ArticlePDF Available

Animal glues: a review of their key properties relevant to conservation



Collagen-based animal glues are widely used in the conservation of artefacts, serving as adhesives, binders and consolidants for organic and inorganic materials. With a variety of different animal glues on the market, such as hide and bone glues, fish glues, isinglass and gelatin, their individual properties need to be well understood in order to choose a glue fit for a specific purpose. This paper reviews a wide range of publications on currently available animal glues, with respect to their specific physical, chemical and mechanical properties.
Collagen-based animal glues are widely used in
the conservation of artefacts, serving as adhesives,
binders and consolidants for organic and inorganic
materials. With a variety of different animal glues on
the market, such as hide and bone glues, fish glues,
isinglass and gelatin, their individual properties
need to be well understood in order to choose a
glue fit for a specific purpose. This paper reviews
a wide range of publications on currently available
animal glues, with respect to their specific physical,
chemical and mechanical properties.
Animal glues are natural polymers derived from
mammalian or fish collagen – the major structural
protein constituent of skins, connective tissue, cartilage
and bones. These glues may exhibit varied physical,
chemical and mechanical properties depending on their
origin and method of preparation. In the manufacture of
objects and artefacts, an extensive traditional knowledge
exists on which animal glues are most suitable for
specific purposes. However, conservators sometimes
lack the confidence to make informed choices between
the different collagen-based glues available when
conserving objects.
The selection and preparation of glues are discussed in
patent descriptions, woodworking and artists’ manuals,
as well as conservation literature and product details
from suppliers [1, 2]. There is also a large amount of
technical research on the properties of collagen and
gelatin published in scientific journals on polymer-
and bio-technology, medical science and the food and
brewing industry. However, much of this literature
is not readily accessible to conservators and it can
be ambiguous or contradictory. This paper seeks to
provide a review of the literature and to identify which
properties of glue need to be considered when making
decisions about conservation treatments.
The applications of collagen-based glue in the
conservation field are diverse, ranging from its use as an
adhesive, consolidant or binding medium for pigments
and filler particles [3–8]. Generally, the following key
properties need to be considered:
chemical structure and denaturation of the protein
gelling properties: gelling temperature (Tgel), gel
strength and setting times.
properties of the glue solution: viscosity, surface
tension and pH.
properties of the dried film: cohesion, adhesion
and final bond strength, mechanical behaviour
in changing ambient environment, and ageing
Types of commercially available animal glue
Hide glues are primarily derived from bovine skins
and those of smaller mammals, although connective
tissue may also be used. Bone glues are predominantly
prepared from fresh (‘green’) bones or sometimes
extracted bones (degreased and demineralised, known
as ossein) from cattle and pigs. Hide and bone glues are
produced and sold as coarse powders, pearls, cubes,
and cakes or plates, though the latter two appear to be
increasingly rare [9]. Commercial gelatin, the purified
active ingredient of any collagen-derived glue (pure
denatured collagen), may be obtained from either skin
or bone sources [9, 10] and is supplied in the form of
thin sheets, plates or powder.
As the name suggests, rabbit skin glues should be
produced purely from rabbit skins [10, 11], though
collagenous waste from various small mammals may
also be used [12]. Some suppliers sell rabbit skin glue
that is mixed with bovine hide glue to alter its properties
[13]. The information on the source, pre-treatment,
or additives provided by suppliers may not always
be reliable, as they may not have been given accurate
information by the manufacturers. It is generally
assumed that most animal glues contain preservatives
of some kind (e.g. sulphur dioxide) [9, 14]. Even rabbit
skin compressed into cubes, a by-product from the felt
industry commercially sold as a raw (and thus usually
thought to be a pure) form of rabbit skin glue [10], has
recently been found to contain preservatives [9]. Some
traditional glues, such as the deer glue used in Japan as
a binder for some inks, are now made from bovine or
porcine gelatin manufactured to match the properties of
the traditional genuine material [15].
The skins of non-oily types of fish [16, 17], as well as
their bones [10, 12, 18, 19], are used to manufacture fish
glues which are sold in liquid form. The swim bladders
of various species are the source for isinglass [20–24],
which is available either in the form of complete dried
bladders or membranes, thin plates or fine strips.
In recent years, fish skin and bone gelatin has also
become available in the food industry as a substitute for
mammalian gelatin [25–27].
A number of industrially manufactured cold liquid
animal glues are available that have modified properties
and a long shelf life. These glues usually contain
additives that alter their natural behaviour, extending
the working time at room temperature, or decreasing
the propensity for biodeterioration and reducing the
dried film’s sensitivity to moisture. However, the exact
composition of industrially tailored collagen-derived
glues and their overall performance may be difficult
to judge, as manufacturers tend to keep their recipes
Animal glues: a review of their key properties relevant to conservation
Nanke C. Schellmann
secret. Furthermore, conservation requirements such
as long-term stability and resolubility are unlikely to
be a priority for commercial manufacturers. Given the
range of additives that may be present in industrial glue
formulations, minimally modified glues represent the
safest option for conservation.
Chemical structure and properties
Chemical structure and denaturation
Collagen consists of long protein molecules composed
of naturally occurring amino acids that are linked in a
specific sequence by covalent peptide bonds. Due to
the spatial conformation of some amino acid groups
(notably proline and hydroxyproline) and the many
ionisable and polar functional groups in the protein
chain, the individual chains form triple-stranded helical
coils that are generally believed to be internally stabilised
by hydrogen-bonding [7, 25, 29–33].
Collagen is insoluble in cold water [30, 34] and is
transformed into soluble gelatin by denaturation, a
process of critical importance for the performance
of the resulting glue. This is achieved by hot water
extraction (hydrolytic breakdown) [9, 34–39]. Pre-
treatment (either acidic or basic) is necessary for
most skin and bone collagen, but is not required for
the extraction of isinglass from fish bladders, which
contain less cross-linkage within the collagen. During
extraction, the bonds (predominantly H-bonds) in the
triple-helix structures of the collagen are broken so
that it separates into disordered ‘random’ coils of single
protein chains, thus completing the transition to gelatin
[40]; in perfect conditions gelatin is pure denatured
collagen. The temperature (Td) at which denaturation
occurs is dependent on the chemical structure of the
proteins in the particular collagen source, notably on
the content of the amino acid derivatives proline (Pro)
and hydroxyproline (Hyp). These are supposed to be
largely responsible for the stabilising H-bonded water
bridges in the triple helix [41] and are present more
abundantly in mammalian collagen than in marine
species [42]. Thus, adult mammalian collagen denatures
at 40–41°C [31, 33], while isinglass and other fish
collagens denature at lower temperatures. The Td of
fish collagens ranges from approximately 15°C for deep
cold water fish (such as cod used for fish glue) [10, 43]
up to 29°C for most warm water species [31, 33, 44],
which are the preferred source of isinglass produced for
commercial clarification of alcoholic beverages. There
are also a few tropical fish species that reach Td levels of
up to 36°C [31, 44].
The process of denaturation is necessary for collagen
to convert to gelatin, which can be used as a glue.
Cleavage of the single protein molecule may also occur
during pre-treatment, extraction and dissolution, and
will significantly affect the properties of the gelatinous
glue. The more vigorous the extraction process (i.e.
the more extreme the pH, the longer the treatment
and the higher the temperature during extraction), the
more bonds within the protein molecule are randomly
cleaved, leading to ever decreasing molecular weights
[34, 37, 39, 45]. Mild extraction at moderate pH and
low temperature yields gelatinous matrices containing
protein fractions of long chain length and high
molecular weight (MW) [38, 46]. As a general rule,
gentle processing is appropriate for the hides of young
mammals, as well as all fish skin and swim bladders,
because they are rich in collagen and the collagen is not
so strongly stabilised by the additional chemical bonds
that develop in older mammals. Furthermore, glues that
are derived from fish cleave more easily on extensive
heating than those of mammalian origin owing to their
chemical structure [40, 46]. Conservators should thus
be aware that when preparing a collagen-based solution,
mild procedures should be employed [4, 46]. Preparation
temperatures for collagen-based glues are generally
recommended to be around 55–63°C. However, there is
little loss of gel strength on heating at high temperatures
(e.g. 80–90°C), even in the case of isinglass, but only if
the solution is kept at these temperatures for no more
than a few minutes [46, 47].
Gelation and gelling temperature (Tgel)
Although the process of denaturation, with the loss of
the triple helix arrangement of the protein molecules,
is irreversible, some helical structure can be restored
during gelling and drying. On gelling the single
random protein coils undergo partial rearrangement
(renaturation) back into collagen-like triple helices [7,
26, 46, 48-50]. However, the misalignment of the single
strands means that renaturation causes nodes (‘junction
zones’) involving only part of certain strands. The
remainder of these strands may form further nodes so
that a continuous three-dimensional network structure
emerges. The degree of renaturation is dependent on
the chemical composition (Pro and Hyp content), the
chain length of the molecules (molecular weight, MW),
concentration in solution and temperature [42, 49, 51].
High Pro and Hyp content, high MW, high solution
concentrations and slow drying at a low temperature
promote a high degree of renaturation and the
development of a highly ordered network structure [34,
37, 48, 52]. The number of nodes that are established
by the formation of H-bonds (and probably also by
electrostatic interaction [42]) within and between the
molecules determines gel strength and the rigidity and
elasticity of the glue matrix [7, 46, 51].
