Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

An investigation of the use of cellulose-based materials to gap-fill

wooden objects
Kate Fulcher, Institute of Archaeology UCL
This is an Accepted Manuscript of an article published by Taylor & Francis Group in Studies in
Conservation, available online: http://www.tandfonline.com/doi/full/10.1080/00393630.2015.1109294

Abstract
The conservation of wooden objects is complicated by their response to changes in their
environment. In particular, filling voids in wooden objects can be difficult when their dimensions
might be expected to alter over time. A short survey of conservators showed that a wide variety of
materials have been, and are being used for this purpose. Following conservation work undertaken
on a wooden coffin, the author wished to investigate the properties of cellulosic materials and their
suitability in the conservation of wooden objects.

This paper shares the results of experiments that were conducted to determine the way in which
hydroxypropyl cellulose and paper pulp fills respond in various conditions, as compared to other fill
materials used by conservators. The compatibility of the hydroxypropyl cellulose and paper pulp fills
with wood is satisfactory enough to be able to recommend their use in the conservation of wooden
artefacts, but ideally further experiments would be carried out, and on a wider variety of fill
materials.

Introduction
The conservation literature describes the use of cellulose ethers and paper products for filling voids
in wood and other materials. Long-fibre tissue has been applied as a barrier applied on the substrate
before the fill is inserted (Gottsman, 1999; Hatchfield & Ryan, 2010; Kariya et al., 2010), and can be
used in combination with cellulose ethers, including methylcellulose and Klucel® G (hydroxypropyl
cellulose) for filling gaps (Gänsicke et al., 2003: 212; Hatchfield & Marincola, 1994 Johnson et al.,
1995). Johnson et al. (1995) applied a paste of Klucel® G and microballoons coloured with pigments
over a void filled with twists of tissue held with Klucel® G. Kariya et al. (2010) used Klucel® G, bulked
with equal quantities of microballoons and cellulose powder, which they applied in layers to prevent
shrinkage. The same authors refer to unpublished studies by (Bruno et al., 2002-3) at the Brooklyn
Museum of Art on the use of Klucel® G with cellulose powder and microballoons for filling wooden
Egyptian statues. Podany et al. (1995) described the use of paper pulp with PVAc and acrylic
emulsions, or with a methylcellulose, in water or ethanol/water mix for use with ancient Egyptian
wooden sarcophagi, sometimes microballoons were also included. The fills were pressed into the
void, which was covered by a thin plastic film, and finished away from the object, before being
adhered into place using an acrylic resin. Elston (1995) described a similar process filling deep, non-
structural, voids with dry paper pulp, sealing the void with methylcellulose and paper pulp and
applying a thin, top layer of methylcellulose and microballoons to give a smooth finish.

The main experimental research into fillers for wooden objects was published in the late 1980s by
Grattan and Barclay (1988), and Barclay and Mathias (1989). Grattan and Barclay assessed 33
selected fill materials, the only cellulose ether fill material tested was Klucel® G. It was mixed 1.5 to 5
parts (by weight) with finely ground chalk or whiting and added to a 5% ethanol solution. The
resulting mixture was found to slump and shrink excessively upon drying. The study’s best
performing fills, and a selection of the remainder, 22 in total, were chosen for further experiments;
the Klucel® G/whiting fill was not included. Mechanical testing was undertaken to determine the
compression modulus, stress at failure, and shape of the stress/strain curve of the 22 chosen fill
materials. Three types of wood were also tested: western red cedar, white oak and basswood.

An acceptable fill material was defined as one that was weaker, or had a lower compression
modulus, than that of the surrounding wooden material, so that the fill materials would deform
before the wood if the wooden object was under stress (Grattan & Barclay, 1988). The same authors
also investigated the effect of changing relative humidity (RH) on the fillers. Cracks in small blocks of
wood were filled with the selected fillers, and subjected to alternating RH from 20% to 80% at seven
day intervals for seven complete cycles. The blocks were observed for any changes. Grattan and
Barclay concluded that the conservator’s ideal wood fill would be “elastic, strong, benign, easily
workable, easily paintable, and strongly adhesive”. They state that as a result of the study two gap-
fillers were adopted for use in the Ethnology Laboratory of the Canadian Conservation Institute:
silicone rubber and microballoons (glass or phenolic) for flexible fills where movement of the wood
may occur (Barclay & Grattan, 1987), and epoxy with microballoons for structural fills where the
wood is unlikely to move significantly (Grattan & Barclay, 1988: 84).

Barclay and Mathias’s (1989) study, focussed on fills composed of epoxy resin and phenolic
microballoons and reported further experiments. Fills of differing ratios of resin and microballoons
were tested under tension and compression to find a workable mix that was weaker than wood. For
the latter, one centimetre cast cubes of fill were compressed and point of failure noted and
compared. For the tensile strength test, the fills were cured between two sticks of western red
cedar, used due to its known weakness (Barclay and Mathias, 1989: 41) and tested in the same
conditions. The site of failure for each was noted and compared. Tomaszewski et al. (1986) looked at
29 gap fillers for wood, focussing on epoxy resins and polyurethanes (PUR) bulked with wood fibres
or wood flour. They concluded none of the fillers had ideal properties, the PUR fillers were
unsatisfactory in terms of compression and swelling, and the epoxy fillers were not sufficiently
permeable to water (Unger et al., 2001: 556-7).

Mintrop (1997) focussed on thermoplastic and elastomeric fillers for wooden figures and a wooden
ceiling. He recommended Plextol (an acrylic emulsion) with a granulated cork bulker, or a filled
silicone rubber as filling materials for wood. He concluded that microballoons were too rigid to use
with wooden objects, snapping easily under stress, and silicone rubber is undesirable for use with
wood because it is impermeable to water vapour (Mintrop, 1997: 76). Young et al. (2002)
experimented with fillers for the structural repair of panel paintings. Their favoured mixture in terms
of handling and finishing (carving) properties, and strength, being strong enough to hold the wood
panels, but weak enough to fail before the wood, was PVA with coconut flour and microballoons.
The coconut flour was found to impart strength and the microballoons ensured a break within the fill
rather than at the wood surface (Young et al., 2002).

Despite literature reports on the use of cellulosic materials to fill wood, there is limited research into
how these materials perform as fillers. To investigate the properties of cellulosic fill materials the
following research questions were addressed:

Is Klucel® mixed with paper pulp an appropriate fill material for wooden objects?

What criteria could be used to judge the above?


Is the notion of using “like with like” achievable and/or desirable?

To address these questions a survey of conservators was undertaken to garner views on the
desirable properties of fill materials for wooden objects and to determine which materials were
being used for this purpose. Secondly, a series of experiments were carried out to determine
whether Klucel® G and paper pulp met the criteria for an appropriate fill material as defined by the
survey and the literature, and to investigate the behavioural similarities of these materials to wood.

