Studies in Conservation

ISSN: 0039-3630 (Print) 2047-0584 (Online) Journal homepage: http://www.tandfonline.com/loi/ysic20

The verisimilitude of verdigris: a review of the

copper carboxylates

David A. Scott, Yoko Taniguchi & Emi Koseto

To cite this article: David A. Scott, Yoko Taniguchi & Emi Koseto (2001) The verisimilitude of
verdigris: a review of the copper carboxylates, Studies in Conservation, 46:2, 73-91

To link to this article: http://dx.doi.org/10.1179/sic.2001.46.2.73

Published online: 19 Jul 2013.

Submit your article to this journal

Article views: 38

View related articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ysic20

Download by: [University Library Utrecht] Date: 17 March 2016, At: 05:00
 The verisimilitude of verdigris: a review of the copper
carboxylates
David A. Scott, Yoko Taniguchi and Emi Koseto

forms the basis of the important organometallic Renaissance

Abstract
pigment and glaze, known as copper resinate, and can be
Copper formates and copper acetates occur both as dissolved in size or gelatin to make a copper proteinate. This
corrosion products on works of art and as pigments. versatility is what makes verdigris such a useful pigment, since,
Some of the characteristics of these compounds are by itself, its use in oil paintings is limited and was quite rightly
reviewed, particularly in relation to their occurrence viewed with a certain amount of suspicion regarding its
in works of art, the deterioration of associated compatability with other pigments and its longevity.
materials, and historical pigment recipes for the
preparation of verdigris [copper (II) hydroxyacetate When bronze objects corrode due to inadvertent exposure to
acetic acid (ethanoic acid), as a result of unsuitable materials
hydrates] and related compounds. X-ray diffraction
being used for packing, storage or display, or when copper
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

analysis is discussed at length and new sets of powder

strips are deliberately exposed to acetic acid vapours from stale
diffraction measurements are provided, for both the
wine or old vinegar to make a pigment, many different forms
basic copper formates and the basic copper acetates,
of verdigris may result, for there is not one copper acetate, but
which represent reference data for the verdigris salts many basic salts whose chemistry is confused and confusing.
[copper (II) hydroxyacetate hydrates], not previously The aim of this review is to elucidate what is known about
available as a comprehensive set of data, as no entries these verdigris compounds and to discuss some of the results
for them at present exist in the International Centre of recent syntheses of them, and of replication experiments of
for Diffraction (Joint Committee on Powder some ancient recipes for the production of verdigris, to ascertain
Diffraction Standards) (ICDD (JCPDS)) files. The what types of compounds actually could be produced.
utility of this new data is reviewed and the optical
characteristics of the different verdigris preparations The dubious nature of the crude verdigris initially made was
recognized historically as being an unrefined substance, since
examined. Many of the historical recipes were
it was common practice to produce a ‘distilled verdigris’ by
replicated in the laboratory, providing further insight
dissolving the crude product in clear vinegar and collecting the
into the complexity of these verdigris salts and their
botryoidal blue-green crystals of the recrystallized salt on
chemical constitution, some of which include the basic
wooden sticks by slow evaporation of the solution. This distilled
copper chlorides, citrates, carbonates and other verdigris is the neutral copper (II) acetate monohydrate, the
copper compounds which are often difficult to best known of the copper (II) acetates; the others being an
identify. The chemical literature relating to the copper assortment of various basic and hydrated salts.
acetates, and some of the classical literature describing
pigment preparations based on verdigris, is reviewed. The copper (I) and copper (II) formates
Introduction Formic acid (HCOOH; methanoic acid) is a common pollutant
in the museum environment due to the use of laminates, glues
Verdigris, loosely defined, is used in common parlance to mean and wood products, which may emanate organic acid vapours,
any green or blue corrosion resulting from the alteration of particularly formic and acetic acids. Problems in the
copper on exposure to the atmosphere, or more particularly, identification of these copper formate salts are principally due
to acetic acid and sometimes formic acid. Verdigris can be a to the fact that no basic copper (II) formates are listed in the
disfiguring form of corrosion of copper alloy objects or works ICDD files. The known compounds are shown in Table 1, which
of art or it may be a pigment, deliberately made by the corrosion follows Gmelin [1-4] and the ICDD files [5].
of copper, or by the conversion of another copper mineral into
verdigris. Since verdigris is soluble in many oils and resins, it The neutral copper (II) formate is freely soluble in water and
is the usual product from direct reaction between copper and

Table 1 The copper formates and basic copper (II) formates.

Compound Formula Crystal System

copper (I) formate Cu(HCOO) not stated in file

copper (II) formate Cu(HCOO)2 not stated in file
copper (II) formate Cu(HCOO)2.H2O monoclinic
basic copper (II) formate 2Cu(HCOO)2.Cu(OH)2.2H2O triclinic
basic copper (II) formate Cu(HCOO)2.Cu(OH)2 not known
basic copper (II) formate Cu(HCOO)2.2Cu(OH)2 not known

73
 R EVIEWS IN CONSERVATION N UMBER 2 2001

formic acid. Apart from a few early citations in the chemical supposed compounds could be confirmed by an attempt at a
literature, there is a paucity of information that has been recent synthesis, which is revealing of the subtleties of the
published concerning the basic copper formates, which are more system, since Rahn-Koltermann et al. could not produce
likely to be found in corrosion contexts, since these are only compound B, Cu(CH3COO)2.Cu(OH)2.5H2O. They discerned
sparingly soluble, so some details concerning them are given distinct data only for compounds A, C and D, but a new,
here. Fowles found that when solutions of the neutral copper additional phase was determined, which has been assigned a
(II) formate were boiled, an insoluble basic salt, formula analogous to compound D, namely,
2Cu(HCOO)2.Cu(OH)2.2H2O, was formed, and that copper Cu(CH 3 COO) 2 .[Cu(OH) 2 ] 4 .3H 2 O, which we shall call
(II) hydroxide readily combined when shaken with a cold compound H. The existence of a new basic salt, not previously
solution of copper formate to form the copper (II) reported, complicates the discussion of these compounds. The
trihydroxyformate salt Cu(HCOO)2.3Cu(OH)2, which tended difficulty with these basic acetates is that there may be a
to revert to Cu(HCOO)2.2Cu(OH)2 [6]. If the freshly-prepared continuum based on the theme of:
trihydroxyformate is added to a dilute solution of the neutral
[Cu(CH3COO)2]x[Cu(OH)2]y.zH2O)
formate, after ten weeks it will change into a green crystalline
powder. If the addition of an excess of a saturated solution of where typical values are x= 1, y=3 and z=2, which equate to
the neutral salt were used, complete conversion required 17 compound D as found by Gauthier, although compounds A, B
days. The compound obtained, Cu(HCOO)2.Cu(OH)2, is a fine and C can all be considered as possible variants on this general
emerald-green crystalline deposit. formula. According to Yamanaka [11], the basic copper acetate,
compound D, has a layered structure of the botallackite type.
For the review of this series of salts, the authors synthesized
The acetate ions are located between the positively charged
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

the basic formates using the same techniques employed by

copper hydroxide layers and can be exchanged with various
Fowles and obtained the X-ray diffraction data given in Table
other anions. This may partially explain why they are difficult
2, which should be of use in further studies of basic formate
to characterize fully, since the presence of other anions, such
salts as corrosion products of metallic works of art, or as
as chloride, sulphate or nitrate, may produce slight variants of
pigment preparations. For example, if copper strips are kept
the parent compound, because of substitution of some of the
in an atmosphere of 80% RH with 200ppm of formic acid,
copper hydroxide layers. This substitution may also help to
dark green corrosion spots form on the copper surface; these
explain the considerable variation in the analytical data for
were analysed by powder x-ray diffraction and shown to
these verdigris compounds, particularly the x-ray powder
provide data very similar to that for our laboratory synthesis
diffraction data.
of 2Cu(HCOO)2.Cu(OH)2.2H2O. A hydrated basic salt is
therefore formed, which illustrates the potential importance The ICDD [12] files recognize only three copper acetates,
of the new data, since otherwise the corrosion formed on the shown in Table 4, of which copper (II) acetate monohydrate,
copper strips would remain unidentified if only the ICDD compound F, Cu(CH3COO)2.H2O, is by far the most common.
(JCPDS) files were relied upon for reference information. Two Compound F is the product that is most easily made with
refractive indices could be determined from the mounted modern, chemically pure, ingredients.
crystals in Cargille liquids, one at µ = 1.620 and the other at
about 1.652. Synthesis and optical properties
Evidence for the existence of copper formates as a patina
In order to research the optical characteristics and x-ray
component, in small concentrations, comes from the work of
diffraction data for the copper (II) acetates, a number of new
Graedel et al. [7], from studies dealing with the corrosion of
syntheses were carried out at the Museum Research Laboratory
the Statue of Liberty, New York. An unusual mention of the
of the Getty Conservation Institute (GCI).
formate anion is by Pey [8], in characterization of some green
pigments in the Hafkenscheid collection, which dates from the Compounds A, [Cu(CH3COO)2]2.Cu(OH)2.5H2O, and B,
early decades of the nineteenth century. One pigment, Parrot Cu(CH3COO)2.Cu(OH)2.5H2O, were synthesized using recipes
green [Papegaaigroen], is described as ‘possibly a formo- from Gauthier, also used by Schweizer and Muhlethaler [13].
arsenite’, which would make Parrot green an analogue of We have already noted however, that Rahn-Koltermann et al.
Schweinfurt green, but no further analytical data was provided could not produce compound B, but in our own laboratory
in the article by Pey. Further research will undoubtedly result studies, we found that several different solid products, five in
in descriptions of the copper (II) formates being found, all, could be removed at different stages of the reaction that
especially as corrosion products. attempted to produce Cu(CH3COO)2.Cu(OH)2.5H2O. The
synthesis of compounds C and D, Cu(CH3COO)2.[Cu(OH2)]2,
The copper (II) acetates and Cu(CH3COO)2.[Cu(OH)2 ]3.2H2O, also followed Gauthier.
Basic verdigris, compound A, [Cu(CH3COO)2]2.Cu(OH)2.5H2O,
The copper acetates (copper ethanoates) are far more common
is a pale blue or blue-green colour and under the microscope
than the somewhat esoteric formates and there is a complex
appears as turquoise-coloured fibrous crystals. The brush-like
family of them, first studied in an often-quoted but virtually
crystal aggregates show intense green, yellow and blue
unobtainable Ph.D. thesis by Gauthier in 1958 [9]. Some details
birefringence under crossed polars with some crystals appearing
of these compounds are given in Table 3: the results in this
almost white. The size of the individual fibrous crystals is very
paper are principally from the recipes given by Gauthier.
small.
Manufactured verdigris could be any one of these compounds,
or mixtures of compounds A and B, or compounds B and D. Basic verdigris, compound B, Cu(CH3COO)2.Cu(OH)2.5H2O,
Acetate C can only occur as a single species on the basis of the is a deep blue-green colour, which under the microscope appears
equilibrium diagram, according to Gauthier. However, dark green to deep blue-green, and may be composed of at
according to Rahn-Koltermann et al. [10], only three of these least two different crystal forms. Most of the particles appear