The ability to form a rigid gel on cooling, which can be
repeatedly reliquefied by reheating, is one of the unique
properties of collagen-based glues. The temperature at
which gelation of the glue solution occurs (Tgel) depends
mainly on the collagen source, but is also affected by
the degree of protein cleavage. Gelation temperatures
decrease with lower denaturation temperature (Td)
and also with increasing cleavage of the molecules.
Mammalian gelatin gels at around 30–35ºC, and cold
water fish gelatin remains liquid down to around 8ºC
[14, 43, 53]. However, this temperature will be lowered
if the preparation temperature of the glue is significantly
Gel strength
Gel strength is a measure of the gel rigidity of gelatinous
glues, and is strongly influenced by the molecular weight
of the constituent proteins [34, 54]. According to
several authors [35, 39], the average molecular weight
(AMW) of animal glues can range from around 20000 to
250000 g.mol-1. It is thought that permanent gelling
does not take place below an AMW of 20000 g.mol-1
[38, p. 43]. Isinglass from sturgeon, if prepared under
mild conditions, reaches average molecular weight values
of well over 150000 g.mol-1 [4, 33, 46], while liquid fish
glue has AMW values of around 60000 g.mol-1 [10, 14],
placing it at the lower end of the range. For most other
commercial collagen-based adhesives, information on
AMW is not readily available.
Characterisation by AMW is only common for fish
glues, which are liquid at room temperature. Most
other gelatinous glues are usually characterised by their
gel strength, as AMW does not describe the molecular
weight distribution and therefore may not always
correlate reliably with the physical and mechanical
properties of a glue [34, p. 60] (Table 1). However, it
would be expected that high AMW adhesives, such as
skin glues, have higher gel strength and viscosity, gel
more rapidly and produce stronger bonds.
Gel strength is strongly influenced by AMW but also
shows a linear correlation with the degree to which
the protein solution renatures during gelation [55],
i.e. the higher the degree of formation of helical
structures, the higher the gel strength. The presence of
salts also influences gel strength, which decreases with
an increasing concentration of ions in solution [42,
Gel strength, also known as Bloom strength, is measured
in grams (g), or Bloom grams (gB), and equals the force
required to make a specified depression into a gel sample
prepared under standard conditions [25, 35, 37, 39].
Manufacturers commonly distinguish between grades
of glue by their Bloom strength, which usually covers
a wide range, being as low as 30 g for weak bone glues
and rigorously extracted hide glues, and up to around
500 g for very strong hide glue [10, 11, 35, 37, 57].
Gelatins derived from tropical fish have significantly
lower Bloom values than mammalian gelatins [58],
since the degree of stabilisation of the triple helix by
H-bonding is lower. Gelatins extracted from cold water
fish do not have specified gel strengths as they are liquid
at room temperature [42].
As gel strength is dependent on the structural
conformation of the gelatinous matrix, it is useful
for estimating the toughness, strength and resilience
of the resulting bond. Furthermore, Bloom strength
also correlates with the water-sorption capacity of the
glue (in gel and solid state), viscosity (at least to some
degree), and gelling temperature (Tgel), which generally
all increase with rising Bloom value. High Bloom glues
require a lower solid content in solution than glues with
a lower Bloom rating to be effective as an adhesive, as
they offer many sites for intermolecular bonding in a
given volume [35, 56]. Mammalian skin glues are usually
considered to have the highest AMW and produce the
strongest gels and films [10], particularly those extracted
by acid pre-treatment. Generally, acid pre-treated glues
(type A gelatins) contain larger fractions of high MW
than alkaline pre-treated collagen derivatives (type B
gelatins), whose MW distribution is skewed towards
lower MW fractions [9, 34, 46].
Open (gelling) time, tack and drying
The setting time of animal glues depends primarily on
Tgel and gel strength. The lower the Tgel and gel strength,
the longer the open time of the solution (i.e. the longer
it takes for the glue to gel). High Bloom hot hide glues
tend to gel rapidly, as gelation occurs at comparatively
high temperatures [10, 11, 14, 39, 59]. Gelatinous
glues derived from fish, which have low Tgel due to
their chemical structure [42, 43, 58], and cold-set liquid
hide glues are convenient to use when long open times
are required. Commercial fish glues usually contain
preservatives [60] and, sometimes, small amounts of
other additives such as colour brightener, deodorizing
agents or fragrance [10]. Liquid hide glues generally
have further additives to inhibit gelation at room
temperature [17, 28]. These are typically salts (e.g.
urea, thiourea) or phenols that extend the setting time
by inhibiting renaturation of the gelatinous matrix [28,
52]. Some manufacturers claim that their liquid hide
glue does not contain gelling inhibitors [17], in which
case the gelatinous matrix must be considerably affected
by molecular cleavage to achieve the comparatively
low MW that is necessary for the glue to be in a liquid
The ability of collagen-based glue to develop tack upon
gelation is a unique property. In general, glues of higher
Bloom strength develop tack faster than lower Bloom
glues. The tack ‘strength’ of glue can be empirically
tested by conservators between two fingertips. Isinglass
solutions may appear to be less tacky than equivalent
concentrations of mammalian gelatin or hide glue, as
they take longer to set at room temperature, since their
lower gelation temperature delays the development of
Drying time generally depends on the ambient
temperature and relative humidity (RH). After gelation,
the glue matrix dries by evaporation of water and this
process can be accelerated by elevating the temperature.
However, collagen-based adhesives should be allowed to
dry as slowly as possible, as a longer period of molecular
mobility after gelation and during drying encourages the
development of highly ordered network structures [52].
This maximises the elasticity and strength (toughness) of
the resulting glue film. Isinglass naturally develops highly
stable and elastic films if dried at room temperature,
being slightly above its Tgel [9].
Properties of gelatinous glue solutions
The viscosity of the glue solution is primarily dependent
on the molecular weight distribution [51]; the greater
the proportion of molecules of higher MW the higher
the viscosity [2, 35]. For a given MW distribution,
the viscosity increases with increasing solution
concentration and decreasing temperature [39, 51, 61].
The degree to which collagen-like helices [62, p. 128]
Table 1 Comparison of the properties of different glue types. The glues are qualitatively ranked relative to one another for each property, i.e. within each individual column. Numerical data is only
referred to in those cases where information was consistent in the literature
PROPERTY molecular weight (MW) [Ref.] gel / Bloom strength [gB] [Ref.] degree of helicity [Ref.] viscosity [Ref.] pH (approximate values) [Ref.]
decreases with rigorous pretreatment
and with excessive/prolonged heating increases with higher MW and
increasing helicity increases with higher MW,
higher Pro and Hyp content
and increasing solution
increases with increasing Bloom,
dependent on isoelectric point
(pI) and pH
influences the viscosity
bone glue low to medium low to medium (down to
50 gB) [9–11,
14, 35,
37, 57,
low to medium low to medium (min.
viscosity around pH
4.5– 5.5)
[10, 14,
69] 5– 7 [5, 10,
35, 37,
hide glue high [10] high (up to 500 gB, hide
glue pearls produce lower
Bloom values than hide
glue grains)
14, 35,
37, 57,
medium to very
high medium to high (min.
viscosity of alkaline
pretreated glue at
around pH 4.5–5.5,
and of of acid
pretreated glue at pH
7.0– 9.0)
[10, 14,
17, 30,
34, 35,
51, 56,
6.5–7.4 (wider variations
are possible) [9–11,
16, 35,
rabbit skin glue high [10] high (up to 500 gB) [9, 10,
14] high to very high high (min. viscosity at
around pH 7.0– 9.0) [9, 10,
14] 5.0–7.5 (wider variations
are possible) [9–11,
110000–168000 (type A
gelatin achieves higher
values than type B
[9, 26,
34, 45,
46, 54]
medium to high (but can
be produced to achieve
Bloom values as low as
75 gB)
medium to high medium to high
(type B gelatin
comparatively more
viscous than type A
[56, 61] 5.0–6.5 [10, 61]
isinglass (from
fish swim
c.150000 and higher up
to 300000 [4, 33,
46] medium to high medium to high highest [4, 22,
46, 65] 6.0–7.5 [19, 61,
71, 79]
fish gelatin
(from fish
skin, bone and
96000–196000 [26, 45,
54] low to medium [58] medium medium to high (min.
viscosity between pH
[43, 54] 3.5–5.0 [10, 43,
54, 58]
liquid fish glue 60000 [14, 57] low to medium high (4000–
6000 mPa.s at
[10, 14,
53] 4.0–6.0 (higher pH values
may be possible) [10, 14]
cold liquid
hide glue
n.a. medium high (4000 mPa.s
at manufactured
[10] 6.5 [10]
n.a. data not available
Table 1 (contd.)
