45th Annual Meeting – Objects Session, Wednesday 31 May 2017, “Archaeological Glass Conservation: Comparative approaches & practicalities of using acrylic resin films as gap fills” presented by Jan Cutajar and Hana Bristow

Jan Cutajar is Research Assistant at the University College of London, and Hana Bristow is Assistant Conservator at the National Museum of the Royal Navy, Portsmouth. They jointly presented on their experiences making acrylic resin fills with Paraloid B-72, an acrylic co-polymer, for use with glass repair. They based their work on techniques previously established by Steve Koob from the Corning Museum of Glass and recently updated at the CCI Symposium in 2011 (See here for the last update: https://www.cci-icc.gc.ca/discovercci-decouvriricc/PDFs/Paper%2035%20-%20Koob%20et%20al.%20-%20English.pdf).


Cutajar and Bristow treated two archaeological glass vessels: one Sassanian glass from UCL, and one beaker from Exeter, as case studies. They had common goals of needing reconstruction, stabilization, and the ability to be studied. They also had similar physical characteristics of degraded but stable glass with relatively good contact between the extant shards, thin walls (as thin as 0.2 mm), and substantive loss around of 35%.

Detail of resin fill in place on the Exeter glass jug (300/1988/G155, Royal Albert Memorial Museum, 2015). Photo by Hana Bristow

In looking for a fill system, they wanted a material that could reinforce weak areas, aid in practical assembly, and be as minimally interventive as possible. Since working with epoxy would require considerable manipulations with the artifact for both direct and indirect casting methods, this was not chosen. Instead, acrylic resin fills were explored because they are lightweight, strong, flexible, thin, detachable for future retreatment, simple to produce and insert, and can be manipulated for color and opacity matching.

Detail of resin tab application on Sassanid glass vessel (6300, UCL Conservation Teaching Collection, 2015), highlighted on the left and blended against the glass on the right. Photo by Jan Cutajar.

Koob’s technique is based on 30% w/v Paraloid B-72, an acrylic co-polymer, in acetone, with ethanol added to slow the evaporation rate, thereby reducing the potential for bubble formation. For coloring, ground pigment can be added to the ethanol before adding it to the resin mixture. The pigmented ethanol should be first decanted to prevent larger pigment particles from being added. The solution is cast and stored in a partially sealed environment for slow evaporation. Bristow felt that B-72 alone was too flexible, so she explored resin mixtures and tested varied proportions using B-72, B-48N or B-44 either straight or mixed in 2:1 ratios but always 30% in acetone. She also tested these opacifiers: fumed silica, marble dust, titanium dioxide and whiting. She cast the test resin mixtures in boxes of the same size, also holding the volume and concentration of the solution and the volume of added ethanol constant. The tests were evaluated after 4-5 days of curing for hardness, plasticity, and appearance. She found that a 2: 1 solution of B-72: B-48N produces a strong film without brittleness. This film was stronger than the B-72 film and not brittle like the B-48N alone or B-44 films.

The resin films are set to cure in a partially sealed solvent atmosphere. Photo by Hana Bristow.

For the opacifiers, Bristow found that fumed silica worked well for adding translucency, and marble for opacity, but whiting and titanium dioxide produced speckled results and were difficult to homogenize with the mixture. She also notes that dry artist pigments are good for tinting but shouldn’t be relied on for opacity as well, because they easily over saturate the mixture, resulting in a cracked and weakened cast. She recommends a maximum of 1.5 micro-spatula scoops per 30 mL resin mix.


Cutajar and Bristow offer some practical notes and tips:

  • Achieving desired film thickness can require some trial and error. Expect about 70% volume shrinkage.
  • Trays should be non-absorbent and easily release the resin. Making or using solvent-resistant boxes lined with release papers or films works well.
  • Enclose the poured resin trays in an acetone rich environment to slow the rate of evaporation. This will help prevent bubble formation.
  • Films should set for at least 4-5 days before removing, otherwise the films are too flimsy for these applications.
  • Films are best to manipulate directly after demolding. Things that can be done are:
    • Texturing
    • Shaping
    • Cutting – determine the size by taking a tracing of the loss area
  • Shape can be adjusted using heat; about 20 seconds under a hair dryer works well. Once warm, hold the cast in the desired position until it cools enough to hold the new shape. This can be done through repeated heating and cooling cycles until desired shape is achieved.
  • Adapt a cast by creating a lip at the edges where joins are very thin. This creates a slight overlap with the adjacent glass. The lip can be created with a heated spatula away from the glass.
  • Bonding can be activated with acetone, but Cutajar and Bristow suggest using more adhesive (Paraloid B-72) to make the join since acetone can compromise a good fit.
  • The film can be cut into tabs and used as reinforcements across joins. Cut the tabs into shape, lay them across the join, and activate with solvent. The tabs are virtually invisible.
  • The film can be used to make recessed fills for backing thin, curved glass, providing local stabilization and weight redistribution.
  • Backing films are easier to apply when they are freshly removed from the solvent atmosphere and retain a slight tack.