74
 Downloaded by [University Library Utrecht] at 05:00 17 March 2016

Table 2. X-ray powder diffraction data for nine different copper formate minerals, some representing the basic copper (II) formates, derived from this study

intense blue pale blue pale blue pale blue pale blue light green green green
16-954 14-804 14-708 32-331 32-332 46-867 XRD770 XRD772 XRD778
Cu(HCOO)2.2H2O Cu(HCOO) Cu(HCOO)2 Cu(HCOO)2 Cu(HCOO)2 Continued Cu(HCOO)2 Cu(HCOO)2.Cu(OH)2 basic salt Cu(HCOO)2.2H2O

d l(%) d l(%) d l(%) d l(%) d l(%) d l(%) d l(%) d l(%) d l(%) d l(%)
4.88 100 7.08 100 4.9 100 5.24 100 7.09 50 1.6743 2 5.21 100 7.35 3 6.19 10 8.54 100
4.68 65 5.23 10 4.6 70 3.969 2 5.56 70 1.6493 10 4.8 88 6.64 100 5.90 10 6.28 30
4.56 85 4.54 30 4.29 45 3.444 3 4.81 100 1.6247 2 3.468 37 5.39 5 5.31 80 5.54 10
4.31 10 3.01 10 3.6 12 3.348 4 4.47 40 1.592 6 3.229 36 4.52 5 4.50 30 5.08 5
T HE

4.22 4 2.86 30 3.39 12 3.282 10 4.15 2 1.5431 1 3.065 43 4.08 10 3.97 50 4.70 10
3.63 10 2.68 40 2.84 18 3.184 2 3.79 5 1.5152 2 2.877 28 3.82 5 3.61 10 4.40 5
3.56 10 2.59 10 2.61 70 3.063 25 3.64 2 1.5074 10 2.482 29 3.71 3 3.29 20 3.82 5
3.51 6 2.39 15 2.44 12 2.8553 6 3.546 50 1.4895 1 2.34 34 3.37 60 3.08 100 3.27 5
3.41 20 2.36 10 2.2 12 2.6688 2 3.519 7 1.4598 7 2.075 32 3.31 2 2.87 20 2.98 5
3.33 18 2.27 10 2.12 6 2.5579 20 3.474 50 1.4421 2 2.06 31 3.19 5 2.62 20
3.28 4 2.14 10 2.02 6 2.5061 6 3.299 60 1.4265 2 3.05 2 2.48 10
3.14 6 2.1 25 1.95 12 2.4894 10 3.235 30 1.4192 10 2.94 5 2.28 20
3.05 8 1.81 10 1.81 12 2.4212 8 2.884 20 1.3859 4 2.79 40 2.22 10
2.836 20 1.59 10 1.75 12 2.275 2 2.781 8 1.3737 1 2.70 3 2.13 5
2.622 40 1.44 10 1.64 6 2.2254 7 2.748 10 1.3693 1 2.65 50 2.05 10
2.6 20 1.27 10 1.53 6 2.1994 2 2.718 3 1.3608 3 2.59 5 1.99 20
2.559 6 2.1324 7 2.643 6 1.2969 2 2.46 40 1.93 5

75
2.439 10 2.1002 3 2.555 4 1.2913 1 2.36 80 1.87 5
2.341 6 2.0741 5 2.538 6 1.2815 2 2.30 10 1.82 10
2.303 6 2.046 1 2.517 4 1.2593 2 2.23 15 1.76 5
2.281 6 1.996 7 2.484 25 1.2221 1 2.14 3 1.68 10
2.239 6 1.9815 <1 2.366 4 1.2164 3 2.02 15 1.61 10
2.208 18 1.905 7 2.339 40 1.1827 1 1.95 30 1.55 5
2.128 6 1.8681 2 2.288 5 1.1658 1 1.87 10 1.48 3
2.081 6 1.861 8 2.249 2 1.1063 1 1.83 15
2.02 6 1.8175 <1 2.144 1 1.0999 4 1.68 5
1.955 10 1.799 12 2.081 20 1.64 5
1.905 6 1.765 <1 2.073 8 1.61 5
1.874 4 1.7406 5 2.06 10 1.59 15
1.848 6 1.7208 5 1.983 2 1.55 15
VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

1.819 10 1.7039 3 1.922 3 1.52 10

1.757 8 1.6822 5 1.886 6 1.51 10
1.73 6 1.6695 2 1.867 1 1.48 15
1.711 4 1.6626 4 1.853 5 1.44 10
1.665 4 1.6407 <1 1.774 30 1.40 5
1.652 6 1.5989 4 1.749 1 1.38 5
1.643 4 1.5926 1 1.741 2 1.33 10
1.626 6 1.5452 3 1.734 20 1.29 10
1.59 4 1.5405 <1 1.69 1 1.26 5
1.529 2 (Cont. next col.) 1.25 5
 R EVIEWS IN CONSERVATION N UMBER 2 2001

Table 3. The nominal compositions of some basic copper formates and basic copper (II) acetates

Compound Formula Crystal System Colour

basic copper (II) acetate (A) [Cu(CH3COO)2]2.Cu(OH)2.5H2O not stated blue

basic copper (II) acetate (B) Cu(CH3COO)2.Cu(OH)2.5H2O not stated pale blue
basic copper (II) acetate (C) Cu(CH3COO)2.[Cu(OH2)]2 not stated blue
basic copper (II) acetate (D) Cu(CH3COO)2.[Cu(OH)2]3.2H2O monoclinic green
Basic copper (II) acetate (H) Cu(CH3COO)2.[Cu(OH)2]4.3H2O not stated blue-green

Table 4. The copper (II) acetates currently recognised in the ICDD files
copper (II) acetate (E) Cu(CH3COO)2 tetragonal pale blue ICDD 27-1126
copper (II) acetate hydrate (F) Cu(CH3COO)2.H2O monoclinic intense ICDD 27-0145
blue-green
copper (I) acetate (G) Cu(CH3COO) not stated not stated ICDD 28-0392

to be composed of bundles of fine fibres of varying orientation, index well below 1.662. Under crossed polars, the neutral salt
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

some curved. The relief varies on rotation, showing differences reveals muted brown, yellow and dark blue colours, none of
in refractive index (RI), the principal RI determined was µ = them very intense, and the particles often show clear extinction.
1.548, the other closer to 1.53. The particles tend to have Kuhn [14] quotes the RI of basic verdigris as 1.56 and 1.53.
undulose or no clear extinction under crossed polars. They These are similar values to those found here for verdigris B,
show an intense birefringence: the greener particles retaining a but some of the other basic salts have refractive indexes higher
finely variegated light green colour under crossed polars, whilst that these two values.
the lighter blue-green fibrous particles appear strongly coloured
Re-examination of the melt-mount preparations after two
with intense blue and light yellow colour. Individual fibres that
years in storage showed that some of them had already
can be seen show a positive sign of elongation. In our
undergone a reaction with the mounting medium; the resin
preparations, some of these B group compounds also occurred
was slightly discoloured and the particles partially dissolved
as irregular crystalline fragments of pale blue-green colour,
and no longer clearly visible. Preparations in melt-mount of
rather than as fibrous particles.
all verdigris pigments should therefore be examined and
Basic verdigris, compound C, Cu(CH3COO)2.[Cu(OH2)]2, photographed freshly made; an unusual caveat to make with
visually a pale turquoise colour, is quite distinctive, since it pigment preparations of this kind.
consists of a mass of well-formed, acicular crystals, some
Some environmental scanning electron microscope (ESEM)
perfectly rectangular and prismatic in shape. Under plane
images for selected verdigris compounds are shown in Figure
polarized light the acicular clumps look light green in colour,
1. Verdigris salts are often present as curved, fibrous crystals,
some of the crystals curved and fibrous, but the majority appear
as the ESEM for compound A in Figure 1a reveals. One
as long prismatic needles which show parallel extinction under
preparation for compound B, gave the microcrystalline
crossed polars. In plane polarized light, in Cargille liquid of
fragments shown in Figure 1b. Compound C was prepared as
µ = 1.580, the needles almost disappear on rotation showing
tiny florets, shown in Figure 1c, and compound D as impressive
that one RI is very close to 1.580 parallel to the long axis of
sheaths of hexagonal plates, as in Figure 1d. A variety of
the crystal. The RI is lowest at 45º to the long axis and about
different crystal morphologies are possible; for further examples
1.560. The needles appear blue, parallel to the wave plate,
see Kuhn [14].
showing that they have a positive sign of elongation.
Basic verdigris compound D, Cu(CH3COO)2.[Cu(OH)2]3.2H2O, Deterioration of associated materials
macroscopically appears a light green-blue colour and
microscopically is composed of pale light green to transparent Kuhn [14] showed that both basic and neutral verdigris samples
crystalline fragments often tinged with light blue-green. These retained good lightfastness, whether painted in linseed oil, egg
particles show only low to moderate relief in melt-mount of RI tempera or gum arabic, and gave colour reflectance curves and
1.662 without any noticeable change in relief on rotation under reflectance maxima for the painted samples. Neutral verdigris
plane polarized light. Under crossed polars, the particles appear shows a reflectance maximum at about 500nm, and the basic
intensely coloured with light straw and strong light blue colours. salts in the range from 480nm-490nm. However, basic and
Some particles appear veined with microcracks or fissures and neutral verdigris manifest a colour change from blue-green to
twinkle on rotation under crossed polars with no clear green, which is apparent after 28 days. Thereafter, the colour
extinction. change is insignificant, following a period of several months’
light exposure. This change is less pronounced with neutral
Visually, neutral verdigris, compound F, is a deep blue-green verdigris than with the basic salts. These colour changes are
colour, and under the microscope is characterized by crystalline probably due to the possible reactions between verdigris and
fragments which often show conchoidal fracture and clear relief resins, oils or proteins.
when mounted in melt-mount of RI 1.662. The particles tend
to show a uniform light-blue colour without visible internal The deterioration of verdigris pigments in association with
defects within the crystalline fragments, which have a refractive painting media is complex: Rasti and Scott [15] found that

76
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

Fig. 1A ESEM Photomicrograph for basic verdigris, Fig. 1B ESEM Photomicrograph for basic verdigris,
compound A, [Cu(CH3COO)2]2.Cu(OH)2.5H2O, compound B, Cu(CH3COO)2.Cu(OH)2.5H2O, which
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

showing bundles of curved, fibrous crystals. This may also be fibrous, but here appears as irregular
fibrous form appears to be the most commonly plate-like crystals, magnification x1000.
encountered morphology in historical basic verdigris
pigments examined by the authors. Some other
verdigris salts may also crystallize as fibrous particles
too. Magnification about x2000.

Fig. 1C ESEM Photomicrograph for basic verdigris, Fig. 1D ESEM Photomicrograph for basic verdigris,
compound C, Cu(CH3COO)2.[Cu(OH2)]2, often compound D, Cu(CH3COO)2.[Cu(OH)2]3.2H2O,
occurring as acicular crystals, but here as clumps of here showing a well-developed floret of hexagonal
hexagonal plates. Magnification x 300. plates. Magnification x3000.

verdigris acted as an effective photostabilizer during the early significantly higher degrees of fragmentation; however, there
stages of diffuse daylight and ultraviolet (UV) exposure of the are problems in correlating degradation due to thermal ageing
painted films. However, Ioakimoglou et al. [16] found evidence between 60–100ºC, with natural ageing processes at ambient
of degradation of a linseed oil-verdigris mixture with oxidative temperature.
conditions involving simultaneous polymerization and chain
Yellowing in the heat-aged films was attributed to the
scission reactions. Transformation of cis- to trans-bonding in
formation of conjugated C=C bonding in oil whose existence
linseed oil is followed by the formation of peroxides whose
may be exacerbated by the carbonyl group in the verdigris.
decomposition is accelerated by verdigris.
This would explain why compounds such as copper acetate
Copper (II) salts can also participate in catalytic cycles through and copper abietate create greater chemical change than
the less stable Cu(I) state, occasionally involving complexation compounds such as malachite, since the two former compounds
with oxygen and unsaturated substrates. Gas chromatography- both have carboxylate groups present. The complexity of the
mass spectrometry (GC-MS) analysis showed that after 64 days possible reactions involved requires further research to define
of thermal ageing, linseed oil and verdigris mixtures showed precisely the reaction between verdigris and linseed oil media.