PROPERTY mechanical strength [Ref.] elasticity [Ref.] stress development
in fluctuating RH
[Ref.] stability in fluctuating
[Ref.] resolubility with age [Ref.]
increases with increasing content of
helical structures increases with increasing molecular
weight, helicity and solution
increases with increasing
helicity increases with increasing helicity decreases with lower original solution
bone glue low to medium more elastic than hide
glue (but more brittle) medium less stable than hide /
rabbit skin glue more resoluble than hide
glue [90]
hide glue high (tensile strength
typically around 39
[76] less elastic (stiffer) than
bone glue and gelatin
from aquatic sources
[26, 58] high [6, 28] more stable than bone
glue, less sensitive
than cold liquid hide
[6, 28] generally thought to be
rabbit skin glue high, but lower than
other hide glue [23, 39,
78] more elastic than hide
glue [23, 84] high [76, 86] less sensitive to
moisture than hide
[5, 39] generally thought to be
high (low Bloom value
gelatin will achieve lower
mechanical strength)
less elastic (stiffer) than
gelatin from aquatic
[26, 58] medium to high less stable than
isinglass generally thought to be
isinglass (from
fish swim
high [4] more elastic than hide
glue very high [4] higher than
mammalian gelatin [79] (contradictory data) [23, 79]
fish gelatin
(from fish
skin, bone and
medium [26, 27,
53] more elastic than
mammalian gelatin (but
more brittle)
medium to high n.a. generally thought to be
liquid fish glue medium [26, 27,
53] more elastic than hide
glue (but more brittle) medium less stable than cold
liquid hide glue [17] resoluble (after 6 months
RH and temperature
cold liquid
hide glue
n.a. n.a. n.a as stable as hide glue
(after 6 months RH
and temperature
[17] resoluble (after 6 months
RH and temperature
n.a. data not available
and intermolecular bonds have developed within the
network (gel/Bloom strength) further contributes to
higher viscosity [63, 64]. Strongly denatured gelatinous
solutions (such as bone glues) or those affected by a
high degree of molecular cleavage will normally have a
comparatively low viscosity. At a given Bloom strength,
alkaline pre-treated (Type B) gelatins are generally more
viscous than acid pre-treated (Type A) gelatins [56]
(Table 1).
Viscosity is an important factor in the choice of adhesive
for bonding or consolidation, as it will affect the degree
of penetration into a substrate. If the viscosity is too
low the glue may penetrate too far into the substrate,
leaving a joint starved of adhesive. For consolidation of
porous materials, high viscosity may prevent adequate
penetration and cause stress to develop at the interface
between consolidated and unconsolidated areas.
Unfortunately, the viscosity values for animal glues given
in the literature and by suppliers vary widely and are
not easily compared. Measurements were often taken
under different experimental conditions and at different
degrees of cleavage in the protein molecules [4, 21, 35,
37, 46].
Isinglass has a much higher viscosity than hide glue at
an equivalent solution concentration and temperature
(above Tgel), which can be explained by its comparatively
high proportion of high molecular weight fractions
(which, in the following paragraphs, will be referred to
as high Molecular Weight Distribution (MWD)) [4, 22,
46, 65]. This is contrary to what is often stated in the
literature and to the traditional beliefs about the handling
properties of isinglass [20, 21, 61]. However, where low
viscosity values have been obtained for isinglass, it is
likely that the particular preparation procedure of the
glue used for the tests resulted in greater cleavage of
the protein molecules [46]. Despite isinglass having a
large fraction of high MW compounds, its low gelling
temperature compensates for this by allowing more
time for the glue to penetrate porous substrates at room
temperature, therefore improving its penetration ability
in comparison to gelatin and rabbit skin glue of similar
high MW fractions, which will gel faster [8, 66].
In order to obtain glue solutions of low viscosity, it is
not always advisable to dilute viscous high Bloom glues
excessively. The use of an over-diluted glue may result
in swelling, leaching or staining of the substrate if it is
water sensitive [67]. In such cases, a glue with a lower
gel strength would be preferable.
Surface tension
Slow gelation and lower viscosity promote uniform film
formation as the glue is able to spread evenly, providing
adequate wetting of the surface. Wetting is improved
with a decrease in the surface tension of the glue solution.
Sauer and Aldinger [68] confirm that a decrease in
surface tension of a gelatinous solution is directly linked
to the presence of fats. Free fatty acids and neutral
fats are regarded as particularly effective in reducing
surface tension even in small concentrations. With the
exception of rabbit skin glue (which has comparatively
high fat levels of around 5% [9, 11]), most animal glues
and gelatins contain less than 1% fat because of modern
manufacturing methods [9, 10, 54, 58, 69] and may
require additives to reduce the surface tension.
Ethanol is commonly added to lower the surface tension
and improve the wetting abilities of collagen-based glues
[21, 70, 71]. In one case beer containing 9% alcohol
was added to fish glue that was used in the conservation
of Boulle-marquetry, and was shown to improve the
wetting properties leading to stronger joints between
the wood and brass components [70]. However, alcohol
may also raise the gelling temperature, speeding up the
gelation and decreasing the time for which the glue
is workable [28, p. 102, 110], and may also promote
swelling of the substrate. Alternatively, surfactants can
be added to lower the surface tension [3, 8, 28, 72,
p. 123].
Animal glues can have an undesirable tendency to foam,
developing small air bubbles in the glue matrix which
can disrupt the uniformity of the dried glue film and
weaken bonds [5, 59]. Natural fats or free fatty acids
present in glues play a vital role in reducing foaming [5,
68, 73], although some authors still express some doubt
that there is a direct correlation between fat content
and tendency to foam [9]. Nevertheless, Skans [73,
p. 66] suggested that a natural fat content of above 5%
would inhibit the development of pinholes in gesso for
gilding. Sauer and Aldinger [68] have demonstrated an
unambiguous dependency of the degree of foaming on
fat content, whereas no direct relationship could be
established with surface tension. They also could not
find any influence of protein degradation products
on foaming, while pH was established to have an
inconsistent effect.
For conservation applications, the choice of adhesive
may be dependent on the pH sensitivity of the substrate
[71]. Collagen-based glues can display varying pH
values that are difficult to predict purely on the basis
of the glue type or treatment during manufacture. The
assumption that glues which undergo alkaline pre-
treatment display a slightly alkaline pH and acid-treated
ones have an acidic pH [39, p. 171] is incorrect. It is
stated in the literature that hide and fish glue solutions
often have a fairly neutral pH in the range of 6.5 to
7.4, although wider variations are possible [9–11, 16,
35, 68]. In general, bone glues tend to be slightly more
acidic [5, 10, 39], with pH levels between 5 and just
below 7 [35, 37]. Pure gelatins from mammals and fish
range between pH 5.0–6.5 and 3.5–5.0 respectively [10,
53, 54, 58, 61]. Isinglass yields solutions with a pH in
the neutral range [19, 61, 71]. Conservators should test
the pH value of the chosen glue before use if sensitivity
of the substrate is of potential concern.
Apart from being a relevant aspect to consider in
conjunction with the sensitivity of the substrate, pH
values also have an influence on the properties of the
glue, as the viscosity increases when the pH of the
solution shifts away from its isoelectric point (pI) [1, 37,
61, 74]. Since proteins and amino acids are amphoteric
in nature (i.e. containing both acidic and basic functional
groups), they have an isoelectric point, which is the
pH at which all positive and negative charges within
the molecule are balanced and the molecule carries no
net electrical charge. If the electrical potential of the
ions is unbalanced, solution viscosity and Tgel increase,
as well as the capacity for water-sorption and swelling
ability, while gel strength decreases [9, 46, 52, 61, 62].
Commercial animal glues extracted by alkaline pre-
treatment (most hide and bone glues, type B gelatins)
usually have a pI of approximately 4.5 to 5.5, whereas
glues derived from acid pre-treated collagen sources
commonly display pI values of between 7.0 to 9.0 [17,
30, 34, 35, 51, 56, 75]. For practical purposes, this
means that glues having a pH near their pI value (such
as bone glues and type B gelatins) will already be at the
lowest possible viscosity, as opposed to those which
have pH values different from their pI, where to achieve
the lowest possible viscosity the pH would have to be
modified to take it closer to the pI (Table 1). The effect
of the pH of a glue solution on its surface tension is
inconsistent [28, p. 75, 68].
Mechanical properties of the dried film
Cohesion, adhesion and bond strength
The cohesive strength of the gelatinous matrix of a glue is
determined by its molecular structure and intermolecular
bonding, as expressed by the Bloom value. To produce an
animal glue film that is as strong as possible in the dried
state, the same rules apply as for obtaining a high gel
strength (i.e. high MW distribution/minimum cleavage
of protein molecules, maximum renaturation/content of
collagen-like triple-helices, high intra-/intermolecular
stabilisation). The cohesion strength of animal glues can
be improved by the addition of a suitable amount of an
alcohol, such as ethanol or glycerine [28, p. 108]. To
achieve strong bonding, chemical adhesion between the
glue and the substrate is as important as high cohesion
within the glue matrix.
Hide glues generally have greater cohesive strength than
the bone glues with highly cleaved molecules, which
display a lower tensile strength and are much more
brittle (Table 1). The tensile strength of hide glues is
typically around 39 megapascals (MPa) (5700 pounds
per square inch, psi) [76]. Mammalian collagen tends
to yield stronger glues than most aquatic sources, owing
to the reduced number of stabilising inter- and intra-
molecular bonds in fish collagen [33, 49]. Cold water
fish gelatins in particular have a lower propensity to
reform helical structures due to their small proportion of
the amino acid derivatives, Hyp and Pro, and therefore
show a comparatively low tensile strength of around 22
MPa (3200 psi) [26, 27, 53]. This value is comparable
to the strength of bovine bone gelatin [54].
A high tensile strength similar to that of hide glue
has been reported for mildly prepared isinglass from
sturgeon [4], making it a useful adhesive for bonding
wooden joints. The literature confirms that isinglass
has often been used for structural woodwork in the Far
East [24, 77]. Although rabbit skin glue has a high gel
strength, it has been stated as having lower cohesion
and bonding strength than other hide glues [23, 39,
78]. This is thought to be due to its high fat content
[9, 23].
