Access to this full-text is provided by Springer Nature.
Content available from Heritage Science
This content is subject to copyright. Terms and conditions apply.
Nabaisetal. Herit Sci (2021) 9:32
https://doi.org/10.1186/s40494-021-00490-8
RESEARCH ARTICLE
Organic colorants based onlac dye
andbrazilwood asmarkers forachronology
andgeography ofmedieval scriptoria:
achemometrics approach
Paula Nabais1, Maria J. Melo1* , João A. Lopes2*, Márcia Vieira1, Rita Castro1 and Aldo Romani3
Abstract
This work presents the first proof of concept for the use of molecular fluorescence signatures in medieval colours
based on lac dye and brazilwood lake pigments. These two important medieval dyes were tested as markers using
their UV–Visible emission and excitation spectra. These medieval paints had been previously fully characterized
through a multi-analytical approach. In this work, molecular fluorescence spectra were acquired in manuscripts dat-
ing from 12th to 15th c., which were produced in monastic scriptoria or workshops. First, the spectral distribution
and relative intensity of the emission and excitation spectra were discussed in detail by comparison with reference
compounds, including reproductions of paints based on medieval technical texts. It was possible to group the spectra
according to recipe specificities. Then, statistical methods (principal component analysis and hierarchical cluster
analysis) were applied to the same fluorescence spectra and the generated clusters were compared with the previous
ones. Principal component analysis was initially employed to eliminate redundancy in fluorescence data, so minimiz-
ing bias on the hierarchical cluster analysis results. Except for some misplaced spectra, the placement of samples per
group was confirmed. The outliers resulted from either a poor signal to noise ratio or occurred because certain paints
were unique, such as the colour produced by mixing lac dye and brazilwood, which was found in manuscripts from
the Alcobaça monastic scriptorium. Previously, by using infrared or Raman spectroscopies, only lac dye could be
detected. Notably, these paints compare well with a recipe that was reproduced from the text by Jean Le Begue, in
which both dyes were required.
Keywords: Lac dye, Brazilwood, Historical dyes, Brazilein, Laccaic acid, Fluorimetry, HCA, PCA, Conservation, Medieval
manuscripts, Photoluminescence
© The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco
mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/
zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Introduction
Colour, a fundamental attribute of our heritage, is fading
in precious artworks. us, in the last decade, progress
has been made in understanding the complex mecha-
nisms of degradation of historical dyes anchored in the
study of their photophysics and photochemistry [1–16].
Another vital ingredient for the understanding of col-
our stability relies on the reproduction of these ancient
colours, which allows for a detailed characterization of
these intrinsically heterogeneous systems [6, 7]. In this
context, in recent years, molecular fluorescence spec-
troscopy has become a powerful analytical technique in
the field of cultural heritage [5, 7–23]. In addition to its
ability to identify organic dyes in complex matrices, by
probing its environment it collects unique data linked to
the paint formulation (other admixed pigments and dyes,
Open Access
*Correspondence: mjm@fct.unl.pt; jlopes@ff.ulisboa.pt
1 LAQV-REQUIMTE and DCR, Faculty of Sciences and Technology, NOVA
University of Lisbon, 2829-516 Caparica, Portugal
2 Faculty of Pharmacy, iMed.ULisboa-Research Institute for Medicines,
University of Lisbon, Av. Prof. Gama Pinto, 1649-003 Lisbon, Portugal
Full list of author information is available at the end of the article
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 2 of 18
Nabaisetal. Herit Sci (2021) 9:32
binding media and other additives, e.g., as fillers, and
their relative proportions) [21, 22]. us, based on pre-
vious applications of microspectrofluorimetry in heritage
science [5, 17–20, 24], in this work, we intend to develop
a data-driven based method to assist the identification of
organic dyes in medieval artworks, built from a database
of fluorescence excitation and emission spectra, in the
UV–VIS [5, 17–20]. rough the specificities of the paint
formulation (the original recipe and its making), we also
intend to extract useful details on their conservation con-
ditions and their place of production. e study of these
complex systems aged naturally for eight to six centuries,
is supported by previous research, in which we built a
database of molecular fluorescence spectra of recon-
structions of medieval colours that included lake pig-
ments and paints for the following colorants: kermes, lac
dye, brazilwood, and cochineal [19]. e photolumines-
cence reference database that will be built for the present
work, combines for the first-time molecular emission
and excitation data acquired in medieval illuminations
for lac dye1 and brazilwood2 chromophores, Figs. 1, 2,
3. What motivates us to test a reference database based
on original medieval colours, and why lac dye and brazil-
wood paints [25, 26]? First, our interdisciplinary studies
demonstrate that the colour we see today, in the selected
manuscripts, has not been restored nor retouched [5,
7, 17–20, 24, 27, 32–36]. Second, we have studied them
extensively because they represent some of the most
important reds/pinks/purplish colours used in medieval
illuminations [24, 26, 27, 32–37], which was one of the
major arts of the European Middle Ages. By providing an
in-depth knowledge of these systems, this research will
contribute to overcoming the challenge of the identifica-
tion and preservation of organic dyes in works of art.