77
 R EVIEWS IN CONSERVATION N UMBER 2 2001

The degradation of wood and cellulosic material by verdigris incompletely reported, or suffer from other deficiencies. It is
and other copper pigments has been studied by Banik [17], important to rectify this situation, since it is difficult to obtain
who found that damaged papers contained copper in both the sufficient information from Fourier transform infra-red (FTIR)
univalent and divalent oxidation states, and noted that the spectrometry, for example, to differentiate between the basic
presence of univalent copper could only be explained by a salts. Low infra-red studies, between 50-600cm-1 can provide
reduction of the copper (II) ion present in the original pigment. interesting information concerning the environment of the
The cellulose in paper samples from damaged areas is copper ions in different verdigris salts, which could be very
depolymerized, with a level of polymerization from 60-80, useful for characterization, but this low IR spectroscopy range
compared with 200-500 in undamaged areas. The cellulose is rarely available and 650cm-1 is a common limitation for most
chain is split at the gluocidic linkage and under acidic conditions museum-based FTIR instruments, consequently a complete IR
the new end groups are aldehydes, whilst under alkaline picture cannot be determined, and X-ray diffraction (XRD) is
conditions the new end groups are carboxyl groups with the preferable. Even here, care must be taken with the
amorphous sections of the cellulose fibres attacked first. instrumentation, since some of the compounds have d-spacings
over 17, which is an extremely high value and some x-ray
The copper content is concentrated in the lumen of the fibre
diffraction methodologies cannot cope with the very small
where concentrations may reach 8-10% by weight. Banik found
angles needed to carry out this measurement. Old-fashioned
that the stability of verdigris-impregnated cellulose could be
Debye-Scherrer cameras can, but not everyone now relies on
enhanced by immersion in magnesium bicarbonate solutions.
this increasingly obsolete instrumentation, and the concomitant
Masri and Friedman [18] found that wool, which contains
investment in time that is involved in developing, processing,
about 3.5% of sulfur, readily adsorbs copper ions by reaction
and reading X-ray diffraction films: instead the new technology
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

with the -S-H groups to form copper mercaptides.

of the General area X-ray diffraction System (GADDS) should
Williams [19] has carried out extensive research into the simplify this work in future, since many diffractometers cannot
compounds that may form between copper and other metallic cope with samples less than 1mg, and samples from most
ions and proteinaceous materials, which should be consulted paintings are much smaller than that.
for further background information. The decarboxylation of
A short compilation of data for compounds A, B, C, D, F
RCOO- groups in proteins is facilitated by inorganic ions such
and H is given in Table 5: van’t Hul-Ehrnreich and Hallebeek
as copper, inducing deterioration. Daniels et al. [20] studied
[24] determined XRD patterns for two versions of basic
the effects of verdigris on silk and deduced an activation energy
verdigris, one of which is given here, for compound A, together
for the degradation. For the deterioration of silk to be
with data obtained by syntheses in the GCI. laboratory and
accelerated, the verdigris must come into direct contact with
data for compounds B, C and D obtained from the collections
the silk. Verdigris may be suspended in a gum binder or, in
of The National Gallery, London, from the synthesis by
Japanese paintings on silk, the painting may be sized with a
Schweizer and Muhlethaler [13]. Orna [25], published some
alum-deerskin glue mixture, which may assist in retarding the
data for X-ray diffraction work on compounds A, B, C and D,
direct action of the copper ions on the silk substrate. Gelatin
but this information was reported without details of the relative
in the glue may act as a scavenger for copper ions since the
intensities of most of the peaks, which limits the utility of the
–S–H groups form a strong linkage, which may help to prevent
publication.
chemical interaction of silk and copper ions. Daniels et al. [20]
found that both benzotriazole and aqueous solutions of Table 6 shows a set of data obtained during the laboratory
magnesium bicarbonate were effective in helping to stabilize synthesis of compound B [13], which produced a series of salts
the degradation of the damaged silk fibres, and since magnesium that could be isolated using different starting conditions. The
bicarbonate has proved effective in previous conservation utility of the new sets of data published here are shown by the
treatments, this is what Daniels recommended as the preferred fact that we were able to identify basic verdigris, B4, from a
choice for stabilization. German manuscript of the fifteenth century in the collections
of the J. Paul Getty Museum [22]. Table 7 provides the XRD
Khandekar and Phenix [21] studied molecular weight changes
data for this latter example, together with two historic verdigris
in egg tempera films to which lead white, vermilion, azurite,
samples from the Hafkenscheid collection of eighteenth-century
or verdigris pigment had been added, followed by artificial
pigments in Amsterdam [8], which look to be a complex
ageing. All samples containing azurite had cross-linked to a
mixture of salts, including probably both A and B. A recent
greater extent than found with vermilion, with molecular
synthesis of basic verdigris by Mactaggart [26] which simply
weights over 250,000. With verdigris, the effects on the egg
wrapped copper foil in vinegar-soaked towel, and data from a
tempera medium were very different: only low molecular weight
replication of a recipe in the Mappae Clavicula [27] (see below),
components, less than 14,000, were found, showing that the
recipe vi, and two further laboratory synthesis of A and B which
verdigris had initiated a breakdown of the proteins.
provided slightly different data.
Verdigris can be dissolved in resins, to produce copper resinate
As a result of this study, a number of themes emerge from a
pigments and glazes, which are of considerable interest in
comparison of the d-spacings of the various compounds. Firstly,
European oil painting, and in proteinaceous media, such as
the fact that some of the verdigris compounds have their most
parchment size, to make copper proteinates [22]; there is
intense d-spacings in the range of 17, 16 or 15. It is very unusual
insufficient space to discuss these aspects here, but there have
that compounds expected to be relatively simple, such as
been recent publications concerning this topic [23].
verdigris, should have such high d-spacings, and this suggests
that some of the crystal planes in the lattice are very tightly
X-ray powder diffraction data spaced, producing complex or unusual structures which have
been sadly neglected by structural chemists as of the date of
The review of the literature highlighted the fact that most of
writing. Secondly, many of the verdigris salts have strong d-
the previously published x-ray diffraction studies have been

78
 Downloaded by [University Library Utrecht] at 05:00 17 March 2016

Table 5. XRD data for some verdigris variants; Compound A (after Gauthier); Compound A (after Van T’Ehrenreich); Compound B (after Schweizer and Mühlethaler); Compound B (lab
synthesis), Compound C (after Schweizer and Mühlethaler); Compound D (lab synthesis); Compound F (ICDD 27-145) and Compound H (lab synthesis).

Cmp A Cmp A Cmp B Cmp B Cmp B2 Cmp B4 Cmp C Cmp D Cmp F Cmp H
after Gauthier after Van T’H after M & S lab syn. lab syn. lab syn. after M & S lab syn. neutral Salt similar to A
XRD 625 Continued XRD690 yt yt D 27-145

I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d I/Io d
100 9.3 20 1.35 40 17.3 40 8.4 60 10.66 100 11.32 80 6.47 20 15.42 3 14.95 100 6.91 100 16.31
40 6.98 4 2.222 70 12.3 5 6.7 50 9.18 95 10.02 10 6.1 100 10.52 100 9.25 35 6.17 6 11.82
T HE

20 6.27 2 2.186 80 11.9 5 5.8 100 8.45 95 8.32 2 3.8 15 6.54 3 5.12 25 5.87 8 9.42
20 5.79 1 2.149 100 9.3 6 4.8 5 7.35 10 7.18 100 3.64 10 5.39 30 4.67 16 5.75 4 7.81
20 5.44 2 2.137 90 8.5 8 4.08 2 6.5 10 6.5 2 3.54 2 4.03 5 4.07 25 5.38 11 6.18
20 5.17 2 2.035 60 8.2 30 3.88 1 6.15 10 5.84 1 3.24 5 3.59 3 3.68 2 4.28 8 5.1
90 4.72 4 2.027 40 7.83 8 3.56 10 5.17 10 5.15 20 3.05 10 3.44 3 3.42 4 4.09 2 4.32
10 4.34 1 1.976 50 4.48 80 3.45 2 4.47 5 4.84 1 2.69 5 3.12 20 3.11 2 4.05 1 3.86
30 4.11 4 1.963 70 4.11 70 3.31 1 4.05 10 4.59 80 2.63 3 2.9 5 2.78 12 3.588 0.4 3.39
10 3.75 2 1.952 20 3.8 100 3.1 10 3.72 30 3.96 95 2.48 40 2.78 10 2.67 20 3.527 2.5 3.084
10 3.55 2 1.928 40 3.42 50 2.83 2 3.62 40 3.6 5 2.39 3 2.46 5 2.58 4 3.455 1 2.863
10 3.4 2 1.891 20 3.21 10 2.73 1 3.4 5 3.38 50 2.3 10 2.41 15 2.45 4 3.434 2 2.596
80 3.16 2 1.874 50 3.06 10 2.64 2 3.2 10 3.24 20 2.16 10 2.2 10 2.29 4 3.292 4 2.227

79
10 3.04 2 1.83 50 2.4 10 2.48 1 2.98 10 3.01 2 2.03 2 1.79 5 2.17 4 3.087
30 2.83 1 1.802 10 2.32 60 2.21 10 2.77 25 2.87 10 1.93 5 1.69 2 2.01 2 3.033
70 2.67 1 1.794 30 2.16 10 2.11 5 2.61 20 2.75 10 1.81 5 1.63 10 1.89 2 2.967
60 2.5 2 1.774 50 2.03 5 2.06 2 2.44 25 2.63 2 1.74 5 1.62 5 1.58 2 2.875
50 2.35 50 1.71 10 1.86 5 2.35 25 2.47 15 1.66 3 1.58 5 1.52 2 2.773
60 2.24 30 1.52 1 2.23 25 2.32 7 1.58 3 1.42 3 1.45 2 2.723
60 2.17 5 2.15 3 1.36 2 2.688
30 2.04 10 2.02 2 2.645
5 1.99 10 1.92 2 2.609
30 1.94 15 1.81 2 2.592
30 1.89 10 1.66 1 2.576
20 1.76 5 1.56 4 2.543
VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

20 1.69 5 1.43 2 2.51

20 1.64 2 2.495
30 1.61 2 2.425
30 1.57 4 2.393
10 1.54 4 2.387
10 1.51 8 2.33
20 1.41 10 2.29
20 1.38 4 2.279
(Cont. next col.) 4 2.226
 R EVIEWS IN CONSERVATION N UMBER 2 2001

Table 6. XRD data for some verdigris variants in the laboratory synthesis of compound B, where five different salts could be
isolated. Note that verdigris B4 is a good match to the verdigris data for the fifteenth-century German manuscript green in Table 7,
and that the historic Hafkenschied pigments show similarities to compounds A and C.