As described by professional bodies (European Confederation of Conservator-Restorers’


Organisations, 2003; American Institute for Conservation of Historic and Artistic Works, 1997) loss
compensation should be reversible, detectable, and easily differentiated from the original material.
Important characteristics of wood fills mentioned in the literature were: reversibility, workability
(ability to handle, shape, colour), stability with ageing, and compatibility with wood in terms of
behaviour with changing humidity. There is some disagreement on whether the fill should be
“strong” (Hatchfield, 1986; Grattan & Barclay, 1988; Podany et al., 1995), or yield easily to changes
in the shape of the wood (Barclay & Grattan, 1987; Johnson et al., 1995; Unger et al., 2001). Several
articles mention a preference for non-toxic materials (Grattan & Barclay, 1988; Podany et al., 1995;
Unger et al., 2001; Young et al., 2002, 85). These characteristics from the literature were used as the
criteria against which the cellulosic materials under investigation were judged.

Survey results [heading]

The complete survey results are reported elsewhere (Fulcher, 2014); a brief summary is given here.
The survey was circulated in April 2012 via the ConsDistList, with 96 respondents, 30 of whom
worked in private practice and 60 of whom worked in institutions, 14 respondents identified
themselves as students. Text answers to some of the questions indicated that their expertise varied
across object types, from furniture to archaeology. The survey did not differentiate between
structural and (purely) aesthetic fills, a failing that was highlighted by many of the respondents. This
makes the results of the survey of limited use, but it has been used to draw some hesitant
conclusions on the role of cellulosic materials.

Materials listed as commonly used for filling small gaps in wooden archaeological or art historical
objects were acrylic resin e.g. Paraloid B72, followed by Klucel® and epoxy resin. The most
frequently mentioned bulking agents were microballoons, followed by paper pulp, paper tissue and
wood. Acrylic resin was often teamed with microballoons, and Klucel® with paper pulp or tissue, but
microballoons were also mentioned in conjunction with Klucel® and acrylic resin with cellulosic
bulkers. When asked to select from a list (Appendix 1), which materials they would not consider
using to fill wood, 16% would not consider Klucel®. 10% said they would not consider any of the
cellulose ethers listed and 12% said they would not consider paper pulp.

Respondents selected from the list in Table 1 the most desirable properties in materials for filling
small gaps in wooden objects. Because the type of fill was not specified the answers to this question
are difficult to interpret, and many respondents commented that the properties of the fill would
have to be adapted for each separate instance because every object is different. However, the
properties agreed on by majority of respondents may be considered to be important for all wood
fills. Most respondents considered retreatability to be an important asset and low amount of
shrinkage on curing was also commonly identified. Respondents indicated it was desirable for the fill
and substrate to react similarly to relative humidity. Many respondents thought a fill with high
flexibility to be desirable as was the ability to shape and colour the fill.

Table 1. Properties of fill materials given for Question 4 in the survey.


Material is similar to wood in terms of reaction to relative humidity
Material is dissimilar to wood in terms of reaction to relative humidity
High flexibility
Low flexibility
High level of hardness (compressibility)
Low level of hardness (compressibility)
Retreatability
No colour change on ageing
High tack when wet
Low tack when wet
Texture
Low amount of shrinkage on curing
High adhesion to substrate on curing
Low adhesion to substrate on curing
Ability to be reshaped after curing (e.g. by sanding)
Ability to colour with pigment
Ability to paint after curing

The most popular filling materials for wood were acrylic resin (e.g. Paraloid® B72) and glass
microballoons, but a wide range of materials were mentioned, in various combinations. Cellulose
ethers were identified by about half of the respondents as a viable option for a filler component, as
was paper pulp, but some text responses indicated unfamiliarity with the products, and a lack of
desire to experiment with materials not previously used. The properties of wood fills that three
quarters or more of conservators judged to be desirable were retreatability, low shrinkage on curing,
ability to colour with paint and pigment, ability to be reshaped after curing, and having a similar
reaction to relative humidity as the wooden substrate.

Experimental
Materials
In order to investigate the notion of “like for like” the materials used in this research are briefly
considered.

Wood

Wood is hygroscopic, which means it absorbs and desorbs moisture readily from its surrounding
environment (Hoadley, 1978), swelling and shrinking with an increase or decrease in RH,
respectively. The anisotropic nature of wood means that the wood swells and contracts by different
amounts in different directions. If the moisture content of the wood changes, but it is restrained by
nails, adhesive, or a fill, the area of the wood next to the restraint will pull or push against it, causing
damage to the wood structure (Gowers, 1971: 45; Plenderleith & Werner, 1971: 126; Young et al.,
2002: 83). Previous decay of wooden cultural heritage objects can make the reaction of the wood to
changes in environment more complicated (Blanchette et al., 1994; Stamm, 1971). This can mean
that filling archaeological wood with the same species of wood is not truly a “like for like” fill
because the fill wood and ancient wood will have aged differently.

Paper pulp
Cellulose is the primary structural constituent of wood (Fengel & Wegener, 1984, 66). It is a complex
natural polymer composed of many units of dehydrated glucose (anhydroglucose) in a chain (French
et al., 2004). Paper is a cellulosic material, obtained from naturally occurring cellulose in a wide
variety of raw materials, including wood (Bajpai 2010). During pulping, lignin and hemicelluloses are
removed from the cellulose, and the process shortens the cellulose chain; the resulting product thus
has different properties from wood despite sharing the same fundamental polymer (Banik & Brückle
2011, 126-130). Therefore, paper pulp cannot be considered an exact “like for like” material with
wood. Paper pulp exhibits the properties of cellulose as influenced by the crystalline and amorphous
regions present. It absorbs liquids in the inter-crystalline amorphous areas, causing the fibres to
swell (Roberts 1996, 73), which has important ramifications for the behaviour of fillers made from of
paper pulp and solvent-based adhesives.

Cellulose ethers

Cellulose ethers are derivatives of cellulose, synthesised by reacting cellulose with alkylating
reagents in the presence of a base, which disrupts the hydrogen bonds, allowing other compounds
to access and bond with the glucose units (Majewicz et al., 2004). The crystallinity of the original
cellulose is usually lost due to the disruption of the strong hydrogen bonds during the process (Feller
& Wilt, 1990: 11). The resulting products are hygroscopic white powders (Feller & Wilt, 1990: 27). A
large range of products can be formed depending on the reagents used, and the products are often
unique to a manufacturer (Horie, 2010: 205). In conservation cellulose ethers are used as adhesives,
sizes, consolidants, and poultices (Horie, 2010: 211). The etherifying process results in a dilution of
the cellulose present so the product contains only a fraction of the original cellulose, for example in
hydroxypropyl cellulose only 40% of the cellulose component remains (Feller & Wilt, 1990: 14). Thus
cellulose ethers cannot be considered to be water soluble versions of cellulose. Most importantly,
ethers lack the crystalline structure of cellulose and instead exhibit an amorphous, non-fibrous
morphology (Feller & Wilt, 1990: 14) much more akin to a gel. Although based on cellulose,
hydroxypropyl cellulose cannot be considered an exact “like for like” fill material for wood.