In the past fifteen years, through a focus on colour,
our team conducted studies of the selected manuscripts
integrating contributions from molecular sciences, his-
tory, art history, codicology, religion and culture [7, 18–
20, 24–27, 32–35, 38–51]. is research, especially on
monastic collections, has demonstrated their cultural and
Fig. 1 Chemical structures of brazilein, laccaic acid A, laccaic acid B, C and E (B, R = CH2CH2OH; C, R = CH2CHNH2COOH; E, R = CH2CH2NH2), laccaic
acid D, and erythrolaccin. For laccaic acid A the main sites for complexation with a metal ion are shown
1 Lac is part of a resinous cocoon secreted by parasitic insects, from the genus
Kerria, on twigs of branches of host trees; for more details please see pp 665
in [25], pp 6–7 in [26] and pp 159–161 in [27]. ese insects are native to the
countries of the southern and south-eastern Asia [25, 28]. e female lac insect
secretes a red resin, sticklac, from which are obtained both the lac dye and the
shellac resin. e complex nature of lac reflects its composition: a terpenoid
resin, which normally represents 68% of the entire matter, a dye that only rep-
resents 10%, and other less representative constituents, such as wax, gluten,
foreign bodies and impurities [29]. e red colour extracted is named lac dye,
which is made of laccaic acids A and B; laccaic acids C, D and E in minor quan-
tities, Fig.1. e refined resin, known as shellac contains the yellowish orange
erythrolaccin, which is responsible for its colour, Fig.1 [20, 27, 30, 31].
2 Brazilwood is a soluble redwood that was used to prepare lake pigments
(dye-metal complexes) for manuscript illumination. e lake pigments were
based on brazilein, Fig.1. In medieval Europe, sappanwood and other soluble
redwoods were known commercially and technically under the name brazil.
For more details see please [24] (p. 256) and [25] (pp. 274–75).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 3 of 18
Nabaisetal. Herit Sci (2021) 9:32
artistic importance, both within Iberian as well as Euro-
pean context [33, 38–42, 44–50]. Our database combines
spectral information acquired in manuscripts preserved
in Portuguese collections, having a Portuguese, French,
or Flemish provenance, and dating from the 12th c. into
the 15th c., Additional file1: Figure S1. e Portuguese
manuscripts were produced in the scriptoria of three
important monasteries, whereas Flemish and French
books of hours in lay workshops, possibly active in flour-
ishing medieval cities [33, 36]. As an example of the
monastic production, the monumental bible of the mon-
astery of the Holy Cross (thereafter Santa Cruz) is one
of the most important medieval manuscripts preserved
in Portugal. Santa Cruz 1 (SC1) is a work of art created
Fig. 2 Lac dye details from manuscripts showing the differences in hues of original paints: from left to right, lac dye in Lv 15 f. 26, SC 1 f. 21v, and SC
20 f. 191. The figures below are macro details of the top row figures (ANTT and BPMP collections)
Fig. 3 Brazilwood details from manuscripts showing the differences in hues of original paints: from left to right, Ajuda Songbook f. 59, IL 15 f. 66,
and ms 22 f. 76v. The figures below are macro details of the top row figures (PNA, BNP and PNM collections)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 4 of 18
Nabaisetal. Herit Sci (2021) 9:32
by the same monastery that prepared Saint Anthony for
the world, its illuminations ornament the Old testament.