Compound B1 Compound B2 Compound B3 Synthesis B4 Compound B5

XRD 689 XRD 690 XRD 691 XRD 696

l/lo d l/lo d l/lo d l/lo d l/lo d

10 11.06 100 11.32 5 9.24 80 6.47 10 11.37
100 10.04 95 10.02 100 6.88 10 6.1 5 10.11
95 8.45 95 8.32 20 6.23 2 3.8 100 9.18
15 7.12 10 7.18 20 5.83 100 3.64 10 8.31
15 6.55 10 6.5 20 5.38 2 3.54 80 6.89
10 5.77 10 5.84 5 4.22 1 3.24 40 6.18
10 5.16 10 5.15 40 3.52 20 3.05 10 5.8
5 4.87 5 4.84 10 3.06 1 2.69 15 5.52
20 4.6 10 4.59 10 2.64 80 2.63 20 4.64
35 3.93 30 3.96 10 2.31 95 2.48 10 4.31
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

15 3.54 40 3.6 5 1.97 5 2.39 10 4.06

10 3.4 5 3.38 5 1.87 50 2.3 10 3.63
5 3.35 10 3.24 20 2.16 5 3.4
10 3 10 3.01 2 2.03 10 3.3
10 2.87 25 2.87 10 1.93 5 3.04
10 2.75 20 2.75 10 1.81 5 1.92
2 2.67 25 2.63 2 1.74 5 1.82
10 2.53 25 2.47 15 1.66 5 1.71
5 2.47 25 2.32 7 1.58
5 2.39 5 2.15
5 2.33 10 2.02
2 2.19 10 1.92
2 2.1 15 1.81
2 2.03 10 1.66
5 1.86 5 1.56
5 1.43

spacings in the range of 12, 11.5, 9.3, 7, or 11, 9, 7, or various Eastlake himself [28], who remarks that ‘...the mode of ‘locking
combinations of these, which are sufficiently unique to allow up’ verdigris may exemplify the means by which all colours
identification of a whole range of substances previously in liable to be affected by damp can be rendered durable. The
doubt. Thirdly, there are less common compounds with strong colour was mixed either with a strong oleo-resinous vehicle or
d-spacings in the range of 6.5, 3.6 and 2.5, such as compound with varnish only...’ [28, p. 458].
B4 with strong lines at 2.6, 2.48 and 2.3.
While Cennini [29], mentions that both lead white and
verdigris are ingredients that assist with the drying and
Verdigris: pigment and uses consolidation of cements, as well as acting as a siccative:
The possibility of verdigris presenting problems of solubility ‘...If you wish this mordant to last [that is, to remain
have long been well-known: for example, Leonardo da Vinci, adhesive] for a week before gilding, put no verdigris; if
translated in Eastlake [28] remarks that verdigris, though you wish it to last four days, put a little verdigris; if you
ground in oil, can only last when it is varnished immediately wish the mordant to be good for a day only, put much
after it is dry, otherwise, he writes ‘...it not only fades, but may verdigris...’ [29, p. 97].
be removed by a wet sponge, especially in humid weather. This
is because of its saline nature; it becomes deliquescent in a Verdigris was also commonly used as a drier for oil paints, as
moist atmosphere...’ [28, p. 458]. well as for mordants, and is mentioned in many early sources,
such as de Mayerne [30] as an additive to dark-coloured paints
This would be particularly true with distilled verdigris; some to make then dry better.
of the basic salts, or other products masquerading as verdigris,
would not necessarily be affected by very humid weather. The The use of verdigris in paintings was common in all periods;
theme of observations on humid conditions continues with for example, in several paintings from the thirteenth to nineteenth

80
 Downloaded by [University Library Utrecht] at 05:00 17 March 2016

Table 7. XRD data for some typical basic verdigris preparations, including three historical samples, two from the Hafkenschied collection, one from the Von Ems fifteenth-century manuscript,
and four laboratory synthesis, one replicating the Mappae Clavicula recipe vi.

Basic verdigris Mappae Clavicula Synthesis Synthesis Hafkenschied Hafkenschied German MS

Mactaggart recipe vi Ayt yt b1 Collection Collection Continued Von Ems
XRD 587 XRD 606

l/lo d l/lo d l/lo d l/lo d l/lo d l/lo d l/lo d l/lo d

100 11.9 100 11.5 5 5.11 0.1 33.2 77 15.75 18 17.8 4 3.154 80 6.47
90 8.26 20 9.34 100 4.68 15 11.2 6 11.84 64 15.7 12 3.135 10 6.1
50 7.06 80 8.16 4 4.04 100 9.29 5 7.86 12 11.8 28 3.086 2 3.8
T HE

40 5.97 70 6.92 50 3.12 1 8.54 3 7.67 5 11.24 23 3.046 100 3.64

10 5.48 40 6.26 7 2.78 30 6.93 83 6.17 91 9.27 7 2.98 2 3.54
70 4.97 30 5.88 14 2.64 1 6.48 34 5.1 8 8.43 7 2.928 1 3.24
70 4.61 30 5.44 7 2.58 1 6.09 2 4.33 2 8.14 6 2.902 20 3.05
10 4.19 10 4.88 11 2.48 2 5.77 100 4.13 9 7.87 8 2.868 1 2.69
10 3.75 10 4.14 10 2.44 1 4.65 39 3.72 5 7.63 4 2.826 80 2.63
70 3.54 10 3.63 7 2.33 0.3 3.62 7 3.39 13 7.18 5 2.802 100 2.48
10 3.29 10 3.47 5 2.27 0.4 3.47 11 3.35 7 6.76 14 2.656 5 2.39
70 3.15 20 3.33 10 2.21 0.1 3.11 7 3.23 100 6.17 7 2.634 50 2.3
10 3.01 30 3.12 7 2.16 0.1 2.24 3 3.15 6 6.06 46 2.612 20 2.16
10 2.932 10 2.557 4 2.02 0.1 2.04 25 3.08 6 5.62 16 2.57 2 2.03
10 2.806 20 2.399 2 1.97 0.1 1.93 3 2.98 8 5.49 6 2.536 10 1.93

81
10 2.673 20 2.353 4 1.92 9 2.92 7 5.23 1 2.51 10 1.81
10 2.54 20 2.3 8 1.87 7 2.82 25 5.09 7 2.487 2 1.74
20 2.424 10 2.238 3 1.62 2 2.68 5 4.89 4 2.456 15 1.66
20 2.291 10 2.036 6 1.59 3 2.66 6 4.55 7 2.443 7 1.58
20 2.205 10 1.987 4 1.57 4 2.65 8 4.44 17 2.41
10 2.065 10 1.574 7 1.59 3 2.64 10 4.34 2 2.359
10 1.984 4 1.52 1 2.56 6 4.24 40 2.229
5 1.773 4 1.52 2 2.54 97 4.13 16 2.217
3 1.49 7 2.49 4 3.99 9 2.2
3 1.34 1 2.358 4 3.95 9 2.105
13 2.229 3 3.82 2 2.068
8 2.219 43 3.72 7 2.029
3 2.199 3 3.59 2 1.975
VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

4 2.068 3 3.63 5 1.783

2 1.757 10 3.478 4 1.77
2 1.757 5 3.45 4 1.756
29 1.68 5 3.419 3 1.736
14 1.671 3 3.391 2 1.729
32 3.358 37 1.68
2 3.269 17 1.671
10 3.233 2 1.642
9 3.219 2 1.585
(Cont. next col.)
 R EVIEWS IN CONSERVATION N UMBER 2 2001

century AD, ranging from a German altarpiece of the thirteenth recorded by Pliny [39] for a variety of medicinal treatments,
century to a painting by Ernst von Liphart dated 1877 [14]. for example:
Laurie [31] was not able to identify with certainty verdigris on
‘Verdigris in a crude state is used as an ingredient in
any manuscript before AD 1400, while from 1419 onwards,
plasters for wounds also. In combination with oil it is a
brilliant greens were found which revealed the crystalline
marvellous cure for ulcerations of the mouth and gums
structure of verdigris. For example, from the Burgundy Breviary
and for sore lips, and if wax is also added to the mixture
[British Library, Harley 2897], Laurie identified finely crystalline
it cleans them and makes them form a cicatrix. Verdigris
verdigris which, he says, is characteristic of fifteenth century
also eats away the callosity of fistulas and of sores
continental manuscripts; he describes these greens as partially
around the anus either applied by itself or with gum of
composed of doubly-refracting crystals, below oil of Cassia (µ =
Hammon’ Book XXXIV, Chapter XXVII [39, p. 154].
1.602) in their refractive index, but of a pale emerald green, not
matched by the blue-green usually associated with verdigris. Our In fact, the use of verdigris as a medicinal compound lasted
laboratory determinations of the refractive indexes show that until the beginning of the nineteenth century AD. Buchner [40]
all of the basic salts are indeed below 1.602. wrote in 1816 of the preparation of ‘Das Kupfersauerhonig’
[Unguentum Aegyptiacum] and in 1895, Kobert [41] reviewed
From the Coram Rege Rolls [folio 1013], dated from
the use of copper salts for medicinal purposes by Paracelsus
AD1500-1700, Laurie determined a verdigris green, but mixed
[1493-1541], and confirmed that they had a beneficial reactivity.
with particles of azurite which suggested the possibility that
the verdigris was not prepared by the corrosion of copper, but
by the dissolution of azurite in vinegar. Laurie also found blue Recipes for the preparation of verdigris
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

particles in some of the copper resinate greens from and related compounds
manuscripts, which may also support this possibility, although
in this case there are more likely to be undissolved crystals of The recipes in this section are considered roughly in historical
verdigris used in making the copper resinate green. Early recipes order, beginning with the earliest known accounts and
often call for the use of vinegar or wine, mixed with copper, proceeding to the most recent. Early recipes often record the
honey and salt and therefore basic copper chlorides as well as manufacture of verdigris from copper strips corroded with stale
copper acetates may be found in these complex verdigris wine, vinegar, grape-skins, urine, curdled milk or yoghurt,
pigments [32]; an interesting mention of them being in a depending on locally available materials.
painting by Veronese in the National Gallery, London [33].
An interesting historical account of the preparation of
As an ingredient of oil-based paints, verdigris embellished verdigris from fermented grapes and copper strips in France
the exteriors of French country houses and Dutch homes; it was given by Benhamou [42], which follows a recipe first
also gave an attractive green to chaises á porteur. Verdigris mentioned by Dioscorides [43], some two thousand years
was mixed with oil or distemper to illuminate stamps and tint earlier: “...hiding one mass or plate (of copper) or else more
paper as well as to colour woods for marquetry. In metallurgical amongst the huskes of grapes, not pressed of late, but growing
processes it was used in soldering and colouring of gold sour, invert it in like sort” Book V, Section 91 [43, p. 232].
jewellery. The tints and colours based on verdigris were usually
The French verdigris production lasted from the twelfth
green, as in the staining of shagreen [34], but it could also be
through to the nineteenth century AD and was, surprisingly, a
used to produce a black on both wood and cloth, perhaps from
monopoly of women who controlled the synthesis and trade
tenorite formation within the fibres. Montet writes: ‘I am told
of this important pigment: any man who attempted to take
that another common use of verdigris is to dye hats black; and
over the business, lacking the secret knowledge, was ridiculed,
a famous dyer of this city told me that he used only verdigris
a social technique which was apparently effective in protecting
to dye woolen cloth black’ [35, p. 213].
this occupation from masculine encroachment.
Verdigris was also used from an early period as a medicinal
Pliny [39] records several recipes, showing that there was
preparation, for example, Weser [36], draws attention to a
probably a long tradition of making the pigment by the first
middle kingdom Egyptian papyrus of about 1500 BC published
century AD, although some of the compounds are simply
by Ebers in 1875 [37] and known as the Papyrus Ebers [38].
corrosion products of copper rather than specific recipes for
Suspensions or ointments made from malachite, copper sulfides,
the preparation of a copper acetate:
verdigris or elemental copper in the form of small hammered
flakes was used for treatment of pain in the muscles, spinal ‘...it is scraped off the stone from which copper is smelted
cord or joints. For example to treat the member of a knee, this or by drilling holes in white copper and hanging it up in
Egyptian papyrus states: ‘Northern salt, gum ammoniac, ibex casks of strong vinegar which is stopped with a lid; the
grease, honey, frankincense, celery, calamine (?), onion, verdigris is of much better quality if the same process is
hammering-flakes from copper, grease of sheep, cumin, oil and performed with scales of copper. Some people put the
natron are ground and it is bandaged therewith’ [38, p. 46]. actual vessels, made of white copper, into vinegar in
earthenware jars, and nine days later scrape them.
Copper has excellent chelating ability, which enables it to
Others cover the vessels with grape-skins and scrape
interfere with the reproduction of many microorganisms, and
them after the same interval, others sprinkle copper
the protein copper-zinc superoxide dismutase has one of the
filings with vinegar and several times a day turn them
fastest reaction rates measured in biochemical systems. In
over with spattles until the copper is completely
laboratory studies Weser [36] demonstrated that copper chelates
dissolved. Others prefer to grind copper filings mixed
of some simple amino acids display enzymatic activity similar
with vinegar in copper mortars. But the quickest result
to the natural dismutase. Some of these ancient recipes for the
is obtained by adding to the vinegar shavings of coronet
treatment of skin infections and muscular pain are therefore
copper’ Book XXXIV, Chapter XXVI [39, p. 152].
shown to have a good biochemical basis, such as the recipes

