45th Annual Meeting – Workshop, May, 29, 2017, “Effectively Using Portable IR and Raman Instruments for Art Object Analysis,” presented by Francesca Casadio and Tom Tague

My personal area of interest and intended future practice is in the conservation of historic interiors. Therefore, I am always keen on portability both in tools and materials as well as forms of analysis. The other advantage to the techniques presented in this workshop is that physical sampling is not required, which is always attractive and music to a curator’s ears.

The workshop met my personal expectations, but the title “Effectively Using…” could have suggested to some that this was going to be more of a “boot camp” for being able to implement these techniques back home. This style of workshop was more of an information/demonstration session and is great for anyone considering buying similar instrumentation and/or for gaining a better understanding of the general benefits and limitations of portable spectroscopy.

Given the short duration of this workshop, I was initially concerned that I might have signed up for a 2 ½ hour lecture without any hands-on component. Participants were encouraged to bring our own samples and indeed at least an hour was dedicated to looking at samples and exploring the instrumentation first-hand. Although we did run over the scheduled time, and were gently shuffled out of the room as hotel staff started to break down tables.

The workshop was led by Tom Tague, Ph.D. Applications Manager at Bruker, and Dr. Francesca Casadio, Director of the Conservation Science department at the Art Institute of Chicago. I really appreciated having these different perspectives. Tague did not assume the role of salesperson during the workshop, but as you would expect he was very positive in his description of the capabilities of the Bruker instrumentation. Casadio kept Tague grounded in the realities of our complex samples and what can be confidently identified using these techniques. At the same time, it was useful to have Tague there to speak to the specifics of the instrumentation and push Casadio a little bit to consider what some of the newer technology could offer. There was also a Bruker sales representative present to assist with running the instrumentation and software and offer information on pricing.

Overall the session was well organized. I know I was not the only attendee who was ecstatic that I got to take home a flash drive loaded with the presenters’ PowerPoint slides. The spectra from my samples that were analyzed were also loaded directly onto this flash drive before the end of the workshop.

The first part of the session did consist of pure lecture. Tague’s presentation focused on specifications of the Bruker portable instruments and descriptions of the techniques.

An interesting tip he offered was using sandpaper to take surface samples. He lightly abraded a painted surface and then placed the sandpaper in front of the portable FTIR (ALPHA)—no additional sample prep necessary.

Having just completed my Master’s degree in conservation I was able to follow the presentation fairly well, but I fear that it may have been overly technical and too fast for someone who does not work with these analytical techniques on a regular basis. Nonetheless, I anticipated this to be an intermediate-level workshop when I signed-up.

Tom Tague in front of demo table at the workshop. He has rotated the ALPHA in front of him so that the laser beam points downward and is working to raise the ALPHA up on some shims in order to fit the sample under the beam at the correct working distance.

 

As would be expected based on the organizers of the workshop, the instrumentation provided and discussed were all Bruker models. Two ALPHA portable FTIR spectrometers were present. The ALPHA is set up to receive different “snap-on” modules. The two modules available for demonstration were the “External Reflectance” module and the “Platinum ATR” module. The BRAVO Handheld Raman spectrometer was also available for interaction.

 

Here are some key facts about each instrument:

 

The base ALPHA starts around $14,000 and each module is on average $6,000 in addition.

 

 

 

ALPHA “External Reflectance”

  • Does not require direct contact with a sample/object
  • No size limitations as long as unit can be mounted/held in appropriate orientation to the sample

    ALPHA “External Reflectance” being used to analyze paper currency. This example was given in Tom Tague’s presentation.
  • Camera integrated in unit to help orient, find appropriate working distance/focus, and document sample location
  • Collects reflectance spectrum NOT absorbance
    • Can collect specular and diffuse reflection; reflective and non-reflective materials can be analyzed
  • Footprint of instrument is about 8” X 11”
  • Weighs about 13lbs.
  • Can be tethered to a laptop
  • About 6mm sampling area
  • Approximately 4cm-1 spectral resolution

ALPHA “Platinum ATR”

  • There is pressure/direct contact with the sample
  • The IR beam does penetrate into the sample

BRAVO Handheld Raman

  • $45,000-$55,000
  • Slightly narrower than 8” X 11” (looks like an oversized ELSEC environmental data monitor; less heavy than the Alpha)
  • Class I safe laser
  • 2mm sampling spot size
    • No camera or viewing capability to help align collection area
  • Object needs to be in contact, but no pressure required
  • Approximately 8cm-1 spectral resolution
  • Fluorescence mitigation built into software/data collection
  • Dual lasers built in and used/activated simultaneously
    • Optimal wavelength and reduced risk of damaging sample
  • Touch screen allows for control and data collection without tethering to laptop
    • Tethering also capable via WiFi to laptop
Example from Tom Tague’s presentation of the BRAVO being used to analyze medieval manuscripts at the Morgan Library and Museum.

