International Symposium on the Restoration of Chinese Coromandel-Lacquer-Screens at the Victoria & Albert-Museum, London

Dipl.-Rest. Irmela Breidenstein: "Damages and Conservation Methods of Chinese Export Lacquer Screens" (unpublished)

I am a free lance conservator since 15 years and I'm now working in my private workshop in Berlin. Since my studies at the University of Applied Sciences in Cologne I have been specialized in the conservation of European and Asian Lacquer. The main emphasis of my work lies in the restoration of what I call "europeanized" lacquer, that means asian export lacquer that has largely been altered and overpainted with european materials and japanned decoration.

1.Paravent Schloß
2. Falkenlust,
Chin. Kabinett
3. aufstehende Risse
 im  Lack
4. Röntgenaufnahme
5. Lackaufbau,

Some years ago I had the opportunity to do a research project on the conservation methods concerning a chinese gold- lacquer screen (1). It had been manufactured probably in Canton between 1705 and 1735. It measures 255 cm in heights and 264 cm in overall width. It is one of three complete six-part folding screens found at Schloß Wilhelmsthal near Kassel. The research on this project included a review on the eighteenth century European literature on the subject of the art of Chinese lacquer and compared the information given there to the lacquer technique employed on the screen.

The second part of the research was devoted to the causes of the damages and of suitable restoration methods. The knowledge and experience I won during the following restoration of the screen helped a lot when in the year 2000 I got the challenging task to execute the restoration in situ of the Lacquer room at Schloß Falkenlust near Bonn (2).
Built between the years 1733 and 1737 by the elector of Cologne, Clemens August, following the plans of Gilles Marie Oppenordt . Oppenordt also used parts taken from two Chinese laquer screens originating from Southern China around 1700 .The magnificent, formerly partitioned in twelve segments and gold-lacquered screen is the striking centre-piece of the middle wall ,whilst parts of a floral, multi-coloured paravent decorate the upper section of the wall and the window bays.

The most obvious, I would say nearly characteristic damage sustained by chinese export lacquer objects, mostly screens and furniture, consists in long parallel splits in the lacquer, running perpendicularly to the direction of the wood grain (3).

The split edges can rise very high in places, forming loose and extremely fragile lamina. Separation of the paint layer occurs above or below the more strongly adherent paper layer between the ground layers, but only rarely directly at the surface of the wooden support.

So the foremost aim of restoration measures on both the screen and the Lacquer Cabinet was to secure open cracks and splits, but above all to secure lifted areas as these are extremely subject to breakage. Open cracks, moreover, continually invite the penetration of moisture in unfavourable ambient conditions, like given in castles, which would only worsen the damage.

First of all I aimed my research on the potential causes of the cracking and lifting of the lacquer layers by examining the composition of the object by x-ray  (4). The screen panels consist of coniferous boards, the joints between them are clearly visible in oblique lighting. In order to determine structural details, an X-ray investigation was made of one panel by way of example. The x-rays revealed that the panel consists of three planks, glued together and reinforced by bamboo splits hammered into notches. The upper, end-grain edge was capped by a board, blunt-mitred at the corners and fitted with a longitudinal tongue, with was secured by means of a splint hammered in from the top.
Clearly visible is the radial grain of the boards, which, together with the overall protective lacquer coating has prevented the panel from warping. The numerous knots, splits and irregularities in the wood are quite visible, as are the gaps between the boards, which in places are quite wide. Defects in the wood were initially filled with putty, as the x-rays clearly indicated. After the surface was thus prepared, the first coat of primer was applied. It consists of a silisious filler, clays and red ochre. The vehicle is protein in nature and it contains a proportion of raw lacquer. D'Incarville says, that for cheaper groundings the Chinese would have taken pig's blood. A quotation of the Cho-keng-lu, 1366 AC says, that "for cheaper goods one takes a mixture of swines blood and rice glue and instead of a linen a hemp fabric".

The second layer comprises a sheet of quite thick, rough-fiber-paper, which in the cross-section of the sample (5) appears as a dark brown line in the primer layer. D'Incarville says in his work "Sur le vernis de la Chine", Paris 1760, that the paper was adhered to the surface with raw lacquer. A thicker layer of ground containing a higher proportion of red pigments follows. Two layers of slightly differing fluorescence, possibly due to different admixtures to the lacquer build the dark, originally highly glossy lacquer layer of the surface. The lacquer is not pigmented and no black colouring has been determined to this point. The depth of tint might have been achieved by means of thick application and burnishings. The composition of the décor too was subject to research but has no relevance for the consolidation of the delamination.

The same structure and lacquering method as described above is found in the gold lacquer panelling in Falkenlust. When we compare the results of our investigation of the Wilhelmsthal-Screen with other Chinese lacquerworks and the eighteenth century descriptions, a high degree of coincidence between the sources and the technical findings becomes apparent. The millenia-old art of Chinese lacquer is based on passed down traditional, time-tested and sophisticated techniques that have been common knowledge among lacquer artists for centuries. This explains why we can hardly expect to find divergences from the established methods and why we find common damages and the same alterations on different export lacquer objects up to the new export boom in the 19. Century.

