SULFUR, IRON, ALUM, AND OTHER INORGANICS IN WOOD
Vasa—Recent Preservation Research
Lars Ivar Elding
The Swedish warship the Vasa sunk in 1628 and was recovered in 1961. It is 900 tons, 60 meters long, full timbered with a 40-50cm hull and had 20,000 objects on it. Since it was found, it has taken 50 tons and PEG and 15 tons of borates to preserve it, and maybe 5-10% of its current mass is PEG. The 200kg pentachlorophenol it was treated with as a biocide has decayed over time and is no longer observable. It went from having 150% water content down to only 10% today. The ship gets some 1.1 million visitors every year. In the 1990’s salt precipitates were discovered, made from sulfate salts reacting with the iron, causing an internal formation of sulfuric acid. This was the subject of a 2002 paper in Nature, but now we know this was just one of several reactions and the situation is rather complicated. From 2003-2006, the “Preserve the Vasa” project aimed to answer some questions: What was the microbial activity? (negligible) How about sulfur-iron chemistry? Removal of the iron catalyst (EDHMA and DPTA dunno if I got those acronyms right though) PEG stability (formic acid found but decay in practice was negligible) Influence on the wood? (Yet to be determined) The next phase, the 2008-2011 “Future for the Vasa” project, asks, Which processes are important? How extensive? How fast? How is the wood effected? Lot of oxygen and a pH 4 is needed to keep reactions going. There are low pH values deep in the timbers, so perhaps reactions are going on there? The surface layer is weak and degraded, then there is an area which is rich in iron 3+ and then deep inside there is iron 2+. Somehow, a high concentration of sulfur inhibits iron-dependent degradation. Cellulose hydrolysis caused by sulfur is an over-simplification. Because of all the visitors, they are limited in the kinds of techniques they can use. In 2004, a new climate system was installed. RH 55% and temp 20C. Iron bolts are being changed out for carbon fiber or stainless steel. The ammonia gas that might help only has 1-2cm penetration, look for a 2010 Studies in Conservation article about it. Right now, we ought to take a sample and store it at low temperature and low pressure and look for change over time. Wish we had samples from 10 years ago, even though it does not look like there have been major changes in that time, who knows?
Iron Removal from Waterlogged Wood and the Effects on Wood Chemistry
Vicki Richards, Ian Godfrey, and Kale Kasi
Recovery in 1985 of an iron steam engine from the 1872 wreck of the SS Xanthro, had electrolysis in sodium hydroxide for years. In 1994, the baseplate was removed and there were some 100 wood chocks under there, coated in tallow (a lubricant for the engine.) From 1991-93 they treated them with dithionite and citrate and PEG 400 to preserve them and remove the iron. Took some 20 days for the iron to start coming out, and it seemed to come out best with dithionite. The solution got pretty acidic, below 3. They tried diethylentriamine penta acetic acid, but it wasn’t as good. Tallow interfered with the FTIR peaks when they tried to look at effects of chelating agents on the deterioration of the wood. I have to admit, I didn’t understand this chemistry very well due to my own lack of background.
Nuclear Magnetic Resonance and Fourier Transform Infra-red Spectroscopic Analyses of Acid-Affected Waterlogged Archaeological Wood
Ian Godfrey, Vicki Richards, Lindsay Byrne, and Emil Ghisalberti
Trying to look at degradation of Vasa to reference wrecks still on the seabed to see what kind of deterioration happens after treatment compared to doing nothing and leaving it in situ. Had to have big samples, 200mg, so there is sample left for other kinds of work also. Seems like there was more cellulose loss from the Vasa, and the inner part has lost more than the outer part. Perhaps areas that have more formic acid, like the outer areas, indicate more PEG since formic acid is a degradation product? There was more acetic acid on the inner parts. Truth is, there is more PEG degradation in the inner core even though there was more formic acid on the exterior. This is another talk where I could not follow all the science very well, a little above my head. I actually wrote in my notes at one point, “Big ol’ brain on Ian!”
The De-Acidification of Waterlogged Archaeological Mary Rose Timbers with Strontium Carbonate Nanoparticles.