 

In terms of the ALPHA “External Reflectance” one of the big selling points is that there is no size restriction or need to balance the object on a stage. The trade-off in allowing data collection without physical sampling is that the spectra generated are in % reflectance. The majority of reference spectra available for free and through the Infrared and Raman Users Group (IRUG) are % absorbance or % transmittance (its inverse). The Bruker software does offer the capability to convert the data using the Kramers-Kronig Transformation. Francesca Casadio seemed to prefer to analyze data from its original state in reflectance. Characteristic peaks for bonds are slightly shifted from their location in transmittance spectra, but at Casadio’s level of experience she is able to take these nuances into account with some ease. She was honest with the attendees summarizing that this form of IR spectroscopy is “not like portable XRF; one needs to have experience and repetition for familiarity with interpreting spectra.”

For those interested in more on interpreting reflectance spectra of art objects Casadio recommended the following publications from a research group in Perugia Italy:

“Reflection infrared spectroscopy for the non-invasive in situ study of artists’ pigments.” C. Miliani, F. Rosi, A. Daveri & B. Brunetti, Appl. Phys. Mater. Sci. Process. 106, 295–307 (2012) (http://dx.doi.org/10.1007/s00339-011-6708-2)
“In Situ Noninvasive Study of Artworks: The MOLAB Multitechnique Approach.” C. Miliani, F. Rosi, B.G. Brunetti & A. Sgamellotti, Acc. Chem. Res. 43, 728–738 (2010) (http://dx.doi.org/10.1021/ar100010t)
“Non-invasive identification of metal-oxalate complexes on polychrome artwork surfaces by reflection mid-infrared spectroscopy.” L. Monico, F. Rosi, C. Miliani, A. Daveri & B.G. Brunetti, Spectrochim. Acta Part -Mol. Biomol. Spectrosc. 116, 270–280 (2013)
“In-situ identification of copper-based green pigments on paintings and manuscripts by reflection FTIR.” D. Buti, F. Rosi, B.G. Brunetti & C. Miliani, Anal. Bioanal. Chem. 405, 2699–2711 (2013)

 

It is important to keep in mind the basis of data collection to understand the limitations of what can be analyzed with the ALPHA “External Reflectance” on a given object. For example, with a varnished painting the spectral reflectance of the varnish will typically only allow the varnish itself to be detected (with some exceptions depending on thickness of the varnish and underlying pigment composition). Similar reflective material properties make plastics easily detectable with this technique. Matte objects are still good candidates for analysis with the ALPHA, but the data will be collected via diffuse reflection. The ALPHA does not seem like an appropriate technique for discerning between individual layers within a given structure unless coupled with other techniques.

One of the ALPHA’s at the workshop was supplied by Casadio from the Art Institute’s lab, and she has extensive experience using the ALPHA. Her presentation was more about working knowledge of the instrumentation. She polled the attendees and focused on case studies mainly of pigment analysis and identification of plastics. Casadio emphasized the benefit of the ALPHA as a mapping tool that does not require sampling. Perhaps one or two samples could be taken from a work of art and more confidently characterized with bench top FTIR and/or GC-MS and then the use of specific materials could be mapped without additional sampling using the ALPHA. Casadio’s case studies often combined multiple analytical techniques. She finds the ALPHA to be a nice compliment to XRF. Overall, Casadio has found the ALPHA to be very useful in characterizing different plastics and also good at detecting deterioration surface products (e.g. zinc soaps) especially with modern and contemporary collections. Casadio noted that the ALPHA detects very strong signal and peaks for waxes and PVA coatings. Casadio has been able to use the ALPHA for collaborations with other institutions and collections, which is another boon of its portability.

I was disappointed that Casadio had not had previous experience with the BRAVO Handheld Raman. At the Art Institute she has a bench top Raman unit. She seemed skeptical about the BRAVO’s capabilities and some of the claims that Tague was making that it could “see” indigo and other organic pigments without surface enhanced Raman spectroscopy (SERS). Casadio stated that in her personal opinion with Raman it is better to bring the art to the unit than the other way around. By the end of the workshop she did seem impressed with the quality of spectra the BRAVO was generating, but there was not enough time to have further discussion and to tease out Casadio’s candid opinion on the instrument.

