AIC’s 45th Annual Meeting – Book & Paper + Research & Technical Studies group talks, May 31, 2017 – “Centuries of Cellulose: Lessons Learned from the Molecular Size of Cellulose in Naturally-Aged Paper Collections” by Dr. Andrew Davis


Andrew began his talk by very graciously acknowledging that many other people have contributed data that informed his paper. Andrew’s work is based on research began by William Barrow, a paper chemist at the Library of Congress until the 1960’s. Barrow’s research on books tried to draw a connection between physical properties and chemical content. He had collected about 1000 books published between 1500 and the 19th century, and he took various measurements such as fold endurance, pH, alum content, etc. He tried to draw connections between those sets of data to predict the ageing characteristics of the paper. This collection was obtained by LoC in the 1970s, and are still used for destructive testing today. But where Barrow used macro and micro scale measurements, Andrew looks to the middle ground: polymer chemistry. For that, he uses size exclusion chromatography, or SEC.

SEC measures the degree of polymerization of cellulose using a roughly 1mm squared sample size. (It may be helpful to think of degree of polymerization as the molecular weight.) The degree of polymerization of a sample can be compared to known references. It should also be noted that papers have a mixture of molecules of different sizes, and SEC provides a distribution curve. The more large molecules in a sample, the less degraded the cellulose, meaning that the paper is in better condition. Andrew discussed several examples of treatments of iron gall ink on paper where SEC was used to show the effects of those treatments on the papers.

Barrow’s research indicated that pH was the best indication of the future physical properties of paper. Andrew took about 80 samples from Barrow’s collection and confirmed that the molecular weights of paper correlate with pH (when the pH drops, the molecular weight drops). Andrew then looked to see if the molecular weight corresponded to physical properties; with newer papers, the molecular weight does tend to be smaller. Poor tear resistance also corresponds to low molecular weight. In general, he found that the molecular weight determined by SEC is a better indicator than pH for future physical properties for both newer and older books.

SEC certainly has advantages. The sample size is ridiculously small. Tells you about the physical building blocks of the paper, giving a better idea of what’s in it and what state the cellulose is in. There are some disadvantages to overcome before this technique is in every lab. The test itself takes a week to do. It requires extremely expensive equipment and organic solvents, and one must have the technical knowledge to interpret the data. Andrew’s ultimate goal is to turn this into a rapid technique that’s affordable, so that the molecular weight distributions of an object can be included in an object’s record and be pulled up by a barcode. That’s an exciting prospect!

Andrew’s work presents a very interesting analytical option that future conservators might have access to. It would be nice to have a predicting model for the degradation of library objects. But it would be even more interesting to see the effects of treatment on paper. It is important that conservators continue to check our own work, and I’m glad to have assistance with that from scientists like Andrew.

AIC’s 45th Annual Meeting – Book & Paper + Research & Technical Studies group talks, May 31, 2017 – “Revisiting paper pH determination: 40 years of evolving practice in the Library of Congress Preservation Research and Testing Laboratory” by Cindy Connelly Ryan

Cindy’s talk was a mightily condensed summary a few of the techniques for measuring the pH of paper that the Preservation Research and Testing Division (PRTD) at the Library of Congress has investigated over the last 4 decades. Her introduction was a summary of the challenges presented by this task. Due to the chemical structure of cellulose and the nature of paper, most methods can only approximate the pH of paper. The method of sample preparation can impact the results of measuring. How paper ages means that there may be a different pH in different regions. The ions that dictate the pH may not be soluble in water, making measuring pH harder. And atmospheric carbon dioxide can react with your solution and affect your results. Notice that I said “solution.” Cindy ended her introduction by noting that you can’t measure the pH of a solid. But you can approximate it, and the PRTD has been trying to identify the best way to do this for decades.

The PRTD’s focus on the pH of paper began in 1971 with the deacidification program. Chemist George B. Kelly used titrated alkalinity and titrated acidity as an “external yardstick”, and four different extraction methods: TAPPI cold method, TAPPI hot method, surface pH measurement, and “pulped” pH method. Kelly determined that for acidic papers, the method of measuring pH didn’t matter much, but the alkaline papers had an acidic pH despite their 1% alkaline reserve. The hot extraction method was shown to be much more accurate with alkaline papers, as it was likely better at getting all the ions into solution. The pulp method came close. Cindy then went on to talk about the uncertain origins of the pulp method (i.e. it’s not discussed in any published literature, but is mentioned frequently in internal documents from the PRTD). (I do wish that Cindy had gone into detail about the process of each method of extraction, because I wasn’t too sure about how each process worked. She doesn’t mention pH strips, gels, or pH meters at all in this talk. And TAPPI stands for the Technical Association of the Pulp and Paper Industry.)

Then Cindy skipped to the late 1990’s (she does mention that a few papers had been published in the decades between). By this time, the PRTD ramps up its documentation efforts, as well as its protocols for sample collection and homogenization. Most of these protocols were put on their website. During the renovation of the instrumental suite in 2007, the lab’s emphasis shifted to developing non-destructive and micro-invasive techniques, which were more appropriate for art objects rather than circulating collections materials. This meant that the sampling methods had to adjust accordingly.