82
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

Attempts to copy some of these recipes mostly resulted in the next day, very carefully grind it on a dry stone. Next,
production of neutral verdigris, or a mixture of neutral and gather some small twigs, place then in the above-
basic forms. Pliny [34] provides another recipe that is also of mentioned hollow box so that two parts of the cavity
interest: are below and a third above, coat the copper sheets on
each side with pure honey over which you sprinkle
‘There is also another kind of verdigris called from the
pounded salt, place them together over the twigs and
Greek worm-like verdigris, made by grinding up in a
carefully cover them with another piece of wood,
mortar of true cyprian copper with a pestle of the same
prepared for the purpose, so that no vapour can escape.
metal equal weights of alum and salt or soda with the
Next, have an opening bored in a corner of this piece of
very strongest white vinegar. This preparation is only
wood through which you can pour warm vinegar or hot
made on the very hottest days of the year, about the
urine until a third part of it is filled, and then stop up
rising of the Dogstar. The mixture is ground up until it
the opening. You should put this wooden container in a
becomes of a green colour and shrivels into what looks
place where you can cover it on every side with dung.
like a cluster of small worms, whence its name. To
After four weeks take off the cover and whatever you
remedy any that is blemished, the urine of a young boy
find on the copper scrape off and keep. Replace it again
to twice the quantity of vinegar that was used is added
and cover it as above...’ XXXV [45, p. 32].
to the mixture...’ Book XXXIV, Chapter XXVIII [39,
p.155]. Banik [46] carried out some research on this recipe and
concluded that the resulting products were often mixtures of
Pliny’s recipe XXVI results in the manufacture of a copper
copper salts, including basic and neutral copper acetates with
acetate of one kind or another, but it is not necessarily apparent
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

some malachite and additional phases that could not be

what range of compounds would result from recipe XXVIII.
characterized. If the salt predominates in this kind of mixture,
In order to investigate this more closely, a replication experiment
then the principal product will be atacamite, as found in our
was made employing a copper mortar and pestle made for the
laboratory work. However, the interaction of the honey with
purpose from pure copper, which is what Pliny implies by
the other ingredients in this recipe is of interest. Some nantokite,
specifying ‘cyprian copper’, the alum used was potassium
CuCl, was found in areas associated with the honey in the
aluminium sulphate, and the salt, sodium chloride. The product
authors’ laboratory work, which suggests that the reducing
formed using this mixture was found to be atacamite; if soda
sugars in the honey have had a very significant effect on the
was substituted for the sodium chloride, then the principal
corrosion of the copper in this case. How the nantokite would
product was found to be chalconatronite: the latter product is
eventually transform to other products is difficult to resolve at
slightly surprising, but all the essential ingredients for
present, but further interaction between the nantokite and
chalconatronite formation are included in the recipe, the copper
copper under humid conditions could be expected to produce
ions from the mortar and pestle, the sodium and carbonate
atacamite. This latter assumption was confirmed by our x-ray
from the soda.
diffraction studies, but other phases were present too, including
Urine is often mentioned in recipes for verdigris preparation. a dark green, which may be a copper chelate complex.
Shier et al. [44] state that the principal components of urine Maekawa and Mori [47] suggested that this dark green complex
are urea, uric acid and creatinine, along with various amounts might form in high humidity and the absence of light, which
of sodium, potassium, calcium, magnesium, chloride, the recipe calls for. The discovery that many ‘verdigris’ pigments
bicarbonate, sulphate and phosphate ions. This complex in illuminated manuscripts are actually basic chlorides that
mixture will be capable of dissolving both verdigris and could easily have been made by Theophilus’s viride salsum
corroding copper to some extent, and is much more active as technique, attests to the historical veracity of these recipes.
an ingredient that water alone, resulting in a very complex
Theophilus’s viride hispanicum is made as follows by:
mixture of organic and inorganic salts.
‘If you want to make Spanish Green, take some plates
Even in Roman times, verdigris was not cheap to produce,
of copper that have been beaten thin, carefully scrape
and the pigment was sometimes adulterated with powdered
them on each side, pour over them pure, warm vinegar,
marble, pumice, gum and even shoemakers’ black. The latter
without honey and salt, and put them in a small hollowed
contained ferrous sulphate, and Pliny [Book XXXIV, Chapter
out piece of wood in the above way. After two weeks,
XXVI] describes the use of a spot test using a papyrus fragment
inspect and scrape them and do this until you have
saturated with an extract of gallnuts for the detection of any
enough colour’ XXXVI [45, p. 33].
iron sulphate adulterant, which will turn black in the presence
of ferrous salts: perhaps the first recorded use of a spot test for It is interesting to note that sodium chloride is specifically
chemical identification. excluded in this recipe, which suggests that many recipes did
indeed use honey and salt. Banik [46] replicated this recipe
Theophilus, writing in his De Diversis Artibus in the twelfth
and found the product to be a complex mixture of basic copper
century AD, [45], records recipes for a viride salsum and a
salts: besides copper acetates and carbonates, some basic copper
viride hispanicum. The viride salsum is to be made as follows:
chlorides were also formed.
‘If you wish to make a green colour take a piece of oak
The Mappae Clavicula, representing a varied body of
of whatever length and width you like, and hollow it
knowledge compiled from the eighth to twelfth centuries AD
out in the form of a box. Then take some copper and
[48], mentions seven different recipes for the preparation of
have it beaten into thin sheets, as wide as you like but
various copper corrosion products, not all of which would
long enough to go over the width of the hollow box.
produce common verdigris. One of the relevant passages is:
After this, take a dish full of salt and, firmly compressing
it, put it in the fire and cover it with coal overnight. The

83
 R EVIEWS IN CONSERVATION N UMBER 2 2001

‘If you wish to make a different azure, take a flask of These three recipes are broadly similar in taking sheets of
the purest copper and put lime into it, half way up and copper and exposing them over warm vinegar. A series of
then fill it with very strong vinegar. Cover it and seal it. experiments carried out in the laboratory showed that the final
Then put the flask in a warm place for one month. Not products were neutral verdigris, and in one case [recipe vi], a
as good as ii but it is suitable for wood and plaster’ iii mixture of two basic acetates.
[48, p. 21].
‘Coat copper, beaten out into sheets with honey and put
Orna [49] examined this recipe, utilizing copper plates in a beneath it in a pot broad laths of wood and pour over it
glass container. After reaction for one month, mixed crystals a man’s urine. Let it stand, covered for fourteen days’
of green, blue and colourless hue were found, in addition to a 96 [48, p. 23].
blue reaction product and unreacted starting materials.
This recipe was replicated using oak wood chunks and pure
Elemental analysis suggested an empirical formula of
copper sheet covered with urine in a jar for fourteen days. A
Cu(CH3COO)2.Ca(CH3COO)2.6H2O, calcium copper acetate
green product was only formed in small patches over the copper
hexahydrate, a water-soluble, deep blue crystalline solid, which
sheet; of a darkish green colour, the product appeared to be
becomes very pale blue when ground to a powder. This
poorly crystalline, since it gave no powder x-ray diffraction
identification was confirmed by XRD; the data for which is
pattern. Examination of the same product with FTIR showed
shown in Table 8. This pigment may have proved useful on
that the spectrum contained significant peaks associated with
wood or plaster, although the original text does state that it is
nitrogen bonds, showing one of the major products must be a
not a high quality pigment. Despite Wallert’s observation [50]
compound between copper (II) and one of the nitrogenous
that the fifteenth century manuscript, the Libro Secondo de
compounds of urine, such as urea, uric acid or creatinine. This
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

Diversi Colourie Sise da Mettere a Oro, known as the Simone

is not unexpected, given the fact that the reactivity of copper
manuscript, is not particularly related to the Mappae Clavicula
ions with both oxygen and nitrogen groups: the identity of the
tradition, it does contain a closely related recipe for a blue
compound formed is, at time of writing, under further research.
pigment that goes as follows:
‘Changing copper- take six minas of salt and four minas
‘To make a beautiful blue at little cost. Take quicklime
of filings or scrapings of copper. Mix the filings in a pot
and green and ground verdigris and sal ammoniac, as
with ground salt, sprinkling vinegar over them and leave
much of one as of the other. Grind these all together
for three days and you will find it has turned green’ 80
with urine and you will see a beautiful blue. Temper it
[48, p. 34].
with the previously described glair, when you want to
work with it’ recipe 39 [50, p. 23]. A laboratory experiment to examine this recipe found that
the copper scrapings did indeed turn green, with the principal
The ‘green’ referred to in the text is an organic sap green,
product identified as atacamite. All verdigris preparations using
while, as we have seen with the Mappae Clavicula recipe iii,
copper smeared with honey to which salt is added and hung
reactions with lime, or in this case with quicklime (calcium
over vinegar, were found to produce principally atacamite.
oxide) are quite likely to produce complex or unusual salts
such as calcium copper acetate hexahydrate. Here, the addition Orna et al. [25] discuss some important manuscript
of sal ammoniac suggests that mixtures of the calcium copper information from the ninth to sixteenth century AD. A Latin
acetates and copper trihydroxychlorides are quite possible, since manuscript of the eighth century AD, The Codex Lucensis 490
ammonium chloride additions tend to result in the replacement at Lucca, variously known as the Compositiones variae, dates
of the acetate groups with basic chlorides. Grinding in urine from eighth to ninth century AD, One recipe calls for:
and the addition of the organic green add to the probable
‘...concerning flowers of copper (iarin). Take pure copper
complexity of the products of this particular recipe, due to
plates and hang them over strong vinegar. Place them in
combination of the copper with organic plant substances and
the sun without touching them. After two weeks, open
with the urine components. In fact, this proved to be the case
the container and take out the plates. Collect the
and a laboratory synthesis of recipe 39 produced soft, turquoise
efflorescence and you will have very pure iarin...’ [25].
blue particles which did not give an identifiable x-ray powder
diffraction pattern. Other recipes from the Mappae Clavicula There are many variants to preparing verdigris just with
continue: copper and vinegar. For example, Mactaggart [26] prepared
basic verdigris using a thick cotton cloth soaked in malt vinegar,
‘Green. If you want to make Byzantine Green, take a
packed in a plastic pot with sheets of copper foil, covered, and
new pot and put sheets of copper in it then fill with very
left at room temperature for six weeks; the data for this salt is
strong vinegar, cover and seal, leave for six months. The
given in Table 7. A verdigris sample prepared by the same
product can be dried in the sun’ v [48, p. 27].
method in the GCI laboratory produced a very different pattern,
Additionally, recipe x mentions that Byzantine green can be a second attempt used the same ingredients, but with burial of
tempered with vinegar, which would convert the basic verdigris the copper foil wrapped in cloth soaked in vinegar, which
salts into neutral verdigris, which would help to attain a purer successfully produced both crystals of neutral verdigris and
product upon re-crystallization. the lighter blue basic verdigris salts, characterized as similar to
compound B.
‘Rouen Green. Take sheets of copper, smear with soap.
Put the sheets into a pot, fill with vinegar, cover, seal Manuscripts from the fourteenth to sixteenth century AD
and place in a warm place for fifteen days’ vi [48, p. 27]. reveal a detailed knowledge concerning the preparation of blue
copper-based pigments. Most of these manuscripts prescribe
‘Take copper strips and scrape them down well and hang
verdigris as the initial material to use instead of copper metal.
them over vinegar. Scrape off and gather the stuff that
A fifteenth-century manuscript, MS 1243 in the Biblioteca
collects on it’ 221-D [48, p. 28].