Elasticity, resistance to impact (toughness) and creep
As for many of the other physical properties of gelatin-
based glue films, the elasticity and stiffness are greatly
dependent on their MW distribution [63], the degree
to which helical structures reform on gelling and the
intra-/intermolecular bonding [7, 26]. The stiffness of
the glue (elastic modulus, known as Young’s modulus
E, mathematically calculated from the ratio of stress to
strain values) increases with a higher ratio of high MW
fractions, higher solution concentrations and with a
greater renaturation level in the network [26, 45, 55].
Stabilisation of the gel network by increased electrostatic
bonding induced by pH levels above or below the pI
also increases the stiffness [42]. Mammalian gelatin
generally has a higher modulus and therefore greater
stiffness than fish gelatin due to its higher network
stabilisation by intra- and intermolecular bonding [26,
58]. Isinglass is also more elastic than mammalian
gelatin [79].
The moisture content of an animal glue has an important
effect on the mechanical properties. Under normal
ambient conditions (50% RH and room temperature)
gelatin-based glue films contain 12–14% of structural
water bound to the polar groups of the protein
macromolecules [52, p. 654]. This water contributes to
the stabilisation of the helical structures within the glue
and a specific amount of water is needed to maintain
structural stability. Above around 25% moisture content
the glue turns from a glassy to a rubbery state at room
temperature [52]. Excessive dehydration of gelatinous
films below a moisture content of 0.2% leads to the
development of covalent cross-links between the protein
molecules, which ultimately renders the glue insoluble
in water [80, p. 509].
In general, gelatinous glue films with a low moisture
content are very brittle regardless of the collagen
source and molecular structure [34, p. 63, 52]. Even
at a normal (12–14%) water content, gelatinous films
undergo brittle fracture under impact. Randomly coiled
structures exhibit much lower resistance to impact
(greater brittleness) than helical glue matrices in the glassy
state. Glue recipes often contain additives such as sugar
alcohols (e.g. glycerine, sorbitol) and polysaccharides
(e.g. dextrins) to improve elasticity and toughness
[28, 34, 81, 91]. One traditional recommendation for
achieving elastic and resilient glue films is the addition
of honey [4, 18, 21, 22, 61, 82]. Sugars are hygroscopic
and so stabilise the protein molecules by introducing
additional hydrogen bonds involving water [25, 83],
inducing an increase in gel strength and viscosity.
Although these additives do not actually plasticise the
glue matrix, they are often referred to as plasticisers in
the literature. A high proportion of fat also improves
elasticity, although it simultaneously reduces the gel
strength of the glue and final bond strength [23, 84].
A higher water content or an excess of hygroscopic
additives generates a reduction in the glass transition
temperature of the glue [61, 81], which can promote an
unwanted tendency to creep (elongation with time).
Stability in ambient environment and sensitivity to
fluctuating levels of moisture and heat
Drying of collagen-derived glue films leads to the
development of high internal stress and tensile forces
within the glue matrix, while increasing humidity
generally causes progressive loss of tension [4, 85]. This
behaviour is dependent on the physical and chemical
structure of the glue. A high degree of collagen-like
triple-helix arrangement in a gelatin film has been
shown to result in a reduced tendency to swell [34, 55],
but is also responsible for increasing stress values due
to stronger cohesion. Isinglass from sturgeon, which
contains a high proportion of helical structures (due to
its high MWD, despite its lower Hyp and Pro content),
develops particularly high stress levels, which it is
suggested are twice as high as in hide glue [4].
If kept under moderate relative humidity conditions over
a long period of time, initial stresses within gelatinous
films relax owing to the absence of covalent cross-links
[86]. Under fluctuating environmental conditions, the
mechanical properties of collagen-based glues are subject
to continual change [85]. Considerable development of
internal stresses will affect the glue’s elasticity, strength
and physical stability and may lead to significant damage
to the substrates [48, 85–87].
At high RH levels (above 85%) animal glue films undergo
a continuing reformation of helical structures. This
will result in new, higher stresses on subsequent drying
and can lead to severe shrinkage due to contraction of
the glue matrix [48, 85, 86]. Cycling of RH can cause
further strain – for rabbit skin glue, non-permanent total
dimensional changes of up to 6% have been reported
as the result of a single RH cycle [76, 86], which are
only partly recoverable. According to Zumbühl [48],
contraction mechanisms compete with plastic relaxation
processes above 65% RH. However, even at high RH
levels plastic relaxation may not sufficiently compensate
for these stresses and continuous cycling further reduces
the ability for stress relaxation [48]. This will result in
permanent shrinkage of the glue matrix (up to 5% for
rabbit skin glue [76]) and, in this case, loss of tension is
only possible by substrate deformation or mechanical
destruction (embrittlement) of the glue film [48].
At low RH levels, more randomly coiled gelatinous
structures (such as bone glues), which have comparatively
low tensile strength and low resistance to stress, induce
relaxation at an early stage by developing cracks in the
glue matrix, thereby preventing high stresses on the
substrate. These glues also show a greater tendency to
creep under stress at high RH levels [61, p. 14]. Although
animal glues containing a high degree of helical structure
exhibit comparatively high stress when exposed to
extreme and fluctuating environmental conditions, they
still display greater stability in their strength properties
than more randomly coiled structures. The strength
properties of hot hide glue have been shown to be less
sensitive to fluctuating RH and temperature than those
of cold liquid hide glue [6, 28]. Liquid fish glues are even
less stable than cold liquid hide glues under fluctuating
conditions [17]. It has also been suggested that a high fat
content, such as in rabbit skin glue, accounts for better
stability in moist conditions [5, 39]. Common methods
for improving the glue film’s hardness and resistance
to water are the addition of tanning agents, such as
aluminium trisulphate (alum), disodium triborate
(borax), sodium acetate or formaldehyde [9, 28, 35, 91].
These salts remove a certain amount of bound water
from the proteinaceous matrix by covalently bonding
to the hydrophilic sites in the glue, thus inducing the
formation of numerous new cross-links between the
protein molecules.
Mechanical properties of animal glues used as gap fillers
Although the excessive shrinkage and brittleness of
animal glues at low RH [88] makes them inferior gap
fillers on their own, modification with ‘plasticisers’ and
bulking agents can alter their properties, improving
their suitability for this application [39, 81]. Hard films
with a minimum tendency to distort can be achieved
by the addition of fillers such as magnesium sulphate
or mineral clays together with sugars and dextrins [28,
The addition of an inert filler dramatically changes
the physical and mechanical performance of animal
glues, depending on the proportion of glue present.
A high pigment concentration significantly reduces
intermolecular bonding within the glue medium [76,
86] and thus impedes dimensional changes of the
matrix in response to relative humidity changes [76]. In
addition, with the lack of chemical adhesion between a
proteinaceous binder and inert filler particles, the glue
is substantially weakened and this leads to low tensile
strength [7]. Therefore high MW glues, with their long
protein strands and ability to develop stabilising H-
bonds, are appropriate for fillers and gesso with a high
pigment concentration.
Ageing characteristics
Whilst substantial research has been published on the
behaviour of collagen-derived glues in a fluctuating
environment, information on the ageing mechanisms
and behaviour on exposure to light seems to be more
limited. According to Michel et al., isinglass from
sturgeon, of all animal glues, best retains its mechanical
properties with thermal and ultraviolet (UV) light ageing
and RH cycling [79, p. 271]. It shows markedly less
change in strength and stiffness than pure mammalian
gelatin. Mammalian gelatin increases in tensile strength
but becomes stiffer and more brittle upon artificial
ageing under UV light, fluctuating RH and temperature.
Isinglass from sturgeon remains much tougher and more
elastic than gelatin [79, p. 274]. It also develops the least
permanent dimensional change, whereas gelatin films
swell or creep slightly during ageing, and other animal
glues show an even more marked effect.
Collagen-derived glues, unless they have been modified
by the addition of tanning agents which causes them
to become relatively resistant to water, generally swell
readily when exposed to water and redissolve when
heated, even after centuries [23, 39]. Neher [89]
established that the Bloom strengths of hide and rabbit
skin glues are not correlated to their water-resolubility
and that all tested samples were completely and equally
successfully reversible after one month of natural
drying. Wooden joints bonded with fish glue or cold
liquid hide glue have also been shown to be detachable
with water after six months of natural ageing or RH and
temperature cycling [17]. An effect of the tannic acids of
oak wood and walnut on their resolubility could not be
established in this study.
The dependence of water-resolubility on original
solution concentration has been demonstrated for aged
and UV-irradiated hide and bone glues at concentrations
of between 2.5 and 20% [90, p. 302]. This research
showed that the lower the original concentration, the
lower the resolubility of the glue film. Bone glues were
more resoluble than hide glues, supposedly because
of their more pronounced molecular cleavage in the
protein matrix (Table 1).
Przybylo tested isinglass from sturgeon obtained from
different suppliers [23], and found that the source,
origin and preparation temperature have no significant
effect on the resolubility of the glue in water after
natural and artificial ageing, as all the films in the test
series remained resoluble. In contrast, Michel et al. [79]
report that their artificially-aged sturgeon isinglass films
were insoluble in water, even though no significant
molecular changes within the protein were detected.