Also included in this work, and of uncertain provenance,
is the Ajuda Songbook, the only surviving medieval song-
book of Galician-Portuguese secular poetry, which dates
to the end of the thirteenth or beginning of the fourteenth
century [45]. Accurate dating of the manuscripts is possi-
ble when a colophon is present, but most of the manu-
scripts are dated by scholars based on the illuminations’
style [47–50]; this information is provided in Additional
file1: Tables S1–S5. Molecular fluorescence spectra were
acquired both insitu and in micro-samples (invisible to
the naked eye, see in the Experimental Section for the
average dimensions). We tested both sets and, although
we had a greater number of spectra acquired insitu, we
chose to present in this work the data acquired in micro-
samples, as it is possible to guarantee a greater reproduc-
ibility in this set of samples.
Our modus operandi will be as follows. We will start by
grouping the data, correlating the spectral distribution
(shape) and relative intensity of the molecular fluorescence
spectra of medieval paints with reference samples that have
been prepared with different concentrations of H+ and Al3+,
different fillers, etc. is experimentation, testing different
concentrations of proton and aluminium ion as well as dif-
ferent fillers, is based on our research in medieval recipes:
we show that these are the essential parameters that have
been manipulated to obtain different hues (from deep red
to purples and pinks), as well as different degrees of opac-
ity; for more details please see [7, 24, 32, 35]. en, we will
apply a combination of chemometric methods (principal
component analysis and hierarchical cluster analysis) and
compare the results with our previous grouping. We will
summarize our main results, stepwise, providing the rel-
evant information for the non-expert in photophysics. e
results obtained from the data processing with the chemo-
metric methods will also be discussed taking into considera-
tion the paint formulation previously characterized through
a multi-analytical approach using Fourier Transform Infra-
red microspectroscopy (μFTIR), Raman microscopy (SERS),
X-Ray Fluorescence microspectroscopy (μXRF), FORS
(UV–VIS), colorimetry [24, 27, 32–35, 38–51].
e statistical methods employed in this work have the
objective of disclosing consistent patterns that help us to
differentiate and to cluster the profiles captured by the
analytical methodologies, providing a systematic analy-
sis of the collected data. As molecular fluorescence spec-
tra, in the solid-state, may suffer the influence of external
factors other than those relevant to this work (the signals
of lac dye and brazilwood chromophores), the methods
were selected to provide a reliable and robust analysis of
the signals. To increase this robustness, a previous analysis
of all spectra is carried out to detect and avoid potential
interferences that vary from the existence of external
impurities to a spectrum that has been acquired with a
low signal-to-noise ratio. In this work, hierarchical cluster
analysis and principal component analysis were used in
sequence. Both methods are unsupervised techniques (so
they do not require a “teaching” algorithm) and they are
relatively simple to implement and use. Principal compo-
nent analysis was initially employed to eliminate redun-
dancy in fluorescence data, so minimizing bias on the
hierarchical cluster analysis results. e simplicity of these
methods when compared to other potential approaches
allow for a better interpretation of results and increases the
robustness of the obtained results. Other methods such as
linear discriminant analysis or partial least squares discri-
minant analysis (both supervised) would require a more
elaborated validation purpose to avoid overfitting and
therefore the generation of spurious results. ree-way
methods such as PARA RAC or some versions of support
vector machines would be better fitted if the data were 3D
fluorescence (which is not the case). Hierarchical cluster
analysis (HCA) also has some limitations as the results
are confined to the samples used in the agglomerative or
divisive algorithm. Different algorithms may also condi-
tion the method’s output. e potential effects of data col-
linearity, noisy data, non-systematic data artifacts or even
the existence of missing data can be minimized by apply-
ing a method like principal component analysis (PCA), as
performed in this work, before employing an HCA algo-
rithm. Additionally, HCA is not a model in the usual sense,
so it cannot be preserved and used to assess new samples.