Klucel® is the brand name for a cellulose ether, hydroxypropyl cellulose (HPC), manufactured by
reacting alkali cellulose with propylene oxide at high temperature and pressure, replacing most of
the primary hydroxyls on the cellulose (Hercules Inc., 2001: 4; Hofenk-de-Graaff, 1981). Klucel® is
available in various grades and viscosity types. Klucel® G was chosen for this study as it is mid-range
in viscosity, and is the most commonly used type in conservation (Feller & Wilt, 1990: 102). Klucel® G
is soluble in cold water and ethanol, and slightly soluble in acetone. These solvents were chosen for
this study as they are commonly used solvents by conservators. When mixed with bulking agents
such as paper pulp, Klucel® G can be used as a filler for areas of loss in wooden objects. It is the
properties of such mixes that were investigated.

The stability of cellulose ethers, including Klucel®, was investigated by Feller and Wilt (1990). They
concluded that high molecular weight Klucel®s (types H and M) were unstable, but results were
inconclusive for Klucel® G (Derbyshire & Rutherston, 1991; Feller & Wilt, 1990). Photochemical
ageing tests indicated that the rate of breakdown of molecular chains by light in Klucel® G was
similar to that cellulose (Feller & Wilt, 1990: 90). Hofenk-de-Graaff (1981) found that Klucel® G
compared well in age testing with other cellulose ethers commonly used in conservation.

Methodology
The experiments were designed to be repeatable and were intentionally low-tech, using only basic
measuring devices and equipment.
The research sought to determine the following:
 Dimensional shrinkage of the fills on curing
 Toughness of fill
 Compressibility/flexibility of fill
 Reaction of fill to various RH levels
 Ease of colouring or shaping fill after curing
All experiments were conducted using flat cylindrical samples measuring 20mm in diameter formed
of an adhesive dissolved in solvent, and a bulking agent.

Klucel® G was used at three concentrations: 3%, 7% and 12%. This range was to represent the use of
Klucel® G as a fluid adhesive at 3%, as a stiffer concentration at 7%, and as a high concentration at
12% when the solution becomes a thick gel. Cold water, industrial methylated spirits (IMS, also
known as denatured alcohol), and acetone were tested as the solvents, to assess the effect of the
solvent, if any, on the preparation and behaviour of the filler. Three bulking agents were used:
Arbocel BC1000 (0.7mm fibre paper pulp), Arbocel BWW40 (0.2mm fibre paper pulp) and glass
microballoons (0.43 grams). The bulking agents were mixed at the ratio with the adhesive that
created a usable fill i.e. dry enough to be manipulated by hand, but without crumbling. The ratio
used by weight of the adhesive was: BC1000 at 30%; BWW40 at 50%; microballoons at 138%.
Samples were prepared in two thicknesses: 3mm and 6mm.

Control samples of the same diameter and in both thicknesses, were made of 50% (by weight to
volume) Paraloid® B72 in acetone with the three test bulking agents, in the same weight to volume
ratios. Paraloid® B72 was chosen for the control based on its popularity in the survey and from the
literature review e.g. acrylic resin (Paraloid® B72) with microballoons was a popular choice for small
gap fills in wood (Gänsicke et al., 2003: 212; Hatchfield, 1986; Hatchfield & Ryan, 2010; Jackson &
Watson, 1995; Simkin & McDonagh, 1995). Acetone was used because it is a good solvent for this
adhesive (Koob, 1986; Rohm & Haas, 2009). A concentration of 50% was chosen because this is a
moderate adhesive strength for this resin.

During the experiments RH was between 45% and 50% and temperature between 17.5°C and 20°C.

Shrinkage
The shrinkage experiment assessed the amount of shrinkage that occurred when HPC cures versus
that of Paraloid® B72. Shrinkage is an important property to consider when choosing a fill material; a
fill that shrinks a great deal on curing may cause problems by pulling at the surrounding material
loosening or even becoming detached from the object. This was confirmed by the survey results. A
full set of samples of the fill materials were left to cure for a week, after which the thickness and
diameter of each sample was measured with a digital micrometer in three different positions for
each dimension. The entire experiment, including preparation of sample mixtures, was repeated
three times.

Toughness
In a conservation treatment, it is preferable that a fill be weaker than the surrounding material and if
the two are firmly attached, that the fill will fail before original material is damaged. (Young et al.,
2002: 95). 20mm diameter pine dowelling was cut into 4cm lengths, and the ends sanded as flat as
possible. Klucel® G was found to be too weak an adhesive to hold the samples onto the dowelling for
the duration of the experiment, so the samples were mounted between two pieces of dowelling
using 50% Paraloid® B72 in acetone as an adhesive. To achieve a good adhesion the ends of the
dowels and the surfaces of the samples had to be consolidated using 30% Paraloid® B72 in acetone.
In retrospect, it would have been preferable not to have applied any adhesive or consolidant to the
samples for this test (and the compression test), which could have been achieved by creating
“dogbone” samples, as used for tensile testing metals and plastics (ASM, 2004). However, the author
did not have access to high precision instruments nor the means to create hundreds of standardised
“dogbone” samples. Every sample was consolidated in the same way, so the results are internally
consistent and comparable. The materials tested for toughness in this experiment are listed in Table
2.

Table 2. Samples used in toughness experiment at a thickness of 6mm.

Adhesive Filler
Arbocel Arbocel
Klucel® G 3% in water Adhesive only Microballoons
BC1000 BWW40
Arbocel Arbocel
Klucel® G 7% in water Adhesive only Microballoons
BC1000 BWW40
Arbocel Arbocel
Klucel® G 12% in water Adhesive only Microballoons
BC1000 BWW40
Paraloid® B72 50% in Arbocel Arbocel
Adhesive only Microballoons
acetone BC1000 BWW40

While testing tensile strength the loading was set at a maximum of 30kg. Loading was increased until
the sample fractured, or the sample tore from the adhesive layer between it and the dowelling,
indicating that the sample was stronger than the adhesive. The site of failure was noted (Pocius,
2002: 43-46). If the adhesive or the substrate fails before the fill, then the fill should be considered
too strong for conservation use.

Compressibility/flexibility
Following the same sample preparation described for toughness testing (Table 2), samples were
compressed in the tensometer to a maximum of 250 kg, believed to be well beyond the maximum
force that would be applied to a fill in a museum object. The effect of the loading on the samples
was observed, and the load at which the samples fractured, if at all, was noted.

Reaction to relative humidity


Four of each sample was prepared and weighed separately with their masses recorded. One set of
samples was placed in a container with silica gel at a low RH (approx. 20%), one set in a container
with a reservoir of water keeping it at a high RH (approx. 90%), and one set with weekly alternating
Prosorb® and Artsorb® cassettes set at approx. 30% and 65% RH respectively. The fourth set of
samples were left open to the ambient environmental conditions in the lab as the control group.
Included with each set of samples were three samples of wood: mahogany, oak and pine, measuring
20mm in diameter and 6mm in thickness. After ten weeks at these conditions, the samples were
removed and weighed again.