The contradictory results of these two studies may be
due to different preparation procedures and artificial
ageing conditions, which varied in the type of light
source as well as cycles of exposure time, temperature
and RH.
Resolubility of animal glues may be reduced in cases
where the protein has come into contact with metal
ions (e.g. metal foils, tools, pigments), or with certain
organic pigments and tannins, either before, during or
even after their application [12, 23, 78]. Resolubility
of collagen-derived glue containing no additives is thus
very much dependent on the environment to which it
has been exposed, rather than being predetermined by
the type of glue. Cold liquid hide and fish glues, the
ingredients of which are often unknown to the supplier
and end user, may already contain additives that promote
cross-linking and, therefore, increase insolubility.
Colour changes on ageing
Hide and bone glues are generally much more strongly
coloured (amber to brown) and less transparent than
gelatin or isinglass because of their higher impurity
content. Higher levels of denaturation and molecular
cleavage also intensify the colour of gelatinous solutions
[47]. This phenomenon may be responsible for the
general observation that the higher the Bloom value, the
less yellow the gelatin [56]. Gelatin and isinglass appear
clear and virtually colourless if dried to thin films, even
though they yield slightly yellow or whitish solutions
[10, 14, 43, 56, 58, 61]. They are also very light fast
and show hardly any discolouration or yellowing with
age [8, 30, 79], which is why they are the only collagen-
derived glues suitable for pigment consolidation.
Isinglass is particularly popular for this purpose, as its
low refractive index, when compared with mammalian
gelatin, causes the least change in appearance of the
pigments after drying [71, 79].
This review of the different types of currently available
animal glue has shown that collagen-derived adhesives
vary in their chemical, physical and mechanical
properties. Being a natural polymer, performance is
partly dependent on the original collagen source, which
determines the glue’s chemical composition, but is also
strongly affected by the extraction and preparation
procedures. Molecular weight distribution is an
important factor which directly influences the protein
solution viscosity and contributes to gel strength and
Tgel. The degree of stabilisation of the protein matrix
by hydrogen and other chemical bonding is determined
by amino acid composition, preparation procedures
and drying time. This has an even greater impact on
the performance of the glue, and significantly affects its
strength, mechanical behaviour, sensitivity to ambient
environment and stability with age. Changes in pH
and the addition of hygroscopic additives (plasticisers)
and salts can alter many of these properties. However,
manipulation of one individual factor cannot necessarily
be realised without simultaneously changing a whole
range of other properties. As most of the properties are
dependent on each other, selection of the appropriate
glue should be based on a correct balance rather than on
individual properties.
It has become evident that much important data that
would allow comparison of the properties of the different
types of glues is still missing. Very few gelatinous
glues have been prepared and tested under the same
conditions, and insufficient characterisation of these
glues makes it difficult to draw exact conclusions for a
general glue type, as physical and mechanical properties
can vary substantially. However, a summary of the data
does reveal general qualitative trends that can be used
by conservators to make well-informed decisions on the
suitability of a particular collagen-based glue for a given
The author would like to thank Shayne Rivers, Senior
Furniture Conservator at the Victoria and Albert
Museum, London, and Dr Ambrose C. Taylor, Imperial
College, London, for their ongoing support in discussing
this paper and their valuable advice.
1 Greber, J.M., Lehmann, E., and van der Werth, A., Die
Tierischen Leime. Geschichte – Herstellung – Untersuchung,
Verwendung und Patentübersicht, Heidelberg (1950).
2 Willers, H., Herstellung von tierischem Leim und seine
Verwendung im Bereich der Tafel- und Fassmalerei nach
Angaben deutschsprachiger Quellenliteratur des 16. bis Mitte
des 19. Jhds., diploma thesis, Akademie der Bildenden Künste
Stuttgart (1980).
3 von Reventlow, V., ‘The treatment of gilded objects with
rabbit-skin glue size as consolidating adhesive’, in Gilded
Wood Conservation and History, ed. D. Bigelow, Sound View
Press (1991) 269–275.
4 Luybavskaya, E.A., ‘Investigation of properties of protein
glues’, in ICOM-CC 9th Triennial Meeting, Dresden, Preprints,
ed. K. Grimstad, International Council of Museums, Los
Angeles (1990) Vol. 1 47–50.
5 Skans, B., Analysis and properties of old animal glues’,
in 7. IADA International Congress of Restorers of Graphic
Art, Uppsala, Sweden, ed. K.J. Palm and M.S. Koch,
Internationale Arbeitsgemeinschaft der Archiv-, Bibliotheks-
und Grafikrestauratoren (1991) 43–50.
6 Buck, S.L., A study of the properties of commercial liquid
hide glue and traditional hot hide glue in response to changes
in relative humidity and temperature’, in AIC Wooden
Artifacts Group Session, American Institute for Conservation,
Richmond, Virginia (1990).
7 von Endt, D.W., and Baker, M.T., ‘The chemistry of filled
animal glue systems’, in Gilded Wood Conservation and
History, ed. D. Bigelow, Sound View Press (1991) 155–162.
8 Nicolaus, K., Handbuch der Gemälderestaurierung,
Könemann Verlag, Köln (n.d.) 230.
9 Wilde, A., ‘Zur heutigen Herstellung von Glutinleim’,
Zeitschrift für Kunsttechnologie und Konservierung 20(2)
(2006) 379–406.
10 Kremer-Pigmente, ‘Bone glue (63000), hide glue (63010–
63020), rabbit skin glue (63025, 63028, 23052), gelatin
(63040), isinglass (63100), Salianski-isinglass 63110,
fish glue (63550), Franklin Hyde Glue (63500 63512)’,
product data sheets, Kremer Pigmente GmbH & Co. KG,
Aichstetten, Germany (January 2007).
11 Natural-Pigments, ‘Rabbit skin glue’, http://www.
21RSGLU (accessed 7 February 2007).
12 Doerner, M., Malmaterial und seine Verwendung im Bilde,
18th edn, Enke Verlag, Stuttgart (1994) 98, 180.
13 Hellmann, A., Hellmann Leim GmbH, Memmingen,
Germany, personal communication (February 2007).
14 Deffner and Johann, ‘Leime’, in Gesamtkatalog 2007/2008
Deffner & Johann GmbH, Schweinfurt (2006) 186–188.
15 Morita, T., ‘ ‘Nikawa’ – traditional production of animal
glue’, in Adhesives and Consolidants, Preprints of the
contributions to the Paris Congress, 2–8 September 1984,
ed. N.S. Brommelle, E.M. Pye, P. Smith and G. Thomson,
International Institute for Conservation, London (1984)
16 Walsh, H.C., ‘Fish glue’, in Handbook of Adhesives, 4th edn,
ed. I. Skeist, Reinhold Publishing Corp., New York (1965)
17 Coerdt, A., ‘Zum Leimen zu gebrauchen – Untersuchungen
zu kaltflüssigen Glutinleimen – Teil 1’, Restauro 113(1)
(2007) 32–38 and Teil 2, Restauro 113(1) (2007) 191–197.
18 Nicolaus, K., Handbuch der Gemäldekunde, DuMont Verlag,
Köln (2003) 226.
19 Petukhova, T., ‘Potential applications of isinglass adhesive
for paper conservation’, The Book and Paper Group Annual
of the American Institute of Conservation 8 (1989).
20 Foskett, S., ‘An investigation into the properties of isinglass’,
Scottish Society for Conservation and Restoration Journal 5
(4) (1994) 11–14.
21 Petukhova, T., and Bonadies, S.D., ‘Sturgeon glue for painting
consolidation in Russia’, Journal of the American Institute for
Conservation 32 (1) (1993) 23–31.
22 Habel-Schablitzky, A., Fischblasenleim Geschichte und
Eigenschaften sowie Anwendung in der Holzrestaurierung,
diploma thesis, Fachhochschule Köln (1992).
23 Przybylo, M., ‘Langzeit-Löslichkeit von Störleim – Tatsache
oder Märchen?’, VDR Beiträge zur Erhaltung von Kunst- und
Kulturgut 1 (2006) 117–123.
24 Lin, S.Y., ‘Fish air-bladders of commercial value in China’,
The Hong Kong Naturalist 9 (3) (1939) 108–118.
25 Fernández-Díaz, M.D., Montero, P., and Gómez-Guillén,
M.C., ‘Gel properties of collagen from skins of cod (Gadus
morhua) and hake (Merluccius merluccius) and their
modification by their coenhancers magnesium sulphate,
glycerol and transglutaminase’, Food Chemistry 74 (2001)
26 Simon, A., Grohens, Y., Vandanjon, L., Bouseau, P., Balnois,
E., and Levesque, G., ‘A comparative study of the rheological
and structural properties of gelatin gels of mammalian and
fish origin’, Macromolecular Symposia 203 (2003) 331–
27 Gilsenan, P.M., and Ross-Murphy, S.B., ‘Rheological
characterisation of gelatins from mammalian and marine
sources’, Food Hydrocolloids 14 (2000) 191–195.
28 Pitzen, C., Die Modifizierung von Glutinleimen
Möglichkeiten der Anpassung an objektspezifische und
verarbeitungstechnische Bedingungen nach Literaturangaben
des 18. bis 19. Jahrhunderts, diploma thesis, Fachbereich
Konservierung und Restaurierung, Möbel/Holz,
Fachhochschule Köln (1991).
29 Woodhead-Galloway, J., Collagen: the Anatomy of a Protein,
London (1980) 23–24.