Examples of the applications of these methods in related
works can be found elsewhere [19, 52, 53].
Experimental
Artworks
e manuscripts studied have been kept in Lisbon,
National Library (Biblioteca Nacional de Portugal, BNP),
National Archives (Arquivo Nacional da Torre do Tombo,
ANTT), Palace of Ajuda Library (Palácio Nacional da
Ajuda, PNA); Municipal Library of Porto (Biblioteca
Pública Municipal do Porto, BPMP); Mafra National Pal-
ace, (Palácio Nacional de Mafra, PNM), Additional file1:
Figure S1. Sample description and representative spec-
tral information that was used in this work is provided in
Additional file1: Tables S1–S5 and S7–S12.
Seventeen manuscripts from the twelfth to the thir-
teenth century from scriptoria of the three main Por-
tuguese monasteries,3 São Mamede of Lorvão (ANTT),
3 e nomenclature of the manuscripts was chosen in accordance with previ-
ous publications. Currently, the nomenclature used within the institutions is
as follows:
Lv 5—Ordem de Cister, Mosteiro de Lorvão, códice 5 (ANTT); ALC 238—
alc-238 (BNP); SC 20—Santa Cruz 20 (BPMP).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 5 of 18
Nabaisetal. Herit Sci (2021) 9:32
Santa Cruz of Coimbra (BPMP), and Santa Maria of Alco-
baça (BNP), were selected. Lv 5 (1183–4), Lv 12 (13th c.),
Lv 13 (13th c.), Lv 15 (1201–50) and Lv 50 (1183); SC 1
(1151–1200), SC 20 (early 13th c.) and SC 21 (early 13th c.);
ALC 238 (late 12th c.), ALC 247 (12th c.), ALC 249 (13th
c.), ALC 347 (12th–13th c.), ALC 412 (1257), ALC 419
(12th–13th c.), ALC 421 (12th–13th c.), ALC 427 (12th–
13th c.) and ALC 446 (13th c.) [32, 38–44].
Other eight manuscripts dating from the thirteenth
to the fifteenth century were included. e Ajuda Song-
book (PNA, 13th–14th c.), the winter Breviary ALC 54
(BNP, 14th–15th c.) and six books of hours, ms 22 (PNM,
1400–20), ms 24 (PNM, 1420/1470), IL 15 (BNP, ca.
1450), IL 19 (BNP, 1420–30), IL 21 (BNP, 1460–70) and
IL 42 (BNP, ca. 1470) [45–51].
Micro‑sampling
Micro-sampling of the manuscripts was performed with
a microchisel from Ted Pella microtools under a Leica
KL 1500 LCD microscope, (7.1× to 115× objective) and
a Leica Digilux digital camera, with external illumination
via optical fibers. e dimension of the micro-samples
ranges from 20 and 50 μm and, as such, invisible to the
naked eye. As we have not yet obtained their weight,
although micro-scales were used, we can use their detec-
tion limit to conclude that they weigh less than 0.1μg.
Micro-samples are stored in microscope slides with sin-
gle cavity and covered with a microscope glass slide. ey
are closed with tape (magic tape 3M) and used as sam-
ple holders. Insitu spectra are collected directly from the
sample by opening the cover. ese sample holders are
then stored in a microscope slide tray cabinet, in a dust-
free enclosure (Ted Pella). e cabinet outer shell is white
polypropylene, and the tray rails are polystyrene.
Micro-samples collection under a microscope, ensures
a selective sampling of the dye paint; that is, the micro-
sample will not include parchment support, ground lay-
ers (mainly applied to metallic colours), or any other
external layers. So, we can certify that fillers and other
additives present in the medieval colours belong to the
paint formulation.