The Mary Rose was the flagship of Henry VIII, dating to 1545. About half survived in silt and was recovered in 1982. Hoping to have money to rehouse it in 2012. There are some 19,000 artifacts as well, and they have a sulfur problem. They are using XANES (xray absorption near-edge structure spectroscopy) to see different oxidation states and distinguish between the sulfate compounds. Before, sulfur was “spectroscopically silent.” SrCO3 good because strontium was a better marker when used at .005 molar, and it became associated with sulfur in the wood. Forms SrSO4? Stronium sulfate as an insoluble material, something more stable to leave in the wood?. Gosh, this was pretty beyond my ability to grasp as well. These heavy-science talks exceed my science knowledge, so I can’t interpret them very well. Sorry about that.
Extraction of Sulfur Compounds from Archaeological Wood by Chemical Oxidation with Sodium Persulfate
Khoi Tran, Fanny Bauchaud and Clement Werner
Trying sodium persulfate to chemically oxidize the iron sulfides, 13.5% worked the best. Buffered the acidic solution with sodium hydrogenocarbonate NaHCO3. On the LaLomilla wreck, they were finding 80% pyrite, 20% silica and pyrite accretions, with 17% ash on the surface and 7% inside of the plank. 5 baths needed to soften accretion so it could be mechanically removed with a brush. Only about 10g removed from a 6kg plank? The PEG treated and freeze dried. There were still minerals in there and PEG could not fully penetrate, so there was some shrinkage. Also tried to do a poultice application on the crusts that came out after the PEG treatment. Cellulose pulp, .14M sodium persulfate, and 0.05 sodium bicarbonate buffer. 10 months later, they still have pyrite but look OK with no effluorescence. Maybe try electrophoresis in a poultice for in-depth extraction of ionic compounds? That has worked in the past to extract alum from wood.
Re-Conservation of Wood from the Seventeeth Century Swedish Warship the Vasa with Alkoxysilanes: A Retreatment Pilot Study Applying Thermosetting Elastomers
Carlos Cabrera Tejedor
In my notes at this point, I have written the remark, “OMG, I am the least illustrious person in this room!” Carlos was using Xiameter PMX-200 silicone fluid CAS # 70131-67-8, Dow Corning SDF-1 (about 75CST viscosity, ranges from 55-90) and the MTMS brand name Xiameter OFS-6070 silane CAS# 1185-55-3 as the crosslinker. Carlos was looking at color using a Munsell color chart, texture, dimensional change, weight, volume variation and microscopic features. His first test was to immerse the Vasa sample in MTMS at 70C to extract PEG over about 20 days, dry it, and catalyze the MTMS. Was a little waxy, heavier, and maybe they could not get out the PEG enough? Second test was soaking PEG out with deionized water, progressive removal of the water with ethanol and then to acetone in 25% increments, impregnating with a little vacuum to help get all the acetone out, cleaning, and then catalysis with DBTDA. The evaluation was OK, except for the volume variation and the microscopic qualities. However, in the third test he removed the PEG in water only, not just in one step with MTMS, and the results were much better. A gallon of silicone oil is $160, and a pint of MTMS is $75. So if I understand right, the alternate MTMS treatment is better than either the traditional one or the standard silicone oil treatment when re-treating PEGged wood from the Vasa. PROS: short time to do it, not too complicated to perform, and means that minimum preventive conservation would be needed to curate the wood. Thin layer of agent (a few microns) allows much of the woody quality to be retained and the morphological features to still be easily observed. Treated wood is hydrophobic, chemically inert, resists acids and bases, and resists change with UV light. CONS include non-reversible, non-re-treatable (only repeatable with same chemicals), price of reagents, and the health and fire hazard.
I have to say I spent a lot of time talking with Carlos during the conference, and liked him very much. He was personable, candid, competent, curious, enthusiastic and open-minded. Before Carlos, silicone oil was firmly in the “do not trust it” pile of treatment options that I had no intention of considering. Now that I have spent some time with Carlos, I feel like I have a much better understanding of the silicone oil issue. For example, I had recently heard that while it is NOT reversible, it is “retreatable.” Indeed, this term “retreatable” has gotten a lot of traction in the conservation world now, admitting that we are not ever going to get everything back out again safely, so “reversible” is in our the theoretical standards and ethics while “retreatable” is kind of the real-world version. However, in chatting about it, Carlos now plans to use the word “REPEATABLE” with silicone oil treatment, because in fact you do not have any other retreatment options with other materials. You can only repeat the original treatment to try to get better results. And Carlos was very candid in saying that when you get good results from the silicone oil treatment, they are really beautiful. To hear him talk, the results are almost seductively, unbelievably beautiful. Only the trained hand could really say, “hmmm, this surface is maybe a little too silky…not quite exactly natural but damn close.” (my own quote) But when it goes bad, it goes really really bad and there is no way back. Carlos’s term, “It is like a bulldozer.” And then there is the issue of the lab where all the LaBelle items are being done. Time will tell, but if the treatments hold up, that will be lovely, and if they don’t… well, the vast majority of all the organics from that ship, thousands and thousands of artifacts, have been treated since 1998 with this material. When you have a hammer does everything look like a nail?