30 Fuchs, R., ‘Pergament – Material, Geschichte, Restaurierung’,
in Pergament: Geschichte Material Konservierung
Restaurierung, ed. R. Fuchs, C. Meinert and J. Schrempf,
Fachbereich Restaurierung / Konservierung, Fachhochschule
Köln (2001) 10–12.
31 Burjanadze, T., ‘Thermodynamic substantiation of water-
bridged collagen structure’, Biopolymers 32 (8) (1992) 941–
32 Holmgren, S.K., Taylor, K.M., and Bretscher, L.E., ‘Code for
collagen’s stability deciphered’, Nature 6677 (392) (1998)
33 Hickman, D., Sims, T.J., Miles, C.A., Bailey, A.J., de Mari,
M., and Koopmans, M., ‘Isinglass/collagen: denaturation and
functionality’, Journal of Biotechnology 79 (2000) 245–257.
34 Rose, P.I., ‘Gelatin – general properties’, in The Theory of the
Photographic Process, 4th edn, ed. T.H. James, Macmillan
Publishing Co., New York (1977) 51–67.
35 Hubbard, J.R., ‘Animal glues’, in Handbook of Adhesives,
4th edn, ed. I. Skeist, Reinhold Publishing Corp., New York
(1965) 114–125.
36 Johns, P., and Courts, A., ‘The relationship between collagen
and gelatin’, in Food Science and Technology of Gelatin, ed.
A.G. Ward and A. Courts, Academic Press, London (1977)
37 Hull, W.Q., and Bangert, W.G., Animal glue – a staff-industry
collaborative report…’, Industrial and Engineering Chemistry
44 (10) (1952) 2275–2284.
38 von Endt, D.W., ‘Technological modifications of protein
materials and the effect on stability. Protein adhesives’, in
Protein Chemistry for Conservators, ed. C.L. Rose and D.W.
von Endt, American Institute for Conservation, Washington
D.C. (1984) 39–46.
39 Rivers, S., and Umney, N., Conservation of Furniture,
Butterworths-Heinemann, Oxford/Auckland (2003) 156–
173, 442.
40 Privalov, P.L., ‘Stability of proteins’, Advances in Protein
Chemistry 33 (1979) 167–241.
41 Ramachandran, G.N., ‘Modification of collagen and gelatin
by chemical reagents’, in Recent Advances in Gelatin and
Glue Research, ed. G. Stainsby, Pergamon Press, London/
New York (1957) 32.
42 Haug, I.J., Draget, K.I., and Smidsrød, O., ‘Physical and
rheological properties of fish gelatin compared to mammalian
gelatin’, Food Hydrocolloids 18 (2004) 203–213.
43 Norland-Products, ‘Fish glue, technical abstract’, http://www. (accessed 7
February 2007).
44 Ogawa, M., Moody, M.W., Portier, R.J., Bell, J., Schexnayder,
H.A., and Losso, J.N., ‘Biochemical properties of black
drum and sheepshead seabream skin collagen’, Journal of
Agricultural and Food Chemistry 51 (2003) 8088–8092.
45 Badii, F., and Howell, N.K., ‘Fish gelatin: Structure, gelling
properties and interaction with egg albumen proteins’, Food
Hydrocolloids 20 (5) (2006) 630–640.
46 Haupt, T., ‘Zubereitung von Störleim Auswirkungen
der Zubereitungstemperatur und -zeit auf Viskosität,
Gelierverhalten und Molekulargewicht‘, Zeitschrift für
Kunsttechnologie und Konservierung 18 (2) (2004) 318–
47 Gelita, ‘Dissolution of gelatine’,
(accessed 17 February 2007).
48 Zumbühl, S., ‘Proteinische Leime – ein vertrauter Werkstoff?
Aspekte zum feuchtephysikalischen Verhalten von Gelatine’,
Zeitschrift für Kunsttechnologie und Konservierung 17 (1)
(2003) 95–103.
49 Gómez-Guillén, M.C., Turnay, J., Fernández-Díaz, M.D.,
Ulmo, N., Lizarbe, M. A., and Montero, P., ‘Structural and
physical properties of gelatin extracted from different marine
species: a comparative study’, Food Hydrocolloids 16 (2002)
50 Hörmann, H., and Schlebusch, H., ‘Reversible and irreversible
denaturation of collagen fibres’, Biochemistry 10 (6) (1971)
51 Alleavitch, J., Turner, W.A., and Finch, C.A., ‘Gelatin’, in
Ullmann’s Encyclopedia of Industrial Chemistry, 5th edn, ed.
B. Elvers, S. Hawkins, M. Ravenscroft, J.F. Rounsaville and
G. Schulz, VCH-Verlag, Weinheim (1989) 307–317.
52 Kozlov, P.V., and Burdygina, G.I., ‘The structure and properties
of solid gelatin and the principles of their modification’,
Polymer Reviews 24 (1983) 651–666.
53 Norland-Products, ‘High tack fish glue, product details’,
(accessed 7 February 2007).
54 Muyonga, J.H., Cole, C.G.B., and Duodu, K.G., ‘Extraction
and physico-chemical characterisation of Nile perch (Lates
niloticus) skin and bone gelatin’, Food Hydrocolloids 18
(2004) 581–592.
55 Bigi, A., Panzavolta, S., and Rubini, K., ‘Relationship between
triple-helix content and mechanical properties of gelatin
films’, Biomaterials 25 (2004) 5675–5680.
56 Gelita, ‘GELITA® Gelatine – für Speise, Pharma und Foto’,
product brochure, Gelita – The Gelatine Group, (n.d.).
57 Kremer-Pigmente, ‘Natürliche Leime und Aquarellbinder’,
in Kremer Pigmente Preisliste 2005/2006, Kremer Pigmente
GmbH & Co. KG, Aichstetten (2005) 58–61.
58 Cheow, C.S., Norizah, M.S., Kyaw, Z.Y., and Howell, N.K.
‘Preparation and characterisation of gelatins from the
skin of sin croaker (Johnius dussumieri) and shortfin scad
(Decapterus macrosoma)’, Food Chemistry 101 (2007) 386–
59 Klinger, J., and Thomas, R., Die Kunst zu Vergolden. Beispiele –
Techniken – Geschichte, Callwey, München (1989).
60 Lee Valley, High Tack Fish Glue, material safety data sheet,
Lee Valley Tools Ltd.,,
Ottawa (January 2007).
61 Haupt, M., Dyer, D., and Hanlan, J., An investigation into
three animal glues’, The Conservator 14 (1990) 10–16.
62 Courts, A., ‘Citrate-promoted helix formation in gelatin’,
Biochemical Journal 83 (1962) 124–129.
63 Normand, V., Muller, S., Ravey, J.C., and Parker, A., ‘Gelation
kinetics of gelatin: A master curve and network modeling’,
Macromolecules 33 (2000) 1063–1071.
64 Yannas, I.V., and Huang, C., ‘Viscoelastic distinction between
helical and coiled macromolecules’, Macromolecules 5 (1)
(1972) 99–100.
65 Leuenberger, B.H., ‘Investigation of the viscosity and gelation
properties of different mammalian and fish gelatins’, Food
Hydrocolloids 5 (1991) 353–361.
66 Webb, M., Lacquer Technology and Conservation,
Butterworth-Heinemann, Oxford (2000) 80–81, 165.
67 Kato, H., ‘The restoration of urushiware for export
with animal glue and urushi’, in Japanese and European
Lacquerware, ed. M. Kühlenthal, München (2000) 81–84.
68 Sauer, E., and Aldinger, W., ‘Oberflächenspannung und
Schaumbildung bei Glutinlösungen’, Kolloid-Zeitschrift 88
(3) (1939) 329–340.
69 Amstel-Products, ‘Product description for bone glue and
technical gelatin/hide glue’, http://www.amstelproducts.
nl/technicalProt/Specification/bone_glue.htm; http://www.
gelatin.htm (accessed 13 February 2007).
70 Triboulot, M.-C., Restauration de panneaux décoratifs –
première partie: Amélioration du collage bois-laiton, ENSTIB
project report, Université Henri Pointcaré Nancy (1998).
71 Fuchs, R., ‘Neue Methoden in der Papierrestaurierung:
Fixierung loser Partikel’, in 8th IADA International Congress
of Restorers of Graphic Art, Tübingen, 19–23 September 1995,
Internationale Arbeitsgemeinschaft der Archiv-, Bibliotheks-
und Grafikrestauratoren (1995) 163–166.
72 Sandner, I., Bünsche, B., Meier, G., Schraum, H.P., and
Voss, J., Konservierung von Gemälden und Holzskulpturen,
Deutscher Verlag der Wissenschaften, Berlin (1990).
73 Skans, B., and Michelsen, P., ‘Die Bedeutung von Fett in
Tierleim für Malzwecke’, Maltechnik-Restauro 92 (2) (1986)
74 Stainsby, G., ‘Viscosity of dilute gelatin solutions’, Nature
169 (4303) (1952) 662–663.
75 Fujita, S., Organic Chemistry of Photography, Springer,
Berlin/Heidelberg (2004).
76 Mecklenburg, M.F., ‘Some mechanical and physical properties
of gilding gesso’, in Gilded Wood Conservation and History,
ed. D. Bigelow, Sound View Press (1991) 163–170.