6. Detail Paravent vor der Rstaurierung 7. Schadensbild und Unterkonstruktion 8. Schematische Darstellung der Schollenbildung 

The causes of the cracking and lifting of the lacquer layers (6) mights be discussed as follows: Fluctuations in relative air humidity cause expansion and contraction movements in the support, and thus lead to contraction cracks and lamina formation of the lacquer layer, which in turn augment the moisture absorption of the wood. In the x-ray-pictures, for example, we can see that a wide gap has opened between the frame and the end of the panel boards, caused by tangential contraction of the frame. In addition, the board joints in the upper area of the panel have opened, while at the middle and the bottom they remained tight. Due to the cracks in the lacquer layer the wood has absorbed moisture, especially through the end-grain-capillaries, causing the wood to swell. A comparison between the schematic derived from the X-ray and the mapping of damage (7) again points up the link between construction and crack formation. No influence on crack formation can be attributed to the bamboo splints on the longitudional joints and only little damage has been caused by the knots.

No connection could be revealed between the construction and the formation of the horizontal splits and cracks in general. The formation of very highly delaminated flakes seems to depend on the forms of radial shrinkage of the wood panels. The cracking in general may well have resulted from a too weak or decomposed binding medium or perhaps from an insufficient drying time of the prime layer before lacquering. Especially thick putty fillings in the support have since formed wide contraction rifts.

I found the ground layers extremely brittle and sanding. The glueing of the paper layer with raw lacquer to the brittle grounding may have caused additional tension and delamination. Climate fluctuation and the anisotropic contraction responses of the wood have combined with the increasing brittleness of the lacquer film to encourage crack formation (8). So I can't say that I found out one single cause for the horizontal cracking. 

Securing measures should be concentrated on reattaching the underlying primer layers to the support. This implies that the lacquer lamina will have to be softened and bent down, in order to be able to adhere the primer layer under them to the layer still solidly bonded to the paper or wood. The lacquer film being under tension, it will tend to return to its former position after cooling and drying, so the adhesive substance used must have an adhesive and cohesive strength great enough to counteract this tensile force. Consequently the adhesive must be relatively highly concentrated and of a corresponding high viscosity. The adhesive should contain as little water as possible, to avoid over-softening of the primer layer and damage to the lacquer film. And finally, to ensure a rational working procedure, drying times should be as short as possible.

A review of restoration materials used in the past reveals that both water and solvent-based adhesives and also hot-seal adhesives like wax-resin-mixtures have been used to secure lifted lacquer on Asian pieces.

To determine their applicability, we proceeded to test these three systems on cracked and flaking areas of the screen, limiting the number of tests to a necessary minimum.

We injected the resin in increasing concentrations and let it dry under pressure for a week. Both, unpolar acrylic resin solutions like Paraloid B 72 and Plexigum PQ 610 and polar acrylic resin solutions like Mowilith in ethylacetate were unsuccessfull; the bonds reopened after a few days. The reasons for the poor adhesion is likely in their good moistening and penetration properties, as a result of which the solutions are absorbed too deeply into the primer layer and wood to remain sufficiently concentrated at the bond.
On the other hand, highly concentrated acrylics, due to their high viscosity, did not sufficiently moisten the bond area. It should also be kept in mind that when synthetic resins are used as adhesives, they turn primer, paper layer and wood support into a plastified, hard mass whose physical behaviour with respect to the neighbouring unsecured areas is unpredictable and can lead to further damage. Their application is extremely time-consuming, the drying times are too long to permit an efficient procedure.

Infiltrated Hot-Seal adhesives form a thick layer under the flakes provide a support and prevent the lamina from moving. But they are not reversible and resist subsequent securing measures.

Water based adhesives are generally well suited to rebond the protein-vehicle primer layers on lacquer pieces. Yet their employment is extremely problematical, since photomechanically damaged, weathered lacquer surfaces, especially Chinese Export lacquers, are irreversibly discoloured by contact with water. In addition, long exposure to moisture softens and swells the hygroscopic lacquer and can cause an expansion of the area being worked on, which will hinder the proper seating of the loose layer. Also, primer layers on Chinese Screens are often highly water soluble and tend to wash out. When one considers using waterbased adhesives for securing measures, it would be advisable to look for methods that preclude moisture damage.

Thus a tentative solution to the problems just described seemed to be the application of a water-impermeable substance to the lacquer surface. In the areas to be bonded with water-based adhesives we therefore applied Plexigum PQ 610 in White Spirit and left this to dry for at least 24 hours before applying the adhesive. After the bond was secure, no signs of heat, moisture or bleaching damage were seen in the isolating or lacquer layer. The film was then removed with white spirits.