Possible Uses of the Calcium Complex of EDTA to Remove Iron
Mags Felter and Anthony Crawshaw
Mammoth tusk needed iron removed in order to improve aesthetics, reduce risk of acidification and reduce risk of oxidation. Godfrey’s 2002 article showed that EDTA or diammonium citrate on bone or ivory strips the calcium and demineralizes the substrate. If you stir in calcium hydroxide to a 1.1molar strength, you can do a treatment for up to 40 days that works better. The problem was that good iron removing chelating agents pulled out Ca because the calcium complex is less stable than the iron complex. Adding in calcium helps.
The Use of an Electric Field for the Removal of Alum from Treated Wooden Objects
Iben V. Christensen, Lisbeth M. Ottosen, Poul Jensen, Inger Bojesen-Koefoed, Hartmut Kutzke, Mikkel Christensen, and Tom Sandstrom
Alum was used extensively in the Scandanavian countries for waterlogged wood until 1960’s. Examples: Hjortspring, Oseberg, and Arby. Alum did not penetrate very deeply, so inner part of the object often left unimpregnated. Internal cracking happens. Alum makes wood heavy and brittle, but not stronger. This experiment put kaolin clay poultices on both ends of the wood with electrodes in the clay so you would not have to put screws in the wood. Trying to make the potassium and aluminum go to the cathode (has citric acid there) and the sulfate to go to the anode (has CaCO3 there.) Have to soak it in water so that the ions can migrate? No, but RH needs to be around 90%. Took 3-4 days, at 4V and 3-5mA current. There was no significant removal of aluminum, but there was good removal of the sulfate and the potassium. Problems: possible electrode reactions, humidity needed, risk of unexpected migration of ions, and that the removal was incomplete, maybe 70-80%. If I understand correctly, the typical method for retreating objects treated with alum is prolonged soaking in water, and maybe this is hard on the delicate degraded objects, so coming up with a method like this where you didn’t have to submerge the artifact would be an improvement?
Past Conservation Treatments and their Consequences: the Oseberg Find as a Case Study (peer reviewed)
Susan Braovac and Hartmut Kutzke
The Oseberg sleds date from 834AD and were excavated in 1904, treated with alum sometime before 1914, and are unstable now…weak and very acidic (pH of 1) with pieces starting to fall off. The treatment involves alum (aluminum potassium sulfate) dissolved in 90C water, a coating of linseed oil and a final lacquer. Alum did not fully penetrate, only going in 5mm across the grain. Cracks and voids inside the wood not easily seen. Aluminum makes lignin complexes? Is sulfuric acid being produced inside the wood? Question is, now does the alum need to be removed, or is there a material that could be added to stabilize things?
Reconservation of Wood Treated with Alum in the 1920’s – Challenges and Strategies.
We must be respectful of past efforts. Otherwise, many of these objects would not be here at all. Some kinds of objects look very different on exhibition as a result of their treatments and it is difficult for the public to interpret what that means. Conservation history is another dimension of an object’s history. Alum was invented as a treatment in 1859 and was popular for about 100 years. It was the first method used to preserve archaeological wood. There was shrinkage and collapse, often, and the wood looked dry after treatment. Glycerol was added in 1900 as a possibility, but that went badly. By the 1960’s PEG began to be used. There was an attitude in the 1970’s and 80’s that all the alum treated wood needed to be re-treated with PEG. There is a certain fashion aspect of conservation in terms of what an object ought to look like. There is also the joy of learning a new method that makes these aesthetic sensibilities change, too. Conservation treatments and the way archaeology was documented in past eras is really interesting and should be part of what we try to preserve. If an artifact is stable, resist the urge to re-treat it in the latest fashion. Consider leaving those old fills and old mountmaking styles in place as its own documentation.