Ease of colouring or shaping


It is desirable for fills to blend with the object to which they have been applied, which might involve
applying colour. The most common ways of achieving this are by mixing dry pigments in with the fill
while it is being prepared, or painting the fill after curing. Conservators may also wish to adjust the
size or shape of a fill after it has cured, by abrasion, cutting, or using solvents applied on swabs.

The ease with which the cured fills could be painted was observed by creating flat areas of each fill,
approximately 5cm square, on a wooden board. Acrylic emulsion paint, watercolour paint (mixed
from solid block using tap water), gouache (premixed in a tube) and egg tempera with pigment
(mixed by the author using fresh egg yolk) in burnt sienna were painted onto the samples.

To determine the ease of finishing the fills, samples of each fill material were swabbed with water,
IMS, and acetone using a cotton wool swab. In addition, a scalpel blade was used to try to adjust the
shape of the samples by carving or shaving the sides, and they were abraded using garnet paper 129
(grade 7/0, grit 240) and Micro-Mesh 1500, a smoother polishing cloth.

Results and discussion


Observations
Klucel® solutions were made at room temperature with newly purchased Klucel® G, stored in an
airtight container. Solutions made with water were clear, those with IMS were slightly hazy and
those with acetone were opaque (Error! Reference source not found.). It was found that using older
stores of Klucel® G produced clearer solutions when combined with IMS and acetone. This is thought
to be due to the absorption of water from the atmosphere which acted as a cosolvent (Hercules Inc.,
2001, 8). Due to the increased granularity of the Klucel® G solutions in acetone, these solutions are
thicker than the water solutions at the same solids concentration, e.g. 7% Klucel® G in water was less
viscous and smoother than 7% Klucel® G in acetone. Solutions in IMS were somewhere in between,
this had a bearing on the ease with which the adhesive could be mixed with bulking agents.

Figure 1. Klucel® G solutions made with, from left to right, water, IMS, and acetone.
The Klucel® fillers that were easiest to mix and formed the most coherent and pliable pastes were
those with a high concentration of adhesive, a small particle size, and compatibility of solvent and
bulking agent (e.g. water with paper pulp). Of the samples that mixed well, those made with
microballoons formed the smoothest paste, but were brittle. Those made with short fibre paper
pulp (BWW40) also had a smooth surface, but were more flexible (see experiment on compressibility
below). The ease of mixing the fill materials indicates the ease with which pigment can be
introduced in order to produce a coloured fill. Those that formed a smooth and pliant putty were
easier to mix with pigments than those that crumbled and separated.

One of the benefits of using Klucel® was the ease with which equipment could be cleaned afterward.
Klucel® is non-toxic, in fact one of its main commercial uses is as a food additive (Hercules, 2001: 3).
If water is used as a solvent the fill can be prepared without the need for air extraction and
equipment can be washed in the sink using tap water, or wiped down using a damp cloth. Paper pulp
is a natural product and entirely non-toxic; as with Klucel®, there are no disposal requirements
(Kremer Pigmente, 2008a & b).

Shrinkage
It was assumed that all samples were created equal in width because they were all formed in a
mould with 20mm diameter holes, half being 3mm thick, and half being 6mm thick. In reality, it
proved very difficult to apply all of the fill mixes evenly into the mould, especially the long fibre
length, low concentration acetone-based Klucel fills because the mixture was bulky and crumbly. The
samples that shrank the most were those that used water as a solvent and paper pulp as a bulking
agent (Error! Reference source not found.2 & 3). The Paraloid® B72 control samples showed the
least shrinkage. Overall, samples with water as a solvent shrank the most. Those using IMS shrank
less, and those using acetone the least.

Figure 2. Size of 3 mm deep samples after curing. Results shown are averages across three measurements each of thickness
and width per sample, and across three repeats of the experiment. Labels refer to solvent/concentration/bulking agent, all
Klucel® G excepting those labelled ‘B72’ for Paraloid® B-72. ‘1000’ refers to Arbocel® BC1000 (long fibre paper pulp); ‘40’
refers to Arbocel® BWW40 (short fibre paper pulp); ‘Mbs’ refers to microballoons.

Figure 3. Size of 6 mm deep samples after curing. Results shown are averages across three measurements each of thickness
and width per sample, and across three repeats of the experiment. Labels refer to solvent/concentration/bulking agent, all
Klucel® G excepting those labelled ‘B72’ for Paraloid® B-72. ‘1000’ refers to Arbocel® BC1000 (long fibre paper pulp); ‘40’
refers to Arbocel® BWW40 (short fibre paper pulp); ‘Mbs’ refers to microballoons.

Samples using paper pulp showed the most shrinkage for all of the solvent types and concentrations.
The samples made with microballoons showed the least shrinkage and the least variability in size
change across sample types. The concentration of the adhesive in the sample solutions did not seem
to affect the extent to which the samples shrank on curing. The vast majority of the solution, at all
concentrations, was the solvent, so the solvent had the largest effect on the way in which the
adhesive cured. The larger shrinkage of samples with water as the solvent was due to the swelling of
the paper pulp. The water absorbed by the pulp enlarges the sample, and when the water
evaporates and the sample cures, the pulp shrinks. Acetone and IMS swell the paper much less, and
evaporate much more quickly, which means that the initial size of the sample is closer to the final
size.

The ideal filler does not change dimensions between its “wet” and “dry” state, so in use the filler will
fill the gap and remain in place once cured. From this point of view, the Klucel® G and paper pulp fills
did not perform well when water alone was used as the solvent. The shrinkage was reduced by using
a non-aqueous solvent such as IMS, or by mixing an organic solvent with water. In the latter case,
the organic solvent reduces the initial swelling of the fibres.

Toughness
The amount of loading applied to cause sample failure is shown in Table 3. 3% Klucel® G was too
weak an adhesive, and the samples came apart whilst being loaded into the tensometer. Only one
sample of 7% Klucel® G made it into the tensometer without breaking. 12% Klucel® G and 50%
Paraloid® B72 both held for testing. Samples were not precisely the same size or shape, having
shrunk slightly on curing, and those made with longer fibre paper pulp had much rougher edges than
those made with microballoons. The results are interpreted comparatively within groups, for
example comparing the microballoon results to each other. However, the results suggest that there
was a significant difference between the strength of Klucel® G at 3%, 7% and 12%, the 12% being
significantly stronger, and in one case (for BWW40) comparable to the Paraloid® B72 samples in
terms of applied load before failure.