77 Burgio, L., Rivers, S., Higgitt, C., Spring, M., and Wilson,
M., ‘Spherical copper resinate on Coromandel objects:
analysis and conservation of matt green paint’, Studies in
Conservation 52 (2007) 241–254.
78 Wehlte, K., Werkstoffe und Techniken der Malerei,
Neubearbeitung, Ravensburg (2000) 55.
79 Michel, F., Geiger, T., Reichlin, A., and Teoh-Sapkota, G.,
‘Funori, ein japanisches Festigungsmittel für matte Malerei’,
Zeitschrift für Kunsttechnologie und Konservierung 16 (2)
(2002) 257–275.
80 Yannas, I.V., and Tobolsky, A.V., ‘Cross-linking of gelatine by
dehydration’, Nature 215 (1967) 509–510.
81 Thornton, J., ‘A brief history and review of the early practice
and materials of gap-filling in the West’, Journal of the
American Institute for Conservation 37 (1998) 3–22.
82 Petukhova, T., ‘A history of fish glue as an artist’s material:
Applications in paper and parchment artifacts’, The Book and
Paper Group Annual of the American Institute of Conservation
19 (2000) 19–29.
83 Choi, Y.H., Lim, S.T., and Yoo, B., ‘Measurement of dynamic
rheology during ageing of gelatine-sugar composites’,
International Journal of Food Science and Technology 39
(2004) 935.
84 Djagny, K.B., Wang, Z., and Xu, S., ‘Conformational changes
and some functional characteristics of gelatin esterified with
fatty acid’, Journal of Agriculture and Food Chemistry 49
(2001) 2987–2991.
85 Hedley, G., ‘Relative humidity and the stress/strain response
of canvas paintings: uniaxial measurements of naturally aged
samples’, Studies in Conservation 33 (1988) 133–148.
86 Mecklenburg, M.F., Tumosa, C.S., and Erhardt, D.,
‘Structural response of painted wood surfaces to changes
in ambient relative humidity’, in Painted Wood – History
and Conservation, ed. V. Dolge and F.C. Howlett, Getty
Conservation Institute, Los Angeles (1998) 464–483.
87 Mecklenburg, M.F., ‘The effects of atmospheric moisture
on the mechanical properties of collagen under equilibrium
conditions’, in 16th AIC Annual Meeting, New Orleans, 1–5
June 1988, American Institute for Conservation, Washington
D.C. (1988) 231–244.
88 Grattan, D.W., and Barclay, R.L., ‘A study for gap-fillers
for wooden objects’, Studies in Conservation 33 (1988)
89 Neher, A.L., Investigation into the Reversibility of Distinct
Strengths of Animal Glues and Five Different Methods of
Reversing in Wooden Joints, final year research project, Royal
College of Art and Victoria and Albert Museum (RCA/V&A)
Joint MA Conservation Programme, London (1993).
90 Fiedler, I., and Walch, K., ‘Fluoreszenzunterschiede von
Leimen an Furnieren’, in Lacke des Barock und Rokoko/
Baroque and Rococo Lacquers Arbeitsheft 81 des Bayerischen
Landesamtes für Denkmalpflege, ed. K. Walch and J. Koller,
München (1997) 297–304.
91 Ulmer, R., and Westebbe, P.H., Modifizierte Glutinklebstoffe,
term paper, Studiengang Restaurierung, Kunsttechnologie
und Konservierungswissenschaften, Technische Universität
München, 2002.
Nanke Schellmann trained as a violin maker in
Mittenwald (Bavaria) before undertaking several years
of internships in the conservation departments of the
National Gallery (Frames) and the Wallace Collection
in London, the Bavarian National Museum, Munich
and the Germanic National Museum, Nuremberg. In
2003, she received an MA in Furniture Conservation
from the Royal College of Art/Victoria and Albert
Museum (RCA/V&A) Joint Conservation Programme,
London, UK. On finishing, she joined the workshop
of Clemens von Schoeler, Munich as a conservator for
furniture and historic wooden interiors. Since 2005
she has attended additional courses in natural sciences
at the Ludwig-Maximilians-University, Munich and is
currently undertaking a PhD at the University of Fine
Arts Dresden, together with the V&A Mazarin Chest
Project and Imperial College, London, in the field of
oriental lacquer conservation.
Correspondence can be sent to:
Nanke Schellmann
Mazarin Chest Project
Furniture, Textiles and Frames Conservation
Victoria and Albert Museum
South Kensington
London SW7 2RL
... The results gave a first indication of the variability in properties exhibited by materials sold under the same trade name. An exhaustive review of animal glue properties by Schellman [1] , which includes a range of academic references, summarizes existing knowledge about these * Corresponding author. ...
... The response of the RSG film to increasing relative humidity is characterized by a gradual decrease in storage modulus E' . This trend directly translates to a loss in stiffness of the material when exposed to high humidity and agrees well with results reported in Schellmann's review [1] . This behavior is observed for all tested glues at each of the four selected temperatures (see Figure S.2 in the Supplementary Information). ...
Full-text available
In this work, the mechanical properties of seven animal glues commercially available for conservation practice are compared. The performance of these materials within ranges of temperature and relative humidity commonly found in uncontrolled environments was quantified on cast glue films using dynamic mechanical analysis (DMA) and tensile testing. The applied experimental protocol resulted in the identification of phase transitions between the glassy and rubbery states and the evaluation in those states of the stiffness, viscoelasticity, strength, and strain at break of each tested adhesive. The presented results allow for the prediction of environmental conditions at which specific glues may fail and may help to select the most appropriate adhesive, from a mechanical point of view, for particular applications.
... Collagen, the highest abundant structural protein, has a broad amino acid composition strictly different from other proteins and has been proved to have a unique triple-helical structure [153]. The collagen protein forms elongated triple-stranded helical coils rather than individual protein chains being coiled globules [154]. These coiled proteins that make up the collagen fibrils give good strength and good flexibility. ...
... The origin of the collagen and the preparation method determine the physical, chemical, and mechanical properties of the adhesive. However, the difference in source becomes less crucial since the chains are shortened to lower the molecular weight of coiled proteins [154]. Unless the addition of tanning agents has altered them, which drives them to become relatively water-resistant, collagen-based adhesives usually swell when exposed to water and redissolve when heated, even after centuries [155]. ...
Full-text available
The industrial market depends intensely on wood-based composites for buildings, furniture, and construction, involving significant developments in wood glues since 80% of wood-based products use adhesives. Although biobased glues have been used for many years, notably proteins, they were replaced by synthetic ones at the beginning of the 20th century, mainly due to their better moisture resistance. Currently, most wood adhesives are based on petroleum-derived products, especially formaldehyde resins commonly used in the particleboard industry due to their high adhesive performance. However, formaldehyde has been subjected to strong regulation, and projections aim for further restrictions within wood-based panels from the European market, due to its harmful emissions. From this perspective, concerns about environmental footprint and the toxicity of these formulations have prompted researchers to re-investigate the utilization of biobased materials to formulate safer alternatives. In this regard, proteins have sparked a new and growing interest in the potential development of industrial adhesives for wood due to their advantages, such as lower toxicity, renewable sourcing, and reduced environmental footprint. This work presents the recent developments in the use of proteins to formulate new wood adhesives. Herein, it includes the historical development of wood adhesives, adhesion mechanism, and the current hotspots and recent progress of potential proteinaceous feedstock resources for adhesive preparation.
... Glues could have been added through successive repairs and restorations. Different kinds of glue were chosen according to their properties and are still currently used by restorers following traditional practices including fish glue (Schellmann, 2007). From the 20 th c. onwards, synthetic water-based vinyl glues could have been used. ...
Full-text available
The PhD study presented here explores the feasibility and the limits of adapting two advanced techniques currently used in archaeometry: 14C geochronology (MICADAS) and isotope geochemistry (MC-ICP-MS NEPTUNE), to the issues of the instrumentarium of the modern period kept in museum collections. The research focused on linseed oil/colophony varnishes; the material corpus was extended to varnishes of pieces of furniture and horse-drawn carriages mainly from the early 18th c. and also includes early gut strings as a testimony of the material history of uses. The 14C geochronology study led to the development of sampling protocols and chemical treatments for analyzing materials according to their nature. The study also provided new insights to deepen our knowledge of the analyzed objects and their uses (impact of restorations, reuse of materials, ingredients, etc.). The Sr isotope study (geographical provenance) was tested on organic binders from the current market. The conservation of the 87Sr/86Sr signature from the bedrocks to the pine resin and colophony (after distillation) was validated for a corpus of pine soil/resin/colophony with a known provenance. The selected colophony provided 87Sr/86Sr measurements consistent with geochemical maps. The state of advancement in the methodological adaptations for both approaches highlights the possibility of overcoming pitfalls such as the sample size or the contamination levels. Three articles were published (Durier et al. 2019, 2021, 2022). This research opens up new perspectives to address the issues of counterfeiting, retouching, maintenance or restoration on musical instruments, and on the provenance study of organic binders for vanishes used in violin making, extended to other heritage objects.
... paper, manufactured from the fibers of Broussonetia kazinoki with no sizing or conservation treatment, the presence of sizing agents based on proteinaceous materials as a ground layer applied to the paper before painting and inking cannot be ruled out in ancient East Asian papers [38][39][40][41]. Furthermore, compounds related to substances of different natures used in the preservation and protection of ancient manuscripts in collection, if present, are likely to be detected during our analysis of the ancient manuscripts from the Pelliot Collection. ...