84
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

Table 8. XRD data for calcium copper acetate hexahydrate, laboratory synthesis compared with ICDD 32-159

Dark blue crystals from Calcium Calcium Copper

Acetate + Copper Acetate Acetate Hydrate
(4 Ca: 1 CU) Molar solutions (CH3COO)4CaCu.6H2O
JCPDS 32-159

I 4 theta d d I
5 118.44 9.6 9.2 1
100 21.28 8.3 8.11 12
100 23.27 7.6 7.89 100
40 30.43 5.82 -
40 32.47 5.46 5.58 30
10 36.16 4.91 4.77 2
10 37.81 4.69 4.60 1
4.06 1
20 47.55 3.74 3.67 12
90 49.56 3.59 3.611 11
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

40 50.96 3.50 3.528 35

20 52.91 3.37 -
20 54.53 3.27 3.284 4
3.236 1
20 57.57 3.10 -
20 59.09 3.02 3.040 6
10 61.91 2.889 -
20 63.45 2.820 2.831 3
30 67.09 2.671 2.687 11
2.670 1
2.628 2
10 70.33 2.552 2.561 1
10 72.20 2.488 2.495 4
5 73.87 2.434 2.447 1
20 77.71 2.318 2.2995 2
10 81.40 2.217 2.2108 1
2.1882 3
40 85.07 2.125 2.1261 6
20 87.96 2.059 2.0663 1
20 89.50 2.025 2.0312 4
20 91.60 1.981 1.9721 1
1.9693 1
50 94.61 1.922 1.9266 3
1.9136 5
10 97.90 1.861 1.8595 2
50 102.97 1.775 1.7742 1
1.7643 3
1.7602 3
20 104.87 1.745 1.7375 2
… more d values
1.5473 1

85
 R EVIEWS IN CONSERVATION N UMBER 2 2001

Riccardiana, translated by Merrifield [27] gives a typical recipe Merrifield [27] records a great deal of relevant information
for a durable azure: from a Paduan manuscript Ricitte Per Far Ogni Sorte di Colouri,
MS 992 from the library of the University of Padua. The
‘...mix well one part of sal ammoniac and three parts of
manuscript is Venetian and probably dates to the middle or
verdigris with oil of tartar until it is soft and paste-like,
late seventeenth century. By this stage, the chemical procedures
or even softer. Then place it in a glass vessel under hot
are becoming quite complex and some of the interpretation is
dung for a day; afterwards, you will find that the green
difficult, as can be gauged from the following recipes:
has turned to best blue. Another way of making the best
blue. Or: mix together three parts of sal ammoniac and ‘To make a good green of verderame - Take 10 parts of
six parts of verdigris with oil of tartar until soft and verdigris, 2 of corrosive sublimate, 1/2 a part of saffron,
paste-like or even softer. Then place the paste into a glass 1/4 of galls of Istria, and 1/2 of sal ammoniac, grind
ampoule, and when it is well-stoppered and sealed, place them up with very strong vinegar (distilled vinegar is
it in a hot oven and let it stand for some days; afterwards, the best), put them into a glass vase, and when the
take it out and you will find it turned into the best vinegar is clear and coloured, let it be decanted and
azure...’ [27, p. 234]. evaporated in a glazed vase. Then pour fresh vinegar on
the remainder, mix again and do as before, until the part
‘Oil of tartar’ is apparently a saturated solution of potassium
which settles ceases to colour the vinegar, and if you
carbonate. Both methods mentioned in this recipe produce
pour the coloured vinegar into shallow open vases, it
bright blue needles or elongated prismatic crystals mixed with
will dry much quicker either in the sun or in the shade.
colourless crystals of varying morphology. The authors
When dry remove the colour gently by dipping the brush
replicated the synthesis in the GCI laboratory using sal
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

in the vinegar, and afterwards grind up the colour with

ammoniac, verdigris and oil of tartar, which produced a dark
a little roch alum and a little gum arabic, so as to make
blue paste. This was then placed in a sealed glass jar at 40ºC
it into a cake’ 16 [27, p. 134].
for five days. The synthesis appears to produce a mixture of
products: one bright blue clump of crystals was ground up Since corrosive sublimate is mercuric chloride, it may be that
and analysed by XRD, with possible matches being potassium this recipe is corrupted by incorrect copying of the original
copper acetate 2K(CH3COO).Cu(CH3COO)2 and ammonium source: the synthesis was not attempted in the laboratory: it is
chloride and/or ammonium copper acetate acetic acid [20-1499] interesting however, how common the ingredients of alum and
4NH4(CH3COO).Cu(CH3COO)2.4CH3COOH [51]. Clearly, ammonium chloride are in these synthetic greens and blues, as
this medieval recipe produces a blue product of somewhat the following Paduan manuscript recipe describes:
intractable nature.
‘To make another brilliant green - Take oz. vi of the
Orna et al. [52] examined the recipes for blue pigments best verdigris, oz. ij of tartar of Bologna, and dr. j ss. of
requiring silver as a starting material and conclusively showed roch alum.; pulverize the whole, and grind each article
that the medieval silver contained some copper and that the separately, then grind them together rather stiffly with
copper content was responsible for the formation of the distilled vinegar, put the powder into a glass vase with a
requisite substances. The Mappae Clavicula, entry ii, in little saffron, and expose it to the sun. Then pour on it a
describing how to make the best azure records: bocale of distilled vinegar, and the longer it is exposed
to the sun the more beautiful will be the colour’ 17 [27,
‘...take a new pot that has never been used for any work
p. 210].
and set in it sheets of the purest silver, as many as you
want, and then cover the pot and seal it. Set the pot in The recipe above calls for ‘tartar of Bologna’, which may be
the must that is discarded from a wine press, and there approximated by the modern cooking aid ‘Cream of Tartar’
cover it well with the must and keep it well for fifteen (sodium potassium tartrate) and roch alum can be substituted
days. Then uncover the pot and shake the efflorescence with potassium aluminium sulphate. Grinding these ingredients
that surrounds the sheets of silver into a bowl...’ [48, together does indeed make a strongly coloured green solution,
p. 18]. with evaporation producing tiny pale green crystals.
Pure silver coupons gave no reaction in the laboratory trials ‘How to refine verdigris - Take the verdigris, grind it
using 5M acetic acid, whilst alloys with some copper content well, steep it in the best vinegar for 3 or 4 days, strain
produced cupric acetate monohydrate, neutral verdigris. Alloy it, then pour the strained liquid on other well-ground
samples treated with acetic acid and horse dung were also cupric verdigris; let it settle for 2 days more, strain it again
acetate, but apparently of a different structure than those gently, leaving the lees of the verdigris at the bottom of
formed with acetic acid alone. Single crystal x-ray the vase; put the liquid which has been strained in a
crystallography work showed that the structure was identical glass vessel with a little saffron, and keep it well covered’
with Cu2(CH3COO)4.2H2O, tetra-µ-acetato-bisdiaquocopper 32 [27, p. 246].
(II), a dinuclear species, whose genesis, Orna states, may be
This recipe is fairly straightforward in intention: the crude
due to the slower reaction rate for the silver-copper coupon
basic verdigris is dissolved in good quality vinegar and the
exposed to the acetic acid and horse dung, compared with acetic
residue decanted from the solution, which is then left to slowly
acid alone. This is a little perplexing, since many copper
evaporate in a covered vessel: this will produce a crystalline
compounds, such as copper acetates, are actually dimeric in
deposit of refined, neutral verdigris. However, the continued
structure and the formula shown above for the dinuclear species
reference to the presence of saffron in some of these recipes
is, in fact, simply neutral copper acetate monohydrate stated
deserves to be discussed. Barkeshli [53] notes the use of saffron
in a more complicated way.
additions to verdigris in Persian miniature paintings, one such

86
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

Table 9. XRD data for copper citrate pentahydrate prepared by reaction of verdigris with lemon juice (Merrifield, Paduan MS
paragraph 33) compared with ICDD 1-191.

verdigris in lemon juice after Merrifield Copper citrate

Paduan MS paragraph 33 C12H10Cu14.5H2O
JCPDS 1-191

I 4 theta d d I
10 21.55 8.2 –
70 24.58 7.2 7.4 25
100 31.43 5.64 5.6 100
10 33.56 5.28 – –
10 35.64 4.98 – –
60 38.67 4.59 4.55 25
60 45.93 3.87 3.85 25
10 49.28 3.61 – –
10 50.69 3.51 3.56 12
10 56.86 3.14 3.13 8
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

10 61.13 2.925 2.92 8

10 67.52 2.655 2.66 8
10 71.34 2.517 2.49 4
10 74.18 2.424 2.14 2
2.19 8
10 86.90 2.083 – –
10 90.41 2.006 1.99 8
10 93.58 1.941 1.93 8
10 96.50 1.886 1.88 6
10 99.55 1.832 1.82 4
10 106.94 1.714 1.70 4
10 110.76 1.659 1.66 4
5 112.98 1.629 – –
1.56 2
5 123.92 1.498 – –
10 127.95 1.455 – –
10 130.18 1.433 – –
10 132.23 1.413 – –
10 136.96 1.370 – –
10 141.62 1.331 1.33 2

recipe, from the sixteenth-century Medad-al Khotoot, states would be able to complex with verdigris to form copper
that: carboxylates, whose formation may lead to enhanced stability
of the pigment and render the copper (II) ions less labile.
‘...for using verdigris, add gum arabic and anzarut gum
[an Iranian equivalent to sarcocolla]. The verdigris The seventeenth-century Venetian manuscript [27] continues
which is made out of yogurt chars paper. To prevent with another transmutation of verdigris:
this problem add a small amount of saffron to make the
‘A most beautiful green colour - Take the powdered
pigment stable’ [53].
verdigris, dissolve it with lemon juice, and let it settle
Barkeshli [53] investigated the possible action of saffron as a for 24 hours; then strain the most fluid portion very
buffer and discovered that in both acidic and alkaline media, carefully, leaving the lees at the bottom of the vase. Put
with acetate ions, that the saffron did indeed possess the ability the strained liquid into a glass vase, and add to it a little
to act as a buffer. The colouring matter of saffron consists of the above-mentioned pasta verde, let it dry, and when
principally of safranal (2,6,6-trimethyl-1,3-cyclohexadiene-1- you use it, add to it some more lemon juice, and the
carboxaldehyde), and crocin,(α–crocin, the di-gentiobiose ester more you add the more beautiful the colour will be, so
of crocetin, which is a carotenoid-dicarboxylic acid). As well that it will be like an emerald; take care, however, that
as acting as a buffer, crocin contains a region of many you do not permit the brush to touch water’ 33 [27, p.
conjugated C=C bonds and carbonyl groups at each end which 268].