Table 3. Results from toughness experiment. Experiments were conducted 3 times (a, b and c);
blanks indicate that the sample failed before testing.

load applied to cause sample failure / kg

a b c AVERAGE

3% Klucel® G
Adhesives only

7% Klucel® G 2.2 2.2

12% Klucel® G 10 8.2 6.1 8.1

50% Paraloid® B72 11.9 15.7 11.2 12.9

3% Klucel® G, BC1000 0.8 1.6 1.2

7% Klucel® G, BC1000 3.4 7.8 5.8 5.7

12% Klucel® G, BC1000 11.4 7.8 11.1 10.1

3% Klucel® G, BWW40 1.8 1.8 1.8

7% Klucel® G, BWW40 7.2 3.6 5.6 5.5

12% Klucel® G, BWW40 16 15.7 9.4 13.7

3% Klucel® G, microballoons 0.5 5 1.2 2.2


Adhesives with bulking agents

7% Klucel® G, microballoons 4.4 8.4 6.4

12% Klucel® G, microballoons 11 9.8 13.2 11.3

B72, BC1000 19 13.5 10.5 14.3

B72, BWW40 11 13.6 16 13.5

B72, microballoons 14.3 14.6 12.5 13.8

The averages from Table show an increase in the force required to fracture the fill as the strength of
the adhesive increased. However, the behaviour of the fill is affected by both the strength of the
adhesive and the type of bulking agent used. The 3% and 7% Klucel® G fills bulked with paper pulp
failed within the fill, pulling and tearing the paper. The 12% Klucel® G fills were very strong, and
failed at the very edge of the fill, in some cases pulling at the Paraloid® B72 adhesive holding the fill
to the wood. The Paraloid® B72 fills were also very strong and failed at the interface of fill and
Paraloid® B72 adhesive on the wood surface.

The toughness of fill required by a conservator varies with the application of the fill. Low
concentrations of Klucel® G will fracture easily, but this may be a desirable characteristic if the
object itself is weak and the fill is required to break first; alternatively the fill may be required to
provide strength, in which case a high concentration can be used. Fills made with acrylic resin, such
as Paraloid® B72, are very hard and strong.

Compressibility
Fills made with paper pulp and Klucel® G readily deformed under pressure, and did not suddenly fail.
In comparison, fills made with microballoons did not compress, but rather resisted compression until
a point of catastrophic failure. Fills made with Paraloid® B72 did not compress or fail at a loading of
250kg. All samples made with long fibre paper pulp (BC1000) and Klucel® G were compressed to the
maximum of 250kg without failure. The samples made with short fibre paper pulp (BWW40) and
Klucel® G compressed to 250kg without catastrophic failure, but the lower concentrations began
crumble or fold at the edges below this load, 3% Klucel® G at about 100kg and 7% Klucel® G at about
150kg.

When the load was removed the 3% fills broke apart, but the 7% fills remained intact. 12% Klucel® G
with short fibre paper pulp compressed under a load of 250kg without disruption, and showed no
signs of damage after removal of loading, although the fills were permanently compressed to a
thickness of 4mm (from 6mm). The Klucel® G fills made with microballoons crumbled, with a
substantial loss of material from the sample. The higher the concentration of Klucel® G, the greater
the load the samples could take. Paraloid® B72 samples with BC1000, BWW40 and microballoons all
withstood compression force to 250kg with no apparent effect. Wooden samples also resisted
compressive force to 250kg with no apparent effect.

Reaction to relative humidity


Although only one set of samples was tested in each of the RH environments, each set included
eighteen samples with each filler (9 at each thickness), and the reactions of the samples were
consistent according to type of bulker (paper pulp or microballoons) and adhesive (Klucel® G or
Paraloid® B72) within each environment. The other variables (strength of adhesive and solvent) had
little effect on the reactions of the samples to RH.

The samples stored at low RH for ten weeks lost between 2% and 10% of their mass, and the
samples stored at high RH gained mass. Control samples kept at room temperature and RH for the
whole of the course of the experiment showed no change in mass. At low RH the samples that
showed the greatest change in mass were the wooden discs. The paper pulp fills lost the second
largest amount of mass. The fills made with microballoons were least likely to change mass from loss
of moisture. No visible change occurred in any of the samples. At high RH all samples increased in
mass. The greatest change was in the samples made with microballoons, which became almost
liquid. The samples made from paper pulp and Klucel® G became pliable and easily compressible.
The different reactions to high RH demonstrate that the paper pulp and Klucel® fills absorbed the
water, whereas in the case of the Klucel® G and microballoon fills the Klucel® absorbed all the water,
at such a high rate that the adhesive began to liquefy.
The aim of the fluctuating high and low RH experiment was to look for visual structural damage to
the samples after ten weeks RH cycles between 30% and 60%. No changes to the samples were
observed. The final week of the experiment was at low RH, and the samples lost a similar amount of
mass to those in the low RH experiment. This indicates that the samples were responding to the
changes in RH, but that these changes were not causing obvious damage; it is possible, however,
that the samples were structurally weakened by the experiment in a process of structural
rearrangement of the cellulose during hysteresis (Banik & Brückle, 2011: 95).

A desirable property of a fill for wood was that it had a similar response to RH, as the wooden
substrate. The tests demonstrated that paper pulp and Klucel® G fills have a similar propensity to
wood to absorb and desorb water in high and low RH. Fills made with Paraloid® B72 are much less
likely to react to RH due to the hydrophobic nature of the adhesive, and those made with
microballoons can be adversely affected, and even liquefied by very high RH.

Ease of colouring or shaping


Colouring

Paints were easiest to apply to the fill materials with smoother surfaces (Figure 4). Fills using
microballoons were therefore the easiest to paint, followed by fills made of short fibre paper pulp
(BWW40) and high concentrations of Klucel® G. Watercolour paint, and to a lesser extent gouache
and egg tempura, absorbed into paper pulp where the bulking agent was exposed, resulting in an
uneven colour wash. Acrylic emulsion covered well on all surfaces but was easier to apply on the
smoother fills.

Figure 4. Paints applied to fills for colouring test. ‘Microbs’ refers to glass microballoons.

Shaping

For the fills tested, those with a smaller particle size, i.e. microballoons or short fibre paper pulp,
were easier to shape and had the smoothest surfaces. A higher concentration of Klucel® G gave a
more coherent and stronger fill that was better able to take the force of a scalpel blade or sandpaper
(Table 4). Water applied by swab was found to have little effect in shaping fills made with high
concentrations of Klucel® G, despite being an excellent solvent for the adhesive. However, water
was successful at removing material from the microballoon samples. Swabs dampened with IMS and
acetone achieved similar results in removing material from all samples. In fact, this removal was too
successful on samples made with lower concentrations of Klucel® G, with the solvents removing
large amounts of material too quickly (Table 4).
Table 4. Results from experiment on shaping after curing, using solvent, scalpel and sanding with
garnet paper and micromesh. No difference was found in terms of ability to shape between the
Klucel G samples prepared with different solvents. Therefore in the results table the Klucel G
samples are listed only by concentration and bulking agent.