Despite the historical richness of ancient records mentioning or describing Chinese ink since the first centuries CE, the manufacture and use of Chinese inks in early history remain obscure. The characterization of inks from dated ancient manuscripts could not only give us new information on the evolution of the manufacture of inks but also on their use. The purpose of this work was to optimize and then validate a Pyrolysis-comprehensive two-dimensional gas chromatography/mass spectrometry (Py-GCxGC/MS) approach to identify compounds characteristic of ingredients used in the manufacture of traditional Chinese inksticks. Performance assessment was accomplished by studying the distribution of compounds characteristic of soot and proteinaceous binder and by performing repeatability tests remaining constant from one analysis to the other. A strategic sampling protocol and the optimized Py-GCxGC/MS method was then applied for the first time to characterize ancient inks manufactures from tiny fragments of Tocharian manuscripts (6th -8th centuries CE) of the Pelliot Collection stored in the Bibliothèque nationale de France as part of the “History of the Tocharian Texts of the Pelliot Collection” (HisTochText) ERC project. The approach allowed us to distinguish compounds attributable to the ink present from those imputable to the paper support and its preparation. Compounds characteristic of soot as well as of proteinaceous material were identified. Moreover, our data provides conclusive evidence for the use of a wide range of additives in the manufacture of inks. Based on the analytical data, this work demonstrates the potential of a Py-GCxGC/MS approach in the characterization of inks from tiny fragments of ancient manuscripts of historical value, making this approach highly relevant to the study of inks manufacture.
The use of engineered wood products (EWPs) is rapidly increasing in the building industry worldwide due to their reliable structural performance, lower weight, renewability, carbon storage, quick on-site installation, and lower levels of construction waste. The adhesives used for manufacturing the EWPs are mainly synthesised from petroleum and natural gas derived chemicals. However, increasing concerns regarding formaldehyde emissions, environmental sustainability and long term security of supply of petrochemicals are the biggest motivation for researchers to develop bio-based adhesives. The main objective of this article is to review recent advances in bio-based adhesives and their application in wood based composite products. Bio-adhesives derived from lignin, protein, tannin and starch and their reported performance are discussed. Bio-based adhesives at various stages of commercial development as reported by the industry and/or in patents are also discussed. Although bio-based adhesives provide a sustainable solution and significantly reduce formaldehyde and volatile emissions, they still pose several different limitations that hinder their industrial and commercial use. The major limitations include: 1) the availability of tannins 2) lack of adhesion for starches 3) poor water resistance for lignin and protein and 4) low strength properties mainly limiting their use to non-structural applications. However, the literature review demonstrates that various modifications, additives and cross-linkers can significantly improve various properties of bio adhesives. The paper also presents a brief summary of the advantages and disadvantages of synthetic adhesives.
For the preservation of painted cultural heritage on wooden substrates, it is important to understand the fracture mechanisms in the multilayer system of which they are constructed and how the environment plays a role in the composites’ physical properties. Past research has investigated the material response of each constituent layer but much more needs to be done to represent the heterogeneous composite structure of easel paintings. In recent years fracture mechanics concepts have been applied to glue and glue/chalk multilayers. However, few experiments have been conducted on multilayers that include oil paint, due to its very long, and impractical drying time, which can be a few years up to decades depending on the type of study. The paper presents a methodology for the use of thermally aged alkyd paint in easel painting reconstructions for mechanical testing, specifically as a substitute for naturally aged traditional linseed oil paint. Elastic and failure properties of the paint have been obtained from environmentally-controlled tensile tests on thin free-film samples. To obtain the characteristic properties of increased elastic modulus and reduced ductility, a thermal ageing protocol has been experimentally developed. The results are compared with data from the published literature, theoretical models and with 30-year-old samples of cold-pressed linseed oil lead white paint tested within this research work. The final methodology provides the research community with a viable way to produce samples that can be used to understand the behaviour of a (simplified) but complete multilayer system.
Full-text available
Η παρούσα εργασία αποτελεί μια βιβλιογραφική επισκόπηση και μελέτη των υποστρωμάτων, των μέσων σχεδίασης και των ιστορικών μεθόδων φωτοαντιγραφής των τεχνικών σχεδίων με έμφαση στον τρόπο παρασκευής, την αναγνώριση και τη διατήρησή τους. Η εργασία συζητά επίσης τις βασικές αρχές και την ισορροπία που θα πρέπει να επιτευχθεί ανάμεσα στη συντήρηση και την πρόσβαση στις συλλογές. Υπό το πρίσμα αυτό παρουσιάζονται ακροθιγώς τα θέματα της αξιολόγησης, της ταξινόμησης, της περιγραφής καθώς και των ερευνών για την εκτίμηση της κατάστασης των συλλογών τεχνικών σχεδίων. Συζητούνται τα θέματα του κλιματικού ελέγχου και των συνθηκών φύλαξης, της αποθήκευσης, του χειρισμού τους και των εκθέσεων τεχνικών σχεδίων και των μέτρα προστασίας από φωτιά, πλημμύρα, κλοπή και βανδαλισμό. Εξετάζονται τα στάδια σύνταξης ενός σχεδίου έκτακτης ανάγκης και παρουσιάζονται οι απαιτούμενες ενέργειες και μέθοδοι αποκατάστασης σε περίπτωση καταστροφής από νερό εστιάζοντας στις ιδιαιτερότητες της τεχνολογίας κατασκευής των τεχνικών σχεδίων. Τέλος, γίνεται παρουσίαση των μεθόδων αντιγραφής και μεταφοράς της πληροφορίας σε άλλο υπόστρωμα καθώς και των μεθόδων μέθοδοι συντήρησης των τεχνικών σχεδίων. The current thesis presents a bibliographical review and study of the materials, the drawing media and the historical photoreproductive methods of technical drawings with emphasis on their manufacturing process, optical identification and degradation causes. The thesis also discusses the basic principles and the appropriate balance between preservation and access to the collections. In this view, the issues of appraisal, arrangement, description and collection conservation surveys are briefly covered. Furthermore, it discusses topics of environmental conditions control, storage and handling, technical drawings exhibitions and security measures against fire, flood, theft and vandalism. It also examines the editing stages of a disaster management plan and lists the necessary restoration procedures and recovery options after water damage focusing on the specifics of technical drawings copying technologies. Finally, after discussing copying and reformatting methods, conservation treatment processes for technical plans are also presented.
Adhesives are effective bridging substances with huge application and market demand. Their application areas are gradually expanding from the traditional wood and construction industries to light chemical engineering as well as medicine and textile industries. This review discusses the composition, bonding mechanisms, synthesis methods and application characteristics of textile-related adhesives. Based on their origin, adhesives are divided into three categories: natural, synthetic and natural/synthetic composite. In a comprehensive study, we further explore the role of adhesives in the development of textiles. This review will facilitate the development of future textiles with functionality, intelligence and diversified applicability.
Canvas paintings are prone to environmental ageing. Hence, the structural conservation of canvas paintings may require lining, a process in which a secondary canvas is adhered to the reverse of the damaged original canvas to provide additional support. Choosing the optimum adhesive in combination with a lining method is crucial and yet challenging, as they should preferably be mechanically and chemically stable and reversible for at least 100 years. Comprehensive data on thermal and long term mechanical behaviour of prevalently used adhesives and their bonded assemblies to canvas is scarce and yet necessary to enable conservators for a proper choice of the materials in terms of durability. In this study, four prevalently used adhesives in the conservation of canvas paintings are investigated and their creep performance is evaluated and benchmarked at three different temperatures and environmental relative humidities (RHs). These adhesives are either bio-based (animal glue-starch paste and beeswax-dammar resin mixtures), or synthetic (BEVA® 371 and an aqueous PlextolTM D540/K360 dispersion mixture). Differential Scanning Calorimetry (DSC) technique is used to study the thermal transitions at different RHs. T-peel and lap shear tests are performed to determine the fracture behaviour and shear strength respectively. An in-house built creep set-up equipped with environmental control is developed which allows investigation of the mechanical creep for different canvas bonded assemblies. The results demonstrate the effects of temperature and relative humidity on the creep behaviour of lined canvases, which are related to their physical response. Moreover, the animal glue-starch paste shows the best creep mechanical performance for this application, while the PlextolTM acrylic dispersion mixture in combination with Mist-Lining is a better alternative when both environment and reversibility are considered.
Full-text available
This book is a printed edition of the Special Issue Historical Wood: Structure, Properties and Conservation that was published in Forests. This is a selection of manuscripts devoted to the conservation and preservation of wooden cultural heritage. The articles present the new methods for conservation of various historical wooden artefacts, reliable modern techniques for characterisation of the wood structure, properties and degree of degradation, and discusses problems and doubts related to all aspects of conservation and re-conservation of wooden cultural heritage. It contains both review and research papers to give the readers a broader picture of the problems and issues related to the conservation of wooden historical objects and structures. We need to remember that wooden cultural heritage is an integral part of our culture and history that define our humanity. We are obliged to protect it, save from oblivion and preserve it for future generations.
A comprehensive guide to the technology and conservation of both Asian and European lacquer.
The comparison of physical and mechanical properties of socks knitted from cotton, linen and bamboo yarn has been carried out, it has shown that the properties of socks from bamboo yarn are close to those of the socks of cotton yarn and surpass them by hygroscopicity in two times.