JAIC , Volume 39, Number 3, Article 1 (pp. to )
JAIC online
Journal of the American Institute for Conservation
JAIC , Volume 39, Number 3, Article 1 (pp. to )




To date, three analytical studies have been published about paper splitting as a method of strengthening paper (Liers et al. 1996; Vilmont et al. 1996a, 1996b; Bansa and Ishii 1997). Paper splitting has been compared with lamination, resizing, leaf-casting, and Parylene impregnation. Physical strength has usually been the most important parameter investigated, but other factors such as appearance and reversibility of the strengthening process have also been considered. Two scientific research projects are still in progress (Galinsky 2000; Kolbe and Banik 2000).


Commissioned by the Bibliotheque Nationale in Paris, the Centre du Papier at the University of Grenoble conducted a study on paper lamination and splitting. The laminate was formed using a polyamide or acrylic adhesive to adhere a paper tissue or polyamide fiber material to one side of the original. Paper splitting was carried out manually under the supervision of Wächter at the Deutsche Bücherei (Vilmont et al. 1996a, 1996b). Three types of core papers were tested: Japanese kozo, and both bleached and unbleached cotton fiber tissues. Two types of modern paper were strengthened by one or the other of these two methods. They were tested before and after artificial aging for their mechanical and optical qualities and the reversibility of the treatments.

When compared to their nonstrengthened, artificially aged counterparts, the strengthened papers showed the following results: paper treated by splitting showed less average thickness increase (55%) than laminated paper (146%); the average tensile strength increase (machine direction) of split paper (125%) surpassed that of laminated paper (74%); the average bursting strength increase was virtually equal for split paper (147%) and laminated paper (143%); only the average folding endurance increase (machine direction) was markedly better for laminated paper (370,927%) when compared to split paper (4,304%). Overall, Vilmont et al. (1996a, 1996b) favored splitting over lamination.

Another study showed that splitting increases the pH of paper by around 3.5 units, and the tensile strength (TS) of resized paper is lower than that of paper strengthened by splitting, especially with regard to degraded, weak paper (Liers et al. 1996).

Helmut Bansa and Ritsuko Ishii (1997) tested six different papers, including degraded historic samples. They used three core adhesives for manual splitting carried out at the Bavarian State Library and compared splitting with leaf-casting, lamination, and Parylene impregnation. The samples were strengthened either before or after artificial aging. The TS of all of the pre-aged, strengthened papers increased markedly when compared to nonstrengthened, aged controls (=100%): split papers with a methyl cellulose core adhesive had the highest average TS increase (329%); papers that received paste as core adhesive showed a TS increase of 172%; papers that received an acrylic emulsion as core adhesive showed a TS increase of 152%; leaf-casting (154% TS increase) and lamination (170% TS increase) compared favorably with splitting; Parylene was less satisfactory (111% TS increase) and had distinctly negative effects on some papers.


Vilmont et al. (1996a, 1996b) found that during artificial aging, papers strengthened by splitting suffered a greater loss of whiteness than papers sealed with a laminate; this phenomenon was not explained. Lamination, however, altered the appearance of the documents and made them less easily legible than the split documents. Vilmont et al. (1996a, 1996b) favored paper splitting over lamination for its more pleasing appearance overall. Bansa and Ishii (1997) observed that none of the four tested treatments—leaf-casting, splitting, lamination, and Parylene coating—were uniformly accepted by a group of evaluators as the one most aesthetically pleasing conservation method; leaf-casting was slightly favored.


To date, only the French team has investigated in detail whether the paper strengthened by splitting later can be separated from its core paper (Vilmont et al. 1996a, 1996b). It found that splitting could be far more easily reversed than any of the lamination processes tested. Soaking the strengthened paper for ca. 10 minutes in water released the two split paper halves from the core paper. The papers laminated with polyamide adhesives required several hours of soaking in organic solvent baths. Höge (1981) reversed splitting treatments by immersing the strengthened paper in a solution of 75% ethanol and 25% water. Gast (1993) and Mowery (1999) also observed that paper strengthened by splitting could be separated again. Wächter maintains that the process can be repeated by refacing the paper, opening the existing split, and replacing the existing core sheet with a new one (Wächter 1999b). It will be interesting to see emergency treatment directives for strengthened paper that has suffered extensive water exposure. Salvage and treatment of water-damaged papers that were strengthened by splitting will require special considerations.


One of the most difficult questions discussed in paper-splitting concerns the preservation of deteriorated iron gall ink manuscripts. In a recent survey of treatments performed at the Institut für Erhaltung von Archiv- und Bibliotheksgut in Ludwigsburg, conservation scientist Anna Haberditzl (1999) considered paper splitting to be an effective salvage treatment for papers suffering from the effects of severe iron gall ink corrosion. Manuscripts that had been stabilized through splitting a few years earlier still exhibited neutral to alkaline pH values, were free of detrimental ferrous ions within the ink lines, and under UV radiation exhibited no fluorescent halos around ink lines, the presence of which can indicate the onset of cellulose degradation. In their recent investigation, Gesa Kolbe and Gerhard Banik agreed that paper-splitting can “chemically inhibit iron gall ink corrosion” (Kolbe and Banik 2000). However, severely deteriorated iron gall ink manuscripts often do not split uniformly and incur substrate losses during the removal of the gelatin-coated facing papers. In the area of the ink application, the embrittled paper remains attached to the facing papers instead of adhering to the core paper. Therefore, paper-splitting “has to be considered a risky procedure … in the treatment of iron gall ink corrosion” (Kolbe and Banik 2000). Robert Fuchs et al. (2000) also observed that while blank areas of 18th-century manuscript pages split in the center, deteriorated ink lines did not undergo splitting and ended up in their entirety on one side of the core tissue.

There has been some speculation regarding the reputed benefit of the gelatin layer brought into contact with the surface of the iron gall ink during splitting. The gelatin is believed to immobilize metallic ions in the ink, an effect that partially fixes the ink and prevents it from bleeding during final washing when the facing supports are released and acids in the paper are rinsed out (Müller 1999b; Wächter 1999b). Kolbe and Banik (1999) investigated the chemical interaction of gelatin with iron gall ink and proved that under certain conditions, the gelatin is able to bind free ferrous ions in the area of ink application, thereby inhibiting their detrimental catalytic activity. These reactions are dependent on a variety of factors including the pH conditions, the reaction time, and the protein-metal ion ratio. It is just this otherwise beneficial interaction between gelatin and deteriorated iron gall inks that complicates the removal of the facing papers and causes losses along the ink lines during paper-splitting.


In the treatment of objects that are not water-sensitive, the paper-splitting process concludes with extensive rinsing cycles that allow the deactivation and subsequent removal of the industrial enzymes that are used to release the facing papers. In the treatment of water-sensitive objects such as iron gall ink manuscripts, however, the aqueous procedures are less extensive to avoid alterations in the appearance of the media. Whether in this case degraded gelatin residues or deactivated enzymes remain on the paper surface has not yet been clarified. A systematic analysis of the paper-splitting process is currently being conducted by Eva Galinsky (2000) at the Institut für Erhaltung von Archiv- und Bibliotheksgut in Ludwigsburg with funding from the Deutsche Forschungsgemeinschaft. In cooperation with the Physiological-Chemical Institute of the University at Tübingen, Galinsky seeks to introduce immobilized enzymes that can be permanently anchored on reusable facing papers. This procedure will establish better control of the enzymes during treatment and will improve their cost-effectiveness.