Ability to shape with solvents

Water IMS Acetone Scalpel Sanding

Material Tears rather


Some pulp Leaves a
Klucel G 12% removed than cuts,
No effect material smoothish
BC1000 faster than leaving fluffy
removed surface
with IMS surface

Removes Removes Easily sanded


Can be
Klucel G 12% Softens and material; material; to leave
shaved with a
BWW40 dulls edge shape can be shape can be smooth
sharp blade
altered altered surface

Material
removed Removes Easily sanded
Some
Klucel G 12% faster than material; Easily to leave
material
microballoons with water; shape can be carveable smooth
removed
shape can be altered surface
altered

Material
Quickly Quickly Removes
Klucel G 7% softened and Sands leaving
removes removes a lot chunks, not
BC1000 removed to fluffy surface
material of material shaveable
some extent

Just about
Material Removes Sands quite
Quickly possible to
Klucel G 7% softened and material; well, leaving a
removes a lot carve with
BWW40 removed to shape can be smoothish
of material very sharp
some extent altered surface
blade

Removes Removes Easily sanded


Quickly
Klucel G 7% material; material; Easily to leave
removes
microballoons shape can be shape can be carveable smooth
material
altered altered surface

Garnet paper
causes
Quickly Quickly sample to
Material Causes
Klucel G 3% removes removes crumble;
removed in sample to
BC1000 large lumps large lumps micromesh
lumps crumble
of material of material removes less
material but
surface fluffy
Ability to shape with solvents

Water IMS Acetone Scalpel Sanding

Garnet paper
causes
Removes
Quickly Quickly Material sample to
Klucel G 3% material;
removes a lot removes a lot removed in crumble;
BWW40 shape can be
of material of material large sections micromesh
altered
more
successful

Removes
Quickly Quickly Very easily Very easily
Klucel G 3% material;
removes a lot removes a lot carved and altered by
microballoons shape can be
of material of material shaved sanding
altered

Sands to a
Removes
Can cut but smoother
some
Paraloid B72 not neat and finish with
No effect No effect material but
BC1000 requires a lot vigorous
hard to
of force rubbing
control
(garnet)

Just about Very hard;


Small amount
Paraloid B72 shaveable but possible to
No effect No effect of material
BWW40 requires a lot sand but slow
removed
of force (garnet)

Removes
Paraloid B72 Cleanly Sands neatly
No effect No effect material
microballoons shaveable and smoothly
controllably

Conclusions
The desirable characteristics of fill materials used in the treatment of wooden objects as defined in
the literature may be summarised as reversibility, workability (lack of shrinkage, ability to take
colour and be shaped), stability with ageing, compatibility with wood in terms of behaviour with
changing humidity, and non-toxicity. In addition it is useful to consider its strength, or likelihood to
deform easily allowing changes in shape of the wooden object, either of which may be desirable in
specific circumstances. The survey indicated that conservators were in agreement with these
characteristics. To be an appropriate fill for wooden objects, Klucel and paper pulp fills would have
to meet these criteria.

Experiments found that fills made of these cellulosic materials fit most of the criteria. The
concentration of the adhesive can be adjusted to increase or decrease the strength and hardness of
the fill thus making it adaptable for a variety of situations. During the course of the experiments and
sample preparation it became clear that water is the best solvent for Klucel® G and paper pulp,
achieving a smooth and pliable fill. However, as conservators may be unwilling to apply a water
based adhesive onto a wooden surface, a more volatile solvent such as IMS may be used with a small
addition of water at about 10% to act as a cosolvent. This ensures that the solution is still clear and
smooth, and that the paper pulp is able to absorb some moisture, but that the amount absorbed is
not large enough to cause significant shrinkage on curing. Some acetone may be added to increase
the rate of evaporation, and therefore decrease curing time, but it is not recommended to use
acetone as the sole solvent with a fill made using paper pulp because it does not sufficiently wet the
pulp, and the components do not combine well.

The experiments found that bulking agents with a smaller particle size produced smoother, more
paintable and shapeable surfaces. Microballoons were the smallest particle tested, but the fills
produced using these were brittle and tended to snap or crumble under loading. A short fibre paper
pulp such as Arbocel BWW40 can achieve a flexible fill because of its small particle size. It is also
possible to mix two bulking agents, as has been done with paper pulp and microballoons (e.g. Bruno
et al. 2002-3), but such mixtures were not tested here and it is unclear how such a fill would function
in these tests. The response to relative humidity of fillers made with Klucel® G and paper pulp was
found to be similar to that of wood with moisture absorbed in high RH conditions and desorbed at
low RH.

Klucel® G and paper pulp were found to provide a fill that is compatible with wood, can be made in a
variety of strengths, and with a variety of solvents to suit the needs of the object and the
conservator. It was observed that the behaviour of the fill depended on the combination of the
adhesive and the bulking agent, and to some extent the solvent, because it determines how well the
other two combine. Thus a fill cannot be chosen for the characteristics of its separate components,
but must be considered to be a system, in which the adhesive and the bulking agent affect each
other’s behaviour and reactions to environmental conditions.

The research conducted was not exhaustive, and much more could be done to investigate the
appropriateness of different fills for wooden objects, and other objects commonly filled by
conservators. Suggestions for further work include investigating curing times within wooden objects
or moulds, testing fill materials made with mixtures of bulking agents e.g. paper pulp and
microballoons, testing the reaction of Klucel® and paper pulp fills to fluctuating relative humidity in
chinks in wood as done by Grattan and Barclay (1988), and a cross-institution review of previously
executed repairs or fills with cellulose ethers (and other materials) to look at real time, real museum
environment ageing issues.

Note

An earlier version of this work was submitted in partial fulfilment of the requirements for the award
of the MSc in Conservation for Archaeology and Museums at University College London (Institute of
Archaeology) in September 2012.

Suppliers
Klucel® G Hydroxypropylcellulose

Made by Ashland http://www.ashland.com/products/klucel-hydroxypropylcellulose

Available from various suppliers including


Preservation Solutions, LLC. Golden, CO, USA. Phone: 303.642.3060

Paraloid® B72 acrylic resin

Made by Dow http://www.dow.com/products/market/construction/product-line/paraloid-


b/product/paraloid-b-72-100/

Available from various suppliers including


Conservation Resources International, LLC. 5532 Port Royal Road, Springfield, Virginia 22151, USA.
Phone: (800) 634-6932

Arbocel paper pulp

Made by JRS http://www.jrs.de/wEnglisch/produkte/arbocel.shtml

Available from
Kremer Pigmente GmbH & Co. KG. Hauptstr. 41-47, DE 88317 Aichstetten, Germany. Phone + 49 –
7565 – 91448-0

Glass microballoons

Many manufacturers. The microballoons used in this research were purchased from
http://www.conservation-resources.co.uk/

15 Blacklands Way, Abingdon, Oxon, OX14 1DY, UK. Phone: +44 (0)1235 553166

Acrylic emulsion paint

Made by Lascaux Studio http://lascaux.ch/en/index.php

Available from numerous suppliers including AP Fitzpatrick http://shop.apfitzpatrick.co.uk/acrylic-


paint-9-c.asp, 142 Cambridge Heath Road, London E1 5QJ

Watercolour and gouache paint

Made by Winsor & Newton http://www.winsornewton.com/uk/

Available from numerous suppliers http://www.winsornewton.com/uk/shop/store-locator

Solvents?