9. Anwärmen der Oberfläche 10. Fixieren der Klebestelle 11. Röntgenuntersuchung der fixierten Schollen12. Detail Paravent nach der Restaurierung 13. Detail Kabinett nach der Restaurierung 

The lacquer layers had lifted up to 4 mm high and the lamina tended to break when pressure was applied. Even the pressure of an injection needle on the split edges could loosen chips. We therefore attempted to increase the elasticity of the layers by applying heat and/or moisture (9). During the application of the glue the surface was warmed with an infrared lamp to a maximum temperature of 60°C. The surrounding area was covered with a heat-reflecting material. In addition, preliminary moistening with destilled water and tenside helped soften the lifted areas and expanded the cracks, facilitating the injection of the adhesive.

The lifted lacquer lamina on the screen tended to return to their former position after cooling and drying. The tensile forces involved are considerably higher than in the case of, say, veneers. Thus with all types of adhesive tested on the screen we used heated boards and clamps (10) to set the loose lacquer. We used an intermediate layer of Chromatco paper, which absorbs extruded adhesive and is permeable to water vapour.

In order to be able to assess the depth of penetration, permeation behaviour and distribution of the adhesive under the lamina we dissolved the adhesives in non-ionic- iodisized salt solution and put it under x-ray (11) after drying.

For warm glue-bonding, Saliansky Isingglass was used in 8, 10 and 20% contrast solution. The adhesive was injected under raised areas and droplets applied under the fine cracks. As is clearly visible in the X-ray, injected glue distributed itself evenly under the raised lamina, but the amount of the infiltrated glue was so low that it was hardly detectable on the film.

We tested acrylic dispersions, Plextol D 498 and Mowilith DM 771, each thinned 1:1 with iodisized salt solution. Both adhesives penetrated very well into fine cracks. While the bonding results were initially very promising, three weeks later we noticed that many of the horizontal oval-shaped lamina along the splits had reopened.

X-ray examination showed that the strongest bonds were those with the thickest layers of glue under the lamina. Evidently we needed a medium to fill out the hollow spaces under the lifted areas. So we began to experiment with gels of rice starch. The paste was injected and distributed under the loose lamina by shifting them. After two hours the lacquer layer had softened sufficiently to be pressed. The resulting bonds turned out to be very stable. However, rice starch produces extremely high-tension films. So we conducted extensive experiments in modifiying the starch and arrived at a paste made of Zin Shofu (Japanese wheat starch) 1:4 in distilled water, mixed in 2 parts with one part of 20% isinglass. By comparison of the other adhesives tested, this starch paste evinced the highest degree of contrast in the X-ray, will say the adhesive layer was quite thick. The advantage of the gel form of the paste is that it has a lower water content than isinglass and largely prevents uncontrolled spreading of the adhesive. In addition, starch is compatible with the primer layer, creates good bonding with fillers without washing them out and permits of rebonding.

After all this preliminary work we decided to lay down the endangered areas of the screen with the starch paste and leaving fine cracks, where no lifting had occurred in their present condition.

The following conservation of the screen gave us the opportunity to practice and to modify our conservation techniques. After the isolation film had dried, we injected a small amount of distilled water in the cracks and warmed the lacquer so that the cracks could open and smoothen. We found out that it was very important to give the lacquer layer enough time to dry after having absorbed the water from the adhesive. After the injection of the starch paste we had to wait at least twenty-four hours before pressing the layers with the clamps. The swelling of the lacquer had gone back so far but the remaining humidity yet improved the elasticity of the lacquer layers.

After having found out a suitable method of securing the layers, the consolidation of the screen was likewise easy to handle: we could lay the panels on a table and use our clamps (12).

In a way we can call it a preliminary exercise for the upcoming restoration of the Lacquer Cabinet in Schloß Falkenlust. Here we were faced with the same kind of damages. Due to unfavourable climatic conditions the lacquer layers, especially of the gold paravent, showed quite dramatical, extensive delaminating and cracks. In this case we had to work in situ and vertically, what meant that the use of clamps was not possible. We couldn't put too much pressure to the panels because they were thinned to about 8 mm thickness and very sensible. Also we had to use an adhesive that would stay in place and not run down the walls. Preliminary securing attempts assured us that in this case, too, our starch-and isinglass- paste was a practicable binding medium. In Falkenlust we didn't have to apply an isolation film because the asian lacquer was covered with a very thick layer of an old mastix varnish. So we lifted our infrared lamp onto the scaffold and began to work. The climatic conditions were very good during the summertime, so that the adhesive could dry rather quickly within one or two days. After drying we could lay down the layers with a heated spatula. This method showed very good results, even when some of the cracks reopened after some weeks. The operation could be repeated and the lacquer layers refixed. (13) 

Under the conditions we have found at Falkenlust there is no guarantee for a long term stabilization if the climate can not be stabilized. But two years after the restoration the consolidated areas remained fixed onto the support and this lets me believe that we could be on the right way regarding the consolidation of lacquer layers with suitable materials that allow further restorations.