87
 R EVIEWS IN CONSERVATION N UMBER 2 2001

This recipe was replicated in the laboratory, using distilled had been used for writing green letters on a manuscript, since
verdigris and freshly squeezed lemon juice. After mixing, the the product is so complex, even without possible slow changes
solution was allowed to stand for 24 hours and decanted from over time.
the sediment of lemon juice fibres. After a few hours a light
In China, verdigris was being prepared both as a pigment
green crust began to crystallize out on the surface of the dish,
and for medicinal use by at least the early centuries BC [55]
and this deposit was removed for analysis. Eventually the
using the familiar technique of exposing copper to the fumes
remainder of the solution solidified into a sticky green mass.
of warm vinegar. Verdigris solutions were used too, for
The identity of the green deposit removed from the surface
application to the surfaces of wooden objects as a stain and to
was determined by Debye-Scherrer powder XRD to be cupric
prevent biodeterioration of the wood. A synthetic blue was
citrate pentahydrate, ICDD 1-091. Citric acid (2-
also prepared as follows:
hydroxypropane-1,2,3-tricarboxylic acid), is, of course, one
of the major components of lemon juice and the transformation ‘...to get a blue pigment from copper one must mix three
of distilled verdigris to cupric citrate in this recipe is not tean [tens of an ounce] of the rust of copper with
unexpected. The X-ray powder diffraction data is given in Table seventeen tean of sal ammoniac and boil this mixture
9. Colour measurements of the two compounds ground as with pure water. Hhiene-pann who lived in the Han
pigment preparations in gum arabic were made using the Dynasty [200 BC - 200 AD] was the inventor of this
Minolta CM-1000 visible spectrophotometer. pigment’ 5 (III) 33 [55, p. 135].
The colour shift towards green, compared with distilled Assuming that the rust of copper is copper (I) oxide, cuprite,
verdigris can be seen in the colour reflectance spectra. Verdigris the recipe was followed with 1.5 grams of cuprite and 8.5 grams
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

has different L*a*b* coordinates and the peak maxima of of ammonium chloride with the result that the light blue-green
copper citrate is around 520nm, while neutral verdigris is, at precipitate that formed was identified as quite pure atacamite,
495nm, significantly more into the blue range of the spectrum, whose x-ray diffraction proved to be identical to atacamite.
confirming the validity of the observations made in the recipe.
Needham [55] mentions a text by Sun Ssu-Mo who refers to
There are no identifications of copper (II) citrate used as a
the colouration of verdigris with indigo to give it a bluer colour
manuscript or painting pigment, to the authors’ knowledge,
which Needham calls ‘malachite’, but since malachite is green
but positive identification would require removal of a small
perhaps azurite is really implied, which would be useful in both
sample and this is rarely possible in manuscript studies,
fresco and scroll painting A Chinese text originally translated
especially since the identification of a copper citrate green from
by Klaproth states:
a very small sample could prove to be problematic. The pasta
verde that is referred to in this manuscript recipe is sap green, ‘...to get a fine green pigment from copper one must
added to the preparation to produce an even brighter green calcine the rust [to make a copper oxide] and then boil
colour, and may have complexing capabilities for coordination it with white alum in a sufficient amount of water. After
of the copper ion. it has cooled it will be green, and one must add some
natron solution [a naturally occurring mixture of sodium
Turner [54] draws attention to the recipe book of Johannes
carbonate, sulphate and chloride] which will precipitate
Alcherius, whose treatise dates from around AD 1411. He not
the green colour called hsiao lu se. This is used in
only gives a recipe for a copper citrate, but also for copper
painting for the colour of plant and bamboo leaves...’ 5
tartrates and various combinations of copper greens mixed with
(III) 33 [55, p. 136].
organic dyestuffs (38: 3–9, 45, 90, 95). A laboratory replication
of one of these copper tartrate recipes produced pale blue-green Needham equates this colour to a green verditer (malachite),
crystals, which were a very close match to copper (II) tartrate the same as the Chinese thung lu. The reaction was replicated
trihydrate (D-2,3-dihydroxybutanedioic cuprate (II) trihydrate) in the laboratory: cuprite, as the calcined copper rust, dissolves
[ICDD 1-158]. The recipes compiled by Alcherius from one very slowly in the hot alum solution: here potassium aluminium
illuminator, Antoine de Compeigne, are also interesting for their sulphate was used as the white alum reagent. Partial dissolution
observations on the corrosive nature of some of these green is effected after several hours of gentle boiling, forming a blue-
pigments, and de Compeigne suggested neutralizing these green solution with a deposit of unreacted cuprite. The exact
properties by the addition of gladiolus juice (38: recipes 300, nature of the reaction products of this solution with natron is
301). probably dependent on the composition of the natron
employed. For example, if the natron is principally a mixture
For the manufacture of an ink, the Ricitte Per Far Ogni Sorte
of sodium carbonate and sodium bicarbonate, then a gelatinous
di Colouri advises:
duck-egg blue precipitate forms when the natron is added to
‘To write green letters - Take strong vinegar, powdered the cooled solution. If the natron contains some sodium chloride
gum arabic, verdigris, and roch alum, all at discretion, and sodium sulphate too, then blue-green crystalline deposits
mixed with the juice of rue, and made so liquid that the can be obtained from the precipitate by filtering and drying.
mixture will flow like ink’ recipe 69 [54, p. 45]. Some recent analyses at the British Museum, London, have
suggested that in many cases, samples labelled as ‘natron’ are,
The mixture did indeed produce a green liquid that could be
in fact, principally composed of sodium chloride. The alum
painted on card, and clearly produced a brighter green colour
content of the recipe reacts to form aluminium sulphate,
that verdigris alone. Juice of rue would add to the solubilization
aluminium chloride or other salts, and stays in solution:
of the mixture, since the porphyrin ring structure of the
aluminium hydroxide is not precipitated from the solution.
chlorophyll in the plant juice could replace copper for
magnesium in the complex, which would help to make a The precipitate gave an XRD pattern matching that for
brighter green ink for writing manuscript lettering. It would chalconatronite, Na2Cu(CO3)2.3H2O. There are no published
be very difficult to determine by analysis if this concoction papers describing the use of chalconatronite as a pigment in

88
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

Chinese paintings as far as the author is aware, but if there or corrosion products of copper resulting from exposure to
had been, there would have been the suspicion that it was found formic acid pollutants within the museum. The same argument
because of alteration from another pigment, such as malachite. applies to the copper (II) acetates, although here the primary
However, the recipe replicated here manifestly shows the interest in the verdigris compounds is as pigments rather than
possibility of finding chalconatronite as an original synthetically corrosion products.
prepared pigment, and not because it had formed by alteration.
The review and replication of historic recipes for making
Banik [47] mentions identifying a mixture of pseudo-malachite
verdigris has shown than many of them produce a variety of
and chalconatronite as the green pigment on a sixteenth-century
basic verdigris, or a mixture of salts. Several recipes do not
illuminated manuscript, which may be an indication that
make copper acetates at all, and instead produce one of the
chalconatronite could have been used as a pigment more
copper trihydroxychlorides, such as atacamite or clinoatacamite
commonly than the present literature would suggest.
[58, 59]. Despite the chloride ion content of these pigments,
Bukhari [56] discusses an Indian manuscript describing there is some evidence to suggest that they are far less
pigment recipes, prepared during the reign of the Mughal ruler, responsible for the degradation of parchment or paper supports
Aurangzeb, known as the Asrarul Khat, written by Fadlu’llah than distilled verdigris. The use of plant juices, such as rue or
Ansari wal Faruqi in AH 1102 (AD 1690). In this manuscript gladiolus, or plant extracts, such as saffron, may not only alter
a good quality of zangar (verdigris) is obtained from recipe 4: the colour of the applied verdigris pigment, but may provide
enhanced stability for the copper and acetate ions in preventing
‘...take one rati of naushagar (sal ammoniac) and half
chemical deterioration of cellulosic or proteinaceous material.
rati of copper scraps, put them in a pot and pour grape
vinegar drop by drop into the vessel and with the help As alchemical knowledge became more sophisticated during
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

of a stick, whose top should be flat, grind the admixture the period from the eighth to sixteenth century AD, so the
in the pot till it becomes zangar...’ [56, p. 3]. recipes for preparing verdigris compounds becomes more
complex, resulting in considerable difficulties in
This recipe will produce a mixture of the basic copper
characterization of the products involved. As a result of these
chlorides, principally atacamite, Cu2(OH)3Cl with copper
difficulties, the complete characterization of some copper green
acetates, depending on the extent to which the naushagar has
and blue pigments in works of art will therefore continue to be
acted to convert the copper scraps to the chloride, before the
a problem for some time to come.
grape vinegar is added; almost certainly the final product from
the original recipe will actually be a complex mixture of salts.
In fact, a laboratory replication of the recipe managed to References
produced neutral verdigris, rather than atacamite, but clearly 1 Gmelins Handbuch der Anorganischen Chemie, Kupfer:
the basic copper chlorides are often associated with these Systemnummer 60 Bd. Cu., Verlag Chemie, Weinheim, Vol. A,
verdigris recipes. Section 1, 1955.
2 Gmelins Handbuch der Anorganischen Chemie, Kupfer:
The same manuscript gives another recipe for the preparation Systemnummer 60 Bd. Cu., Verlag Chemie, Weinheim, Vol. B,
of ordinary verdigris as: Section 2, 1961.
‘...put the specks of copper in a copper pot, crush them 3 Gmelins Handbuch der Anorganischen Chemie, Kupfer:
Systemnummer 60 Bd. Cu. [B], Verlag Chemie, Weinheim, Vol.
with curd and let the mixture be dried up for three days
B, Section 3, 1965, pp 962-7.
and nights. The zangar will be ready for use...’ [56, p. 4].
4 Gmelins Handbuch der Anorganischen Chemie, Kupfer:
Banik [57] states that the use of sour milk instead of vinegar Systemnummer 60 Bd. Cu. [B], Springer-Verlag, Berlin and New
was common in Russia and Serbia too, and that the chemical York, Vol. B, Section 4, 1966, pp 1555-64.
composition of the prepared pigment was similar to malachite. 5 International Centre for Diffraction Data, ICDD Powder
Diffraction File - Organic Phases, ICDD, Joint Committee on
A recipe, dating to the eighteenth century, for an artificial blue
Powder Diffraction Standards, Swarthmore, Pennsylvania, 1982.
pigment actually produced posnjakite Cu4SO4(OH)6.H2O rather
6 Fowles, G., ‘Note on the basic copper formates’, Journal of the
than a carbonate or acetate. Chemical Society 107, 1915, pp. 1281-2.
The copper chlorides are often mentioned as being particularly 7 Graedel, T.E., McCrory-Joy, C. and Franey, J.P., ‘Potential
destructive for paper and parchment, but in the study by Banik, corrosion of metals by atmospheric organic acids’, Journal of the
Electrochemical Society 133, 1986, pp. 452-3.
there was no proof of a connection between the presence of
basic copper chlorides and the decomposition of the parchment 8 Pey, I., ‘The Hafkenscheid collection’ in Hermens, E., Ouwerkerk,
A. and Costaras, N., eds, Looking through Paintings, Archetype
substrate. Publications, London and Uitgeverij de Prom, Baarn, 1998, pp.
465-500.
Conclusions 9 Gauthier, J., Etude de quelques propriétés des sels neutres et
basiques de cuivre des acides formique, acétique, propionique &
The review of the verdigris pigment group has shown the Propositions données par la Faculté, Ph.D. dissertation, Procede
considerable complexity of the basic acetate salts, and has Sertic, l’Université de Paris, Lyon, 1958, pp. 17-19; 27-43.
highlighted the need for further chemical research into their 10 Rahn-Koltermann, G., Buss, D.H., Fuchs, R. and Glemser, O.,
structure, properties and synthesis. The information derived ‘Zur Kenntnis basischer Kupferacetate’, Zeitschrift für
from the review and from the laboratory work undertaken has Naturforschung 46b, 1991, pp. 1020-24.
helped to update and correct the powder x-ray diffraction 11 Yamanaka, S., Preparation and properties of anion-exchangeable
layered basic salts, Nippon Ion Kokan Gakkaishi 7, 1996, pp. 8-
information and has shown, in passing, that FTIR can be used
17 (Japanese text).
to differentiate between the neutral and basic verdigris group.
12 The ICDD Powder Diffraction File-Inorganic Phases, ICDD, Joint
Revised data is also presented here for the copper (II) formates, Committee on Powder Diffraction Standards, Swarthmore,
which should be of use in future examination of formate salts Pennsylvania, 1992.