I was most excited for the practical demonstration with the instruments especially because I had come armed with over 10 samples. I was anticipating that I may not even get to analyze one sample, but was very pleased that I was able to look at 7 samples with the BRAVO portable Raman. This much time with the instrument was due in part to many participants not bringing samples.

If a similar workshop is organized in the future, it might be good to have participants sign up ahead of time for slots with the instrument if they are interested in analyzing a specific sample. It was a fairly large group – about 18 participants. Attendees that did not bring samples were still interested in watching the process of collecting data and interpreting the spectra. This was challenging; even with three instruments there tended to be 5-7 people crowding around a laptop screen. Dividing us into smaller groups, having the laptops hooked up to a projection screen, or further limiting the number of participants may be additional considerations for future workshops.

It seemed like the majority of participants were conservators rather than conservation scientists. I personally do not work with spectroscopic techniques on a regular enough basis to be able to confidently interpret spectra on the fly. Francesca Casadio was able to offer her expertise and interpretation while working with samples from the participants, but neither Tom Tague nor his Bruker colleague could offer specialized interpretation. Some of the participants seemed frustrated that the instruments were not connected to an art materials database for instant gratification and matching.

Both Tague and Casadio strongly emphasized the importance of each institution building its own reference database specific to the collection. The IRUG database was promoted, but as a supplement to an institution’s own reference database. Neither of the instructors felt that the database that comes with the Bruker software was appropriate for art materials.

My personal goal during the workshop was to pit these portable instruments against their stationary counterparts and to pit the two complimentary techniques against each other. Therefore, I brought known samples from my institution’s reference collection of traditional paints. All the paints were oil-based and mixed with some degree of lead white. The reference pigments I chose were mostly organics (indigo, madder, cochineal). Colonial Williamsburg has had the opportunity to partner with the College of William and Mary in order to perform SERS on objects in the paintings collection. My colleagues and I were curious to see how this portable unit compared to spectra produced with SERS. With the minimal time, I chose to focus on the BRAVO because our institution already has a bench top FTIR.

Tom Tague was set-up at the BRAVO “station” during the practical session, and as I stated previously he was not comfortable offering any interpretation of the data. I was excited to review the spectra we collected back at my home institution (Colonial Williamsburg Foundation/CWF) alongside Kirsten Travers Moffitt, the conservator in charge of our materials analysis lab. Moffitt performs a lot of FTIR analysis on our collection, but has less experience with Raman.

All the organic paint spectra from the BRAVO were certainly “neater” than what I am used to seeing in terms of raw data from a bench top Raman with oil paint samples. I personally would attribute the quality of the spectra to the dual laser capability. I’m not sure how much impact the fluorescence mitigation had because the spectra were still pretty noisy and it was challenging even for Moffitt to distinguish significant peaks. It appears that the fluorescence of an oil binder is still problematic with the BRAVO. In Tague’s presentation he showed an example of indigo detection with the BRAVO, but this was on an illuminated manuscript, where fluorescence of the binding media would be less of an issue.

At CWF we only have a reference database for IR spectra, but looking at possible peaks in the indigo/lead white sample spectrum, the characteristic peaks for indigo that Tague mentioned (545, 1463, 1578) do not appear to be present. It seems that the lead white is dominant, with a strong peak around 1050. In conclusion, Tague is partially right that the BRAVO can detect some organic pigments, but likely only if they are present in high enough concentrations (not mixed) and are not in highly fluorescent binding media (like oil).

Spectra (representing a shorter and longer collection time) from the BRAVO of the CWF reference of indigo and lead white in oil.tic peaks for indigo that Tague mentioned (545, 1463, 1578) do not appear to be present. It seems that the lead white is dominant, with a strong peak around 1050. In conclusion, Tague is partially right that the BRAVO can detect some organic pigments, but likely only if they are present in high enough concentrations (not mixed) and are not in highly fluorescent binding media (like oil).

The other samples I looked at were reproduction wallpaper samples from Adelphi. I was curious to see if we could detect anything useful about the pigments on an object that would normally be challenging to sample and could not be brought to the lab if it were installed in a historic interior.

The resulting spectra were less noisy than those of the oil paint reference samples, again likely due to the non-oil binding medium on the wallpaper.