Accelerated Aging of Recent Oak, Impact from Iron Ions and Oxygen on Mechanical Properties in the Longitudinal Direction (poster)
Gunnar Almkvist, Charles Johansson, Ingela Bjurhager
Known that iron2+ in the Vasa is contributing to iron catalyzed processes in the wood (Fenton type reactions?) , they attempted to impregnate fresh oak with iron ions. They think that the changed wood preoperties from the use of PEG occur in the radial and tangential directions because the PEG reduces the compression strength, but the iron affects the axial tension strength. They found a 39% decrease in strength. What about the ability of sulfur and PEG to inhibit degradation from iron?
Removing Iron Compounds from a Waterlogged Wooden Gun-Carriage Using the Chelating Agent DTPA (poster)
Ebba Phillips and Inger Nystrom Godfrey
Needed to treat an oak gun carriage that was eroded, cracked, and had high levels of iron and sulfur. Problem with citrate or EDTA is that you cannot put it down the drain. DTPA (diethylene triamine pentaacetic acid) is a chelating agent that can be poured down the drain. They used only 1%, which is half of what others have used, in 1000L of water, and removed a kilogram of iron in the first 40 days. You need to do analysis of how much iron is coming out so you know when to renew your system.
Excavation and Stabilization of a 17th Century Wicker Basket: New Application of a Known Method
Jill Barnard, Liz Goodman, and Nancy Shippen
Nancy Shippen is now at the MAC lab. In 2006, a 17th century willow basket was found in a controlled excavation and block lifted. There were 2.5cm concretions inside. They sprayed it with 20% PEG 200 and then 20% PEG 4000. It was then vacuum freezer dried. But how to remove the concretions? They decided on glass beads of 44 micron diameter that have been used on bone, as the round shape of a glass bead was less likely to become embedded in the basketry. They used the lowest setting on the machine, a scalpel, bamboo and compressed air. Found this to be rapid and gentle, and they brush consolidated with 5% Butvar in IMS as they went. When pieces came off (the willow had some intact bark) they used 2-3% Klucel G. On the interior, they did leave some crusts in place if they were too tenacious. Jim Spriggs comments that he’s had some success with vacuum tweezer in these situations, too. Someone asked if the accretion could have been a clay meant to be there, but it seems the entire matrix was this material, so no, it was not meant to be part of the basket.
Ship Caulking – the “Leftovers” from the Ship Conservation Projects (poster)
Anette Hjelm Petersen
Caulking between ship planks made of wool soaked in tar is removed during conservation in order to treat the wood as well as get a good look at the tool marks. But there might be interesting information in that caulking material. In the period 1015-1450, textile are rare in Danish sites. When was the caulking material unspun yarn and when was it textile? Are there differences in the caulking material between warships and merchant ships? She is exploring these issues.
Rehabilitation and Monitoring of Waterlogged Archaeobotanical Remains from Southwest Florida’s Pineland Complex.
Donna L. Ruhl
Unfortunately, there was no presenter for this poster. The abstract involved a site excavated from 1988-1995, dated 50-1710AD. The archaeobotanical collection including things like seeds and wood is being curated at the Florida Museum with some grant funding from the National Endowment for the Humanities. Because this was the last day of the conference and I didn’t know there was no presenter, I missed the chance to look at the poster.
Conservation of a Marine Composite (Copper/Textile) from the 19th Century Shipwreck “Patris” in Greece
P. Patsiri, C. Margariti, S. Rapiti
The Patris was a wheel steamship sunk in 1868 in the Aegean sean, with its two halves at different depths. It was recovered in 2006. The object being treated was a firehose fitting, where the hose attaches to a hydrant. Calcareous accretions were removed with 10% formic acid on a local compress, and they also had 1% Hostacor in deionized water. The fiber ID indicated the hose material was flax or hemp with a little bit of cotton. There were also wood traces, but not enough to ID. The hose also suffered from active corrosion as bronze disease and secondary deposition of copper as newly formed cuprite. Sodium sesquicarbinate was used to treat the copper alloy.
Check out additional posts on the business meeting, specific WOAM personalities, the flavor of WOAM, and Lars Andersen’s advice to me on freeze drying at the AIC’s news blog.