89
 R EVIEWS IN CONSERVATION N UMBER 2 2001

13 Schweizer, F. and Mühlethaler, B., ‘Einige grüne und blaue 34 Gopfrich, J., ‘The granulated donkey? Shagreen: some aspects of
Kupferpigmente: Herstellung und Identifikation’, Farbe und Lack conservation’ ICOM Committee for Conservation, 12th Triennial
74, 1968, pp. 1159-73. meeting, Lyon, Vol. II, James & James Ltd, London, 1999,
14 Kuhn, H., ‘Verdigris and copper resinate’ in Roy, A., ed., Artists’ pp. 685-90.
pigments, a handbook of their history and characteristics, 35 Montet, J., ‘Second Mémoire sur le vert de gris’, Histoire et
National Gallery of Art, Washington and Oxford University Press, mémoires de l’Académie Royale des Sciences Vol. 1753, 1757,
New York, vol. 2, 1993, pp. 131-58. pp 609-11.
15 Rasti, F and Scott, G., ‘The effects of some common pigments on 36 Weser, U., ‘Biochemical basis of the use of copper in ancient
the photo-oxidation of linseed oil-based paint media’, Studies in Egyptian and Roman medicine’ in Black, J., compiler, Recent
Conservation 25, 1980, pp. 145-56. Advances in the Conservation and Analysis of Artifacts, Summer
16 Ioakimoglou, E., Boyatzis, S., Argitis, P., Fostiridou, A., Schools Press, London, 1987, pp. 189-93.
Papapanagiotou, K. and Yannovits, N., ‘Thin-film study on the 37 Ebers, G., Papyros Ebers: die Arzneimittel der alten Ägypter in
oxidation of linseed oil in the presence of selected copper hieratischer Schrift, Verlag W. Engelmann, Leipzig, 1875.
pigments’, Chemical Materials 11, 1999, pp. 2013-22. 38 Ebbell, B., The Papyros Ebers, Oxford University Press, London
17 Banik, G., ‘Discoloration of green copper pigments in manuscripts 1937.
and works of graphic arts’, Restaurator 10, 1989, pp. 61-73. 39 Pliny the Elder, The Natural History, tr. H. Rackham, Harvard
18 Masri, M.S. and Friedman, M., ‘Effect of chemical modification University Press, Cambridge, MA and William Heinemann,
of wool on metal ion bonding, Journal of Applied Polymer Science London, 10vols [original edition 1933] 1979.
18, 1974, pp. 2367-77. 40 Buchner, J.A., Repertorium fur die Pharmacie: Uber das
19 Williams, R.J.P., ‘Copper-protein compounds’ in Welch, A.J. and Kupfersauerhonig (Unguentum Aegyptiacum), J.L. Schrag,
Chapman, K., eds, The Chemistry of the Copper and Zinc Triads, Nürnberg, 1816, p. 143.
Royal Society of Chemistry, Cambridge, 1993.
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

41 Kobert, R., ‘Uber den jetzigen Stand der Frage nach den
20 Daniels, V. and Leese, M., ‘The degradation of silk by verdigris’, pharmakologischen Wirkungen des Kupfer’, Deutsche Medikal
Restaurator 16, 1995, pp. 5-63. Wochenschrift 21, 1895, pp. 5-7.
21 Khandekar, N. and Phenix, A., ‘Some observations on the effects 42 Benhamou, R., ‘Verdigris and the entrepreneuse’, Technology and
of a selection of pigments on an artificially aged egg tempera Culture 25, 1984, pp. 171-81.
paint film’, AICCM Bulletin, 2001 (in press). 43 Dioscorides, The Greek Herbal of Dioscorides, ed. R.T. Gunther,
22 Scott, D.A., Khandekar, N., Schilling, M.R., Turner, N., Tanigushi, Hafner Publishing, London and New York, [original edition,
Y. and Khanjian, H., ‘Technical examination of a fifteenth-century 1933] 1968.
German illuminated manuscript on paper: a case study in the 44 Shier, D., Butler, J. and Lewis, R., Hole’s Human Anatomy and
identification of materials’, Studies in Conservation 46, 2001, Physiology, W.C. Brown, Chicago, 1996, p. 213.
pp. 93-108.
45 Dodwell, C.R., ed., Theophilus: De Diversis Artibus, Thomas
23 Woudhuysen-Keller, R., and Woudhuysen, P., ‘Thoughts on the Nelson, London, 1961.
use of the green glaze called ‘copper resinate’ and its colour-
changes’ in Hermens, E., Ouwerkerk, A. and Costaras, N., eds, 46 Banik, G., ‘Discolouration of green copper pigments in
Looking through Paintings, Archetype Publications, London and manuscripts and works of graphic arts’ Restaurator 10, 1989,
Uitgeverij de Prom, Baarn, 1998, pp. 133-44. pp. 61-73.

24 van T’Hul-Ehrnreich, E.H. and Hallebeek, P.B., ‘A new kind of 47 Maekawa, S. and Mori, H., Theophilus, De Diversis Artibus: A
old green pigment found’, ICOM Committee for Conservation, Japanese Translation of the First Volume, Bijyutsushi, Tokyo,
3rd Triennial meeting, Madrid, Paris, 1972, pp. 2-8. 1965, p. 25.

25 Orna, V.A., Low, M.J.D. and Julian, M.M., ‘Synthetic blue 48 Smith, C.S. and Hawthorne, J.G., ‘Mappae Clavicula: A little
pigments: Ninth to sixteenth centuries, 2 - silver blue’, Studies in key to the world of medieval techniques’ Transactions of the
Conservation 30, 1985, pp. 155-60. American Philosophical Society, NS, Vol. 64, 1974, pp. 3-127.

26 Mactaggart, P., Letter to the author, Los Angeles, September 22, 49 Orna, M.V., ‘Copper-based synthetic medieval blue pigments’ in
1996. Archaeological Chemistry, American Chemical Society,
Washington, D.C., 1996, chapter 9, pp. 107-115.
27 Merrifield, M.P., A Treatise on Painting, Written by Cennino
Cennini in the Year 1437, and First Published in Italian in 1821, 50 Wallert, A., ‘The Libro Secondo de Diversi Colourie Sise da
with Introduction and Notes by Signor Tambroni, E. Lumley, Mettere a Oro: A Fifteenth-century technical treatise on
London, 1844. manuscript illumination’ in Wallert, A., Hermens, E. and Peek,
M., Historical Painting Techniques, Materials and Studio Practice,
28 Eastlake, C.L., Materials for a History of Oil Painting, Vol. 1, Getty Conservation Institute, Los Angeles, 1995, pp. 38-47.
Longman, Brown, Green and Longmans, London, 1847, p. 458.
51 Hardt, H.-D. and Streit, G., ‘Zur Kenntnis der
29 Cennino d’Andrea Cennini da Colle di Val d’Elsa, Il libro dell’arte, Alkalikupferacetate’, Zeitschrift fur anorganische allgemeine
tr. D.V. Thompson, Yale University Press, New Haven, Chemie 350, 1967, pp. 84-91.
Connecticut, 1932.
52 Orna, V.A., Low, M.J.D. and Baer, N.S., ‘Synthetic blue pigments:
30 de Mayerne, Sir T.T., Pictoria sculptoria & quae subalternarum Ninth to sixteenth centuries, 1-literature’, Studies in Conservation
artium, 1620: le manuscrit de Turquet de Mayerne, ed. M. Faidutti 25, 1980, pp. 53-63.
M. and C. Versini, Audin, Lyon, 1974 [British Museum, MS
Sloane 2052]. 53 Barkeshli, M., ‘The presence of saffron in Persian miniature
paintings and its use as an inhibitor for the destructive effects of
31 Laurie, A.P., The Pigments and Mediums of the Old Masters, verdigris’, ICOM Committee for Conservation, 12th Triennial
Macmillan, London, 1914. meeting, Lyon, Vol. II, James & James Ltd, London, Vol. II, 1999,
32 Rahn-Koltermann, G., Glemser, O., Oltrogge, D. and Fuchs, R., pp. 489-94.
‘Grünspan, Ein bedeutsames Pigment für das Scriptorium des 54 Turner, N., ‘The recipe collection of Johannes Alcherius and the
Mittelanters’, Naturwissenschaftliche Rundschau 46, 1993, painting materials used in manuscript illumination in France and
pp. 222-7. Northern Italy, c. 1380-1420’ in Roy, A. and Smith, P., eds,
33 Penny, N., Roy, A. and Spring, M., ‘Veronese’s paintings in the Painting Techniques: History, Materials and Studio Practice,
National Gallery, technique and materials: part II’ National Preprints of the Contributions to the Dublin Congress, IIC,
Gallery Technical Bulletin 17, 1996, p. 33-55, n. 32, 33. London, 1998, pp. 45-50.

90
 T HE VERISIMILITUDE OF VERDIGRIS : A REVIEW OF THE COPPER CARBOXYLATES

55 Needham, J., Science and Civilization in China 5: Chemistry and

Chemical Technology, Cambridge University Press, Cambridge, Authors
1974, part 3, section 33, pp. 244-5.
David A. Scott, B.Sc., BA, Ph.D., FIIC, FRSC, is the Senior
56 Bukhari, Y.K., ‘Pigments’, Marg 16, Bombay, 1963, pp. 12-16
Scientist in charge of the GCI Museum Research Laboratory.
57 Banik, G., Stachelberger, H. and Wächter, O., ‘Investigation of
He was a lecturer in conservation at University College,
the destructive actions of copper pigments on paper and
consequences for conservation’ in Science and Technology in the London, Institute of Archaeology, Department of Conservation
Service of Conservation, Preprints of the Contributions to the and Materials Science, from 1981-7. In 1987 he joined the
Washington Congress, IIC, London, 1982, pp. 75-8. GCI as Head of the Museum Services Laboratory. His principal
58 Scott, D.A., ‘A review of copper chlorides and related salts in interests are the analysis of Museum objects, the
bronze corrosion and as painting pigments’, Studies in characterization of pigments, ancient metals and
Conservation 45, 2000, pp. 39-53. microstructure, and the archaeometallurgy of pre-Hispanic
59 Scott, D.A., Copper and bronze in art: Corrosion, colorants and Colombia and Ecuador.
conservation, The Getty Conservation Institute, Los Angeles, 2002
(in press). Yoko Taniguchi, BA in Archaeology, Tsukuba University,
1996; MA in Archaeometry, Tokyo National University of Fine
Arts and Music, 1998, Graduate intern, GCI Museum Research
Laboratory, 1998-9; Assistant of Preventive Conservation
Laboratory, Graduate course of conservation, Tokyo National
University of Fine Arts and Music, 1999.
Emi Koseto, B.Sc. in chemistry from the University of Tokyo
Downloaded by [University Library Utrecht] at 05:00 17 March 2016

(1995); MA in conservation science, National University of

Fine Arts and Music (1998); Graduate Intern, GCI Museum
Research Laboratory, 1999-2000; 2001, Research associate at
the National Museum of Japanese History, Sakura.

91