 

 

Spectra collected from two different greens on a wallpaper reproduction from Adelphi using the Bruker BRAVO. Peaks are labeled, but a match has not been identified.

 

 

Despite the better quality of the spectra, we still did not have the resources (i.e. a good reference database for Raman and experience working with Raman spectra) to confidently characterize the pigments present. I am sharing this to illustrate Casadio’s point that the ALPHA and BRAVO require a certain level of expertise and do not provide instant answers.

One of the other participants, Ann Getts, a textile conservator at the De Young Museum in San Francisco, brought various sequins from a costume in storage with a suspicious vinegar odor. Getts had time to look at one of the sequins with both ALPHA modules, and her case study demonstrates some of the trade-offs with the non-contact “External Reflectance” module.

She began with the “External Reflectance” module and the first hurdle was getting the instrument positioned at the appropriate working distance from the sample. Without an adjustable stand, we had to use trial and error to shim up the ALPHA so that the camera could focus on the sequin. The resulting spectrum suggested cellulose acetate (as suspected by Getts initially), but even Casadio still felt insecure about drawing any concrete conclusions based on this spectrum. Then the sequin was analyzed with the “Platinum ATR” module and right away Casadio concluded that indeed it was cellulose acetate.

Each of these instruments has their advantages and disadvantages. Overall the ALPHA seems like a good bang for your buck given the duality of the modules. The price point is pretty reasonable also considering the portability.

The BRAVO is fairly new technology and the dual lasers seem promising, but at this point it does not seem like a must have for the average institution. I would encourage anyone thinking about purchasing any of these instruments to consult with both of the workshop leaders.

Francesca Casadio using the ALPHA to analyze a Roy Lichtenstein painting installed on the wall.

 

In general I would specifically recommend the ALPHA to:

  • Institutions that have a lot of sampling restrictions
  • Institutions with a lot of oversized works
  • Institutions that focus on modern and contemporary art (especially with plastics and large Color Field paintings)
  • Institutions with a conservation scientist on staff

 

In general I would specifically recommend the BRAVO to:

  • Institutions that have a lot of sampling restrictions
  • Institutions wanting to focus on analysis of paper-based art
  • Institutions with a lot of oversized works
  • Institutions that already have staff with Raman expertise
  • Institutions looking to purchase a Raman instrument

This blog represents my personal conclusions and understanding of the workshop. I would encourage any of the other participants and the instructors to post in the comments if they have differing opinions or think that I have misunderstood any of the technical aspects of the instrumentation.

43rd Annual Meeting-Book and Paper Session, May 15, 2015, "16-17th Century Italian Chiaroscuro Woodcuts: Instrumental Analysis, Degradation and Conservation" by Linda Stiber Morenus, Charlotte Eng, Naoko Takahatake, and Diana Rambaldi