ProSorb

Made by Long Life for Art, Germany http://www.cwaller.de

Available from Conservation By Design http://www.conservation-by-


design.com/category.aspx?id=710

ArtSorb

Made by Fuji Silysia Chemical Ltd., Japan. (http://www. fuji-silysia.co.jp/english)


Available from Preservation Equipment
https://www.preservationequipment.com/Store/Products/Conservation-Materials/Other-
Materials/Art-Sorb%C2%AE

Silica gel

Available from numerous suppliers including Conservation Resources


http://www.conservationresources.com/Main/section_20/section20_07.htm

Wood samples

Available from various suppliers including craft shops

Garnet paper 127, grade 7/0, grit 240

Available from numerous online suppliers

Micro-Mesh 1500

Available from Conservation Resources http://www.conservation-


resources.co.uk/index.php?main_page=product_info&products_id=655

References
American Institute for Conservation of Historic and Artistic Works 1997. Commentaries to the
guidelines for practice. Commentary 23 – compensation for loss. Available from:
http://www.conservation-us.org/index.cfm?fuseaction=page.viewPage&pageID=555 [accessed June
2012].

ASM International 2004. Introduction to Tensile Testing. In: Tensile Testing (2nd ed.). Ohio: ASM
International.

Bajpai, P. 2010. Environmentally Friendly Production of Pulp and Paper. Hoboken, N.J.: Wiley.

Banik, G. & Brückle, I. 2011. Paper and Water: A Guide for Conservators. Oxford: Butterworth-
Heinemann.

Barclay, R.L. and Grattan, D.W. 1987. A Silicone Rubber/Microballoon Mixture for Gap Filling in
Wooden Objects. In: K. Grimstad, ed. ICOM 8th Triennial Meeting, Sydney, Australia, 6-11
September, 1987: preprints. Los Angeles: Getty Conservation Institute, pp. 183-187.

Barclay, R.L., and Mathias, C. 1989. An Epoxy/Microballoon Mixture for Gap Filling in Wooden
Objects. Journal of the American Institute for Conservation 28(1): 31-42.

Blanchette, R.A., Haight, J.E., Koestler, R.J., Hatchfield, P.B. & Arnold, D., 1994. Assessment of
deterioration in archaeological wood from ancient Egypt. Journal of the American Institute for
Conservation, 33(1): 55-70.

Bruno, L., Kariya, H., March, T., Posner, K., Moy, S., and Caspi, S., 2002-3. Treatment reports on
Methethy statues BM nos. 50.77, 51.1, 53.22. Unpublished. New York: Brooklyn Museum.

Derbyshire, A. & Rutherston, J. 1991. A review of evaluation of cellulose ethers for conservation.
V&A Conservation Journal, 1.
Elston, M., 1995. Technology and Conservation of a Polychromed Wooden Sarcophagus. In: C.E.
Brown, F. Macalister & M.M. Wright, eds. Conservation in ancient Egyptian collections: papers given
at a conference organised by the United Kingdom Institute for Conservation, Archaeology Section,
and International Academic Projects, held at London, 20-21 July 1995. London: Archetype, pp. 13-21.

European Confederation of Conservator-Restorers’ Organisations 2003. E.C.C.O. professional


guidelines (II): code of ethics. Available from http://www.ecco-eu.org/about-e.c.c.o./professional-
guidelines.html [accessed June 2012].

Feller, R.L. & Wilt, M. 1990. Evaluation of Cellulose Ethers for Conservation. LA: The Getty
Conservation Institute.

Fengel, D. & Wegener, G. 1984. Wood: Chemistry, Ultrastructure, Reactions. Berlin; New York: de
Gruyter.

French, A.D., Bartoniere, N.R., Brown, R.M., Chanzy, H., Gray, D., Hattori, K. & Glasser, W. 2004.
Cellulose. In: H.F. Mark, ed. Encyclopedia of Polymer Science and Technology, Volume 5, 3rd ed.
Hoboken, N.J.: Wiley-Interscience, pp. 473-507.

Fulcher, K. 2014. Survey on Material Used to Fill Wooden Objects During Conservation. Journal of
Open Archaeology Data, 3:e2.

Gänsicke, S., Hatchfield, P., Hykin, A., Svoboda, M. & Mei-An Tsu, C. 2003. The Ancient Egyptian
Collection at the Museum of Fine Arts, Boston. Part 2, a Review of Former Treatments at the MFA
and Their Consequences. Journal of the American Institute for Conservation, 42(2): 193-236.

Gottsman, S. 1999. Investigation and Conservation of an Egyptian Mummy Board from the Royal
Albert Memorial Museum, Exeter. SSCR journal: the Quarterly News Magazine of the Scottish Society
for Conservation and Restoration, 10(2): 5-10.

Gowers, H.J. 1971. The treatment of ethnographical wood. In: Preprints of the contributions to the
New York Conference on Conservation of Stone and Wooden Objects, 7-13 June, 1970. London:
International Institute for Conservation of Historic and Artistic Works, pp. 45-52.

Grattan, D.W., & Barclay, R.L. 1988. A Study of Gap-Fillers for Wooden Objects. Studies in
Conservation, 33(2): 71-86.

Hatchfield, P., 1986. Note on a Fill Material for Water Sensitive Objects. Journal of the American
Institute for Conservation, 25(2): 93-96.

Hatchfield, P. & Marincola, M.D. 1994. Compensating Losses: Tissue Paper Fills for Sculpture. In: E.
Pearlstein & M. Marincola, eds. Loss Compensation: Technical and Philosophical Issues. AIC Objects
Specialty Group Postprints. Proceedings of the Objects Specialty Group Session 2. Washington: AIC,
pp. 57-71.

Hatchfield, P., & Ryan, G. 2010. Selected case studies in the treatment and preservation of Middle
Kingdom polychrome wood in the collections of the Museum of Fine Arts, Boston. In: J. Dawson, C.
Rozeik & M. M. Wright (eds.) Decorated Surfaces on Ancient Egyptian Objects: Technology,
Deterioration and Conservation: Proceedings of a Conference held in Cambridge, UK on 7-8
September 2007. London: Archetype in association with the Fitzwilliam Museum and Icon
Archaeology Group, pp. 78-86.
Hercules Inc. 2001. Klucel® hydroxypropylcellulose: physical and chemical properties. Wilmington,
DE: Aqualon. http://www.ashland.com/Ashland/Static/Documents/ASI/PC_11229_Klucel_HPC.pdf
[accessed July 2014].

Hoadey, R.B. 1978. The Dimensional Response of Wood to Variation in Relative Humidity. In: N.S.
Brommelle, A. Moncrieff & P. Smith, eds. Conservation of Wood in Painting and the Decorative Arts.
Preprints of the Contributions to the Oxford Congress, 17-23 September 1978. London: International
Institute for Conservation of Historic and Artistic Works, pp. 1-6.

Hofenk-de-Graaff, J. 1981. Hydroxy Propyl Cellulose, a Multipurpose Conservation Material. In: J.R.J.
van Asperen de & M. Berends-Albert, eds. Preprints of the ICOM Committee for Conservation 6th
Triennial Meeting, Ottawa, 21-25 September 1981. Paris: International Council of Museums, pp.
81/14/9-1 – 81/14/9-7.

Horie, C.V. 2010. Materials for Conservation: Organic Consolidants, Adhesives and Coatings, 2nd ed.
Amsterdam; Oxford: Butterworth-Heinemann.

