. Will a thin linseed oil film do this too? Any research into this?

Oxygen dried linseed oil would keep some double bonds which are required for fluorescence.

Share this post


Link to post
Share on other sites
On 8/20/2020 at 6:11 AM, HoGo said:

Looking at the Barlow SEM picture  (particulate) I immediately thought of old crumbly  plastics... varnish that is 300 years old may be similar to old plastics and piece broken off will possibly crumble into particles at the break. Modern varnish (even of the same composition) may break or cut cleanly just because it is fresh. Just my thoughts...

That's what it looks like,  crumbly varnish with  a thin layer of french polish on top.   

Share this post


Link to post
Share on other sites

A varnish that's predominantly linseed oil is not going to wear like the Cremonese varnish does IMHO, coating a violin in linseed oil would be like coating a violin in a thin coat of rubber, its going to deaden the tone, not accentuate it. I think anyone that has a bottle of linseed oil knows what dried linseed oil looks like, nothing like rosin

Share this post


Link to post
Share on other sites

Frankly, any coating you put on an instrument is going to add mass and thereby damping. Varnishes of any kind do not improve the acoustic properties of an instrument. "Do no harm" is the guiding principle - less is more. 

What is the purpose of varnish? To protect thin bits of wood and accentuate the natural beauty of the materials. That's it. 

Share this post


Link to post
Share on other sites

I don't like unvarnished violins, they are too wild. You might be right that the sound is optimal without varnish, but we are used to the sound of varnished violins.

Share this post


Link to post
Share on other sites

Anyway I still think the whole oil/resin ratio is irrelevant in this case when it's estimated with tests, from +300 year old fragments.

Share this post


Link to post
Share on other sites
1 hour ago, Strad O Various Jr. said:

Actually unvarnished wood is somewhat raw sounding and brighter, varnishing with a good oil varnish tends to smooth over the highs and accentuate the midrange, that has been my experience, definitely improves the tone.

Varnishing with a good spirit varnish or a good essence varnish appears to do precisely the same, given the quantity of excellent violins that use these methods. I don't disagree with your observations, but the effects are subtle when good practices are followed and extreme (in the negative) when any kind of finish is overapplied.

1 hour ago, Peter K-G said:

I don't like unvarnished violins, they are too wild. You might be right that the sound is optimal without varnish, but we are used to the sound of varnished violins.

I don't mean to suggest that white violins are more tonally optimal, sorry. But I get my hackles up when varnish is spoken of as a major beneficial tonal/timbral component, as it frequently is in pulpy newspaper articles and unfortunately occasionally in trade periodicals. It is not.

Share this post


Link to post
Share on other sites
18 hours ago, uguntde said:

Oxygen dried linseed oil would keep some double bonds which are required for fluorescence.

So you are saying drying with no uv?

Share this post


Link to post
Share on other sites

Using expanded/super exploded view of the belly I can count five to six different layers.  Can anyone else see that or is my monitor a piece of junk?

Assuming my monitor works I see a possible glue layer, a doubtful eggwhite layer or evaporated gamboge, for example. - or in plain terms something that didn't sink down below the surface.

Maybe my monitor is a piece of junk.  

Share this post


Link to post
Share on other sites
On 8/20/2020 at 10:11 PM, HoGo said:

Looking at the Barlow SEM picture  (particulate) I immediately thought of old crumbly  plastics... varnish that is 300 years old may be similar to old plastics and piece broken off will possibly crumble into particles at the break. Modern varnish (even of the same composition) may break or cut cleanly just because it is fresh. Just my thoughts...

Or sag because it is too fresh...:D

Uncured varnish.JPG

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.