The presenter, Linda Stiber Morenus, began her discussion of these complex prints with a description of the printing process. Chiaroscuro woodcuts were intended to emulate chiaroscuro drawings, which were comprised of black chalk shadows and white chalk highlights on colored paper. Color oil-based printing inks were first used to print 14th-century textiles, being used on paper by the mid 15th-century. The chiaroscuro woodblock prints required two to five separate woodblocks, inked with different shades lighter and darker than the midtone colored paper.
In order to better characterize the media, Morenus collaborated with art historian Takahata, and conservation scientists Eng and Rimbaldi from the Los Angeles County Museum of Art (LACMA). In addition to prints at LACMA, the team studied prints from the British Museum and Library of Congress. Out of over 2000 surveyed woodcuts, 72 were studied in depth, with X-ray Fluorescence (XRF), Fiber Optic Reflectance Spectroscopy (FORS), and Raman spectroscopy. Inorganic compounds were indicated by XRF analysis. FORS was especially helpful for detection of indigo. Raman spectroscopy provided additional information about organic colorants.
Renaissance artists’ manuals, such as Cennino Cennini’s Libro dell’Arte guided the research by providing information on the most likely colorants for printing inks. Inorganic pigments included lamp black, lead white, ochres, vermillion, verdigris, and orpiment. Organic pigments included indigo and a variety of lake pigments.
After providing background information, the presenter began to focus on deterioration and conservation of the chiaroscuro prints. The prints from the Niccolo Vicentino workshop had a high lead content. The inks typically had a low vehicle-to-pigment ratio, tending to turn gray around the edges, due to the presence of lead sulphide. Verdigris corrosion was also a common problem, as found on “Christ Healing the Paralytic Man” by Giuseppe Niccolo Vicentino, as well as 13 other prints from the same workshop. Typical copper-induced paper degradation included yellow-brown halos around inked areas and cracks in the paper.
Fading and discoloration were major problems for the organic colorants, such as indigo and the yellow lakes. Morenus compared copies of Ugo da Carpi’s “Sybil Reading a Book” in the British Museum and the Library of Congress, finding clear evidence that the indigo in the British copy had faded. The British Museum had confirmed the presence of indigo through Raman spectroscopy. At least 8 of the prints were found through XRF to have high levels of calcium in the same areas where indigo had been identified, suggesting the presence of chalk-based lakes. Organic greens had shifted to blue or brown where organic yellows had faded or become discolored.
The presenter concluded with suggestions and caveats for conservation treatment. First, she advised conservators to exercise caution in aqueous treatment, in order the preserve the topography of the prints. The woodblock creates a relief impression in the paper, and the layering of the inks adds another level of texture that might be altered by humidification, flattening, washing, or lining treatments. The low binder content also makes the inks more vulnerable to saponification and loss during alkaline water washing. Morenus warned that the hydrogen peroxide color reversion treatment for darkened lead white would be particularly risky, because the white lead sulphate end product has a lower refractive index than basic lead carbonate original pigment. This means that treated lead white becomes more translucent, and the lower “hiding power” shifts the tonal balance of the print to appear darker overall.
For exhibit recommendations, Morenus suggested that we should always expect to find fugitive organic colorants in chiaroscuro prints, so exhibit rotations should be planned accordingly. Maximum exhibit conditions should be 5 foot-candles (50 lux) of visible light for 12 weeks of exposure, no more often than every three years. She also indicated that overmatting should be avoided to reduce the risk of differential discoloration.
During the Question and Answer period, Morenus clarified the color order used in printing. Some prints were inked from dark to light, but most were printed with the lightest color first.
I thoroughly enjoyed learning about these beautiful prints, but I think that the discussion of the lead white conversion treatment-induced refractive index shift was the most important “take-away” from the presentation.

39th Annual Meeting – Joint Paintings/Research and Technical Studies Session, June 3, “Raman Revealed: A Shared Internet Resource for the Cultural Heritage Community” by Suzanne Quillen Lomax

Suzanne Lomax presented on IRUG’s (Infrared and Raman Users Group) latest efforts to distribute data for Raman spectra.  She began the talk with a brief discussion on the history and mission of IRUG and their new initiative to create a Raman spectra database due in large part to a $239,650 two-year IMLS grant awarded to the Philadelphia Museum of Art in partnership with IRUG.    The 118 institutional members will contribute to the database which will be used by individuals, scientists, conservators, and students to study cultural heritage.  The Raman database will be maintained on their website.  IRUG has biennial conferences and their website www.irug.org contains information on grant funding and the conferences.  All of the coauthors for this paper are board members of IRUG.

Suzanne described the model for the database and compared it to the widely used infrared database.  By 2009, the IR database was 100% digitalized, available on CD, and in two print volumes.  The latest edition contains over 2,000 infrared spectra. On the current IRUG website, members are able to search terms and match by keyword resulting in a hit list for searched components.  The resulting spectra provide in their file name link the mode of collection and where it was collected.  The largest represented group in the IR database is organic dyes and pigments followed by mineral pigments.  Raman spectra are currently being collected and added to the database.

Suzanne also stressed the growing use of IR and Raman data use in the field and how this is being reflected in papers at IRUG conferences specifically related to art and archaeology.  She provided examples in which mineral pigments as well as synthetic organic pigments have been identified though used of the database and how Surface enhanced Raman spectroscopy (SERS) can be used to overcome the problems of fluorescence by using nanoparticles to magnify the signal.

Objectives of the IRUG database will include a website interface with the ability to upload data by users, software, a translator to transform native data into IRUG standard, a searchable library, an interface for keyword searches, data download, and spectra printing.

Suzanne is chair of the newly formed Raman review committee, which reviews spectra and format.  The format to be used by IRUG is JCAMP-DX (ASCII) files for universal access.  This will also allow batching of spectra for submission.  To learn more about the format refer to the IRUG website.

The first batch of spectra has been pledged but the invitation is open to new contributors.  Interested people should contact Suzanne or Beth Price, the project manager from the PMA.  Currently users cannot upload data but can do searches on the website.

A comment after the talk reminded the audience that it is a free database though users need to contribute 10 spectra to get access to the searchable version.