To address the new challenge of micro samples (or none at all), the PRTD tried to make surface pH measurement work, but found that tideline formation and sensitive media made that difficult. “Miniaturization” was another method the PRTD tried. For this technique, sample size can be a few milligrams to a few micrograms, depending on the paper to water ratio and other details of sample preparation. They found that slurrying helps, but filtration makes no difference in pH measurement.  In addition, controlling the amount of carbon dioxide was key to getting an accurate reading with acidic papers. Both purged bags and sealed vials were tried, with comparable standard deviations but slightly different pH readings. The pH readings from the micro methods agreed fairly well with macro methods.

One of the best takeaways from Cindy’s talk was when she shared that during their renovation, the PRTD was sending out samples to a contractor for pH measurement, and papers that had an alkaline reserve were coming back with an acidic pH. Their conclusion was that not every method is appropriate for every type of paper, and that sample preparation can also affect results.

Here were some tips about each of the methods, taken from Cindy’s recap slide: The Hot Method is closest to objective measurement, but takes two hours per sample. The Pulped/Blender method generally agrees with the hot method, but is faster. The ISO cold method has a much higher standard deviation than the TAPPI cold method. Thea Surface pH method has the highest standard deviation of any method tested, and is difficult with alkaline papers, thick boards, and boards with adhesives. This method also causes tidelines. And the Mini Method is also difficult with thick boards, but the results comparable in repeatability to large scale extraction methods.

So, what does it take to accurately measure the pH of a piece of paper? A focus on repeatability and an optimistic attitude! The scientists and preservation specialists at the PRTD struggle with many of the same challenges that the rest of us do, albeit with fancier equipment. It sounds like just getting a ballpark figure for pH is as close as we can hope for for now. The PRTD is still investigating methods, and we should all look forward to their results!

Finally, one cool tip: You can make your own micro blender with a homemade Mylar blade attached to a Dremel tool!

42nd Annual Meeting – Photographic Materials Session, May 31, “Technical Research on The Diane Arbus Archive” by Janka Krizanova

Janka Krizanova’s fascinating talk on her work with the Diane Arbus Archives at the Metropolitan Museum of Art in New York City provided an overview of the first eight months of her two-year-long research scholarship. The Diane Arbus Archives contains 600 vintage prints, 800 work prints, 6,200 rolls of film, 6,500 contact sheets, and the artist’s library and equipment. Krizanova’s goals for the project are to characterize the materials in the archive, assess the stability of materials in the collection, create a long term plan for the preservation of the archives, and create guidelines for the exhibition of materials in the collection.
In her research, Krizanova examined:
• technical and historical literature on 20th century photography
• manuals and books of samples of photographic papers issued by the industry (in addition to Paul Messier’s Historic Photographic Papers Collection)
• other characterization studies, such as that on the Thomas Walther Collection at the Museum of Modern Art
Krizanova began her study by conducting a survey of the collection. She found that Arbus’ body of contact prints had the widest variety of photographic processes and the most varied condition states. Arbus worked primarily with silver gelatin prints, but also utilized the stabilization process for contact sheets and temporary proofs, as these prints were much faster to process than silver gelatin prints. However, they were not designed to be long lasting.

An image from Krizanova’s talk: an advertisement for “The Kodak Ektamatic 214 Processor”. The image was scanned from a postcard, purchased by Krizanova from:,0224814815,language,E.html

Stabilization prints are made by using a special photographic paper with an incorporated developer. Arbus used Kodak Ektamatic Paper (boxes of which are housed in the archives). A negative in an enlarger is used to expose the paper, which is then fed into the processer. It first passes through an alkali bath, which activates the developer in the paper. The paper then moves directly into an acidic stabilization solution, which complexes the unexposed silver. This silver is “stabilized”, but not fixed. The whole process is over in a matter of seconds. In addition to describing this photographic process, Krizanova discussed the intriguing condition issues seen in the collection, including spotting, darkening, lightening caused by applied pressure, and discoloration even when stored in an ambient room environment. Krizanova is working on establishing a set of terms to describe the condition issues presented.
The technical characterization of some of the silver gelatin and stabilization prints will involve:
• Measuring paper thickness
• Documenting printing on the verso
• Microscopic documentation of the surface texture
• XRF in the D-max and D-min areas
• UV examination
• Paper fiber sampling
• Spectrophotometric measurements
• Microfadometer readings
Krisanova has begun the first four categories of characterization. I really look forward to hearing the results of her work and her characterizations of the interesting and complicated condition issues seen in the stabilization prints.
Here are two excellent questions asked (and answered) at the end, loosely paraphrased:
Q: Will you fix and wash the stabilization prints?
A: No, they are to be preserved as is.
Q: Would you consider freezing the stabilization prints in an attempt to preserve them?
A: No one is doing that right now, as far as anyone knows. Currently, it is not an option for us, but we will certainly address this question again later in the study.