Jackson, T. & Watson, J. 1995. Conservation of Waterlogged Chinese Lacquerware from the Warring
States Period 475–221BC. The Conservator, 19(1): 45-51.

Johnson, C., Head, K., and Green, L. 1995. The Conservation of a Polychrome Egyptian Coffin. Studies
in Conservation, 40(2): 73-81.

Kariya, H., Bruno, L., Godfrey, J., and March, T. 2010. Treatment of a Dynasty 18 Painted Coffin,
37.47E a-e (Abbott Collection 405A). In: J. Dawson, C. Rozeik & M. M. Wright (eds.) Decorated
Surfaces on Ancient Egyptian Objects: Technology, Deterioration and Conservation: Proceedings of a
Conference Held in Cambridge, UK on 7-8 September 2007. London: Archetype in association with
the Fitzwilliam Museum and Icon Archaeology Group, pp. 97-105.

Koob, S. P. 1986. The Use of Paraloid B-72 as an Adhesive: Its Application for Archaeological
Ceramics and Other Materials. Studies in Conservation, 31(1): 7-14.

Kremer Pigmente 2008a. 59750 - Arbocel® BC 1000. Available from: http://www.kremer-


pigmente.com/media/files_public/59750_SHD_ENG.pdf [accessed July 2012].

Kremer Pigmente 2008b. 59770 – Arbocel® BWW 40. Available from: http://www.kremer-
pigmente.com/media/files_public/59770_SHD_ENG.pdf [accessed July 2012].

Majewicz, T.G., Erazo-Majewicz, P.E. & Podlas, T.J. 2004. Cellulose Ethers. In: H.F. Mark, ed.
Encyclopedia of Polymer Science and Technology, Volume 5, 3rd ed. Hoboken, N.J.: Wiley-
Interscience, pp. 507-532.

Mintrop, B. 1997. Dauerelastische Kittmaterialien für Holzobjekte zur Verwendung in Durch Quellung
und Schwindung Stark Beanspruchten Rissen und Fugen. Unpublished thesis for diploma. Cologne,
Germany: Fachhochschule Köln.

Plenderleith, H.J. & Werner, A.E.A. 1971. The Conservation of Antiquities and Works of Art:
Treatment, Repair and Restoration, 2nd ed. London; New York: Oxford University Press.

Pocius, A.V. 2002. Adhesion and Adhesives Technology: An Introduction, 2nd ed. Munich: Hanser
Publishers; Cincinnati: Hanser/Gardner Publications.

Podany, J., Elston, M., Thoresen, L. & Maish, S.L. 1995. The Use of Paper Pulp Based Fill Material for
Compensation of Losses to Sculpture. In: J. Heuman, ed. From Marble to Chocolate: The
Conservation of Modern Sculpture; Tate Gallery Conference, 18-20 September 1995. London:
Archetype, pp. 59-64.

Roberts, J.C. 1996. The Chemistry of Paper. Cambridge: Royal Society of Chemistry.

Rohm and Haas 2009. Paraloid® Acrylic Resins: Thermoplastic Solution Grade & Solid Grade Acrylic
Resins. Available from: http://talasonline.com/photos/instructions/paraloid_solvent_guide.pdf
[accessed Feb 2014].

Simkin, M. & McDonagh, E. 1995. Redisplay of the Egyptian Collection at the Horniman Museum. In:
C. E. Brown, F. MacAlister and M.M. Wright, eds. Conservation in Ancient Egyptian Collections:
Papers Given at a Conference Organised by the United Kingdom Institute for Conservation,
Archaeology Section, and International Academic Projects, held at London, 20-21 July 1995. London:
Archetype, pp. 103-108.

Stamm, A.J. 1971. Wood Deterioration and its Prevention. In: Preprints of the Contributions to the
New York Conference on Conservation of Stone and Wooden Objects, 7-13 June, 1970. London:
International Institute for Conservation of Historic and Artistic Works, pp. 1-10.

Tomaszewski, K., Wieczorek, K., Chrzanowski, A. 1986. Z Badan Nad Opracowaniem Kitów Do
Uzupelniania Ubytków w Drewnie Zabytkowym Na Bazie Zywic Pochodzenia Krajowego. (Studies on
the Preparation of Domestically Made Resin-Based Putty to Fill in Missing Fragments in Ancient
Wood). Rocznik Przedsiebior Stwa Panntswo Wego Pracownie Konserwacji Zabytków 1: 259-283.

Unger, A., Schniewind, A.P. & Unger, W. 2001. Conservation of Wood Artifacts: A Handbook. Berlin;
London: Springer.

Young, C., Ackroyd, P., Hibberd, R. & Gritt, S. 2002. The Mechanical Behaviour of Adhesives and Gap
Fillers for Re-Joining Panel Paintings. National Gallery Technical Bulletin, 23: 83-96.

Figure headings

Figure 1. Klucel® G solutions made with, from left to right, water, IMS and acetone.

Figure 2. Size of 3mm deep samples after curing. Results shown are averages across 3 measurements
each of thickness and width per sample, and across 3 repeats of the experiment. Labels refer to
solvent / concentration / bulking agent, all Klucel G excepting those labelled “B72” for Paraloid B72.
“1000” refers to Arbocel BC1000 (long fibre paper pulp); “40” refers to Arbocel BWW40 (short fibre
paper pulp); “Mbs” refers to microballoons.

Figure 3. Size of 6mm deep samples after curing. Results shown are averages across 3 measurements
each of thickness and width per sample, and across 3 repeats of the experiment. Labels refer to
solvent / concentration / bulking agent, all Klucel G excepting those labelled “B72” for Paraloid B72.
“1000” refers to Arbocel BC1000 (long fibre paper pulp); “40” refers to Arbocel BWW40 (short fibre
paper pulp); “Mbs” refers to microballoons.

Figure 4. Paints applied to fills for colouring test. “Microbs” refers to glass microballoons.

Appendix 1. List of materials given for Question 2 and 3 in the survey.

Epoxy resin
Acrylic resin (e.g. Paraloid/Acryloid B72)
Acrylic resin emulsion (e.g. Primal B60A, Lascaux 98HV)
Hydroxypropyl cellulose (e.g. Klucel G)
Methyl cellulose (e.g. Methocel)
Ethyl hydroxyethyl cellulose (e.g. Ethulose)
Polyvinyl Acetate (e.g. Mowilith)
Polyvinyl Acetate emulsion (e.g. Vinamul)
Polyvinyl Butyral (e.g. Butvar B98)
Cellulose nitrate
Starch paste
Silicone rubber
Animal glue
Urushi (Japanese lacquer)
Waxes
Linseed oil
Plaster
Paper mache
AJK / BJK dough (mix of thermoplastic resin, jute fibres and china clay
powder)
Chalk/gypsum (whiting)
Clay powder (e.g. kaolin)
Paper pulp
Glass microballoons
Phenolic microballoons
Wood flour
Sawdust
Cork granules
Fumed silica
A mixture of fillers (e.g. microballoons and paper pulp)
Long fibre tissue
Balsa wood
Commercial wood fillers

You might also like