Microspectrouorimetry measurements
Fluorescence excitation and emission spectra were
recorded with a Jobin–Yvon/Horiba SPEX Fluorog 3–2.2
spectrofluorometer coupled to an Olympus BX51M
confocal microscope, with spatial resolution controlled
by a multiple-pinhole turret, corresponding to a mini-
mum 2μm and maximum 60μm spot, with 50× objec-
tive. Beam-splitting is obtained with standard dichroic
filters mounted at 45°, in a two-place filter holder. For a
dichroic filter of 570nm, excitation may be carried out
until about 560nm and emission collected after about
580nm (“excite bellow, collect above”). e optimization
of the signal was performed daily for all pinhole aper-
tures through mirror alignment, following the manufac-
turer’s instructions, using a rhodamine standard (or other
adequate references). Fluorescence spectra were cor-
rected only for the wavelength dependence of the excita-
tion-source intensity. For the study of red dyes, two filter
holders with two sets of dichroic filters are employed,
for lac dye the set of 500 and 600nm and for brazilwood
the set of 540 and 600nm. is enables both the emis-
sion and excitation spectra to be collected with the same
filter holder. A continuous 450W xenon lamp, provid-
ing an intense broad spectrum from the UV to near-IR,
is directed into a double-grating monochromator, and
spectra are collected after focusing on the sample (eye
view) followed by signal intensity optimization (detector
reading). e pinhole aperture that controls the area of
analysis is selected based on the signal-to-noise ratio. For
weak to medium emitters, it is set to 8μm, in this work
for very weak signals 30μm spot was also used (pinholes
5 and 8, respectively) with the following slits set: emis-
sion slits = 3/3/3 mm (6 nm bandpass) and excitation
slits = 5/3/0.8 mm (final bandpass of 2 nm). Emission
and excitation spectra were acquired on the same spot.
With our experimental set, usually, excitation spectra are
acquired with a higher S/N than emission spectra. For
more details on the experimental set-up please see [5, 18,
19].
Fourier transform infrared microspectroscopy (microFTIR)
Infrared analyses were performed using a Nicolet Nexus
spectrophotometer coupled to a Continuμm microscope
(15× objective) with a MCT-A detector cooled by liq-
uid nitrogen. e spectra were collected in transmission
mode, in 50μm areas resolution setting 4 or 8 cm−1 and
128 scans in the 4000–650 cm−1 spectral range, using a
ermo diamond anvil compression cell. For some infra-
red spectra, the system was purged with nitrogen prior
to the data acquisition; for all infrared spectra the CO2
absorption at circa 2400–2300 cm−1 was removed from
the acquired spectra. To improve result robustness, more
than one spectrum was acquired from different sample
spots.
Data analysis
Spectral pre‑treatments
Excitation and emission spectra were tested, separately
and in combination. Each spectrum was pre-processed
by normalization to unit area. Spectra of different sam-
ples were then arranged vertically forming two data
blocks (excitation and emission) forming two-way matri-
ces (samples versus wavelength). ese data blocks were
analysed in separate and in combination. In the latter
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 6 of 18
Nabaisetal. Herit Sci (2021) 9:32
situation, the two blocks were merged by horizontally
concatenating the matrices. e best results were accom-
plished with the combination of the excitation and the
emission spectra. Further details on the selection of pre-
processing methods can be found in [54].
Chemometric methods
Hierarchical cluster analysis (HCA) was applied to the
spectral data towards the discrimination and classifica-
tion of artworks using different dye families. Dendro-
grams were developed considering data from the entire
artworks’ dataset including the whole wavelength range.
e HCA method was applied not directly to the fluores-
cence excitation and emission spectra but, to the result
of a principal component analysis of these data. e first
three components encompassing slightly more than 95%
of the total variance were used by the HCA method. e
Ward’s algorithm was used to perform the clustering
approach and the Euclidean distance selected. Prior to
the application of all chemometric methods, the datasets
were mean centred. All chemometric analyses and data
manipulations were performed with Matlab Version 8.6
(R2015b) (e Mathworks, Natick, MA) and the PLS
Toolbox Version 8.2.1 (Eigenvector Research, Manson,
WA).
Results anddiscussion
Lac dye identication inmedieval manuscripts
bymolecular uorescence: probing theinuence
ofthemanufacturing processes andadditives
e main spectral features used to establish the four
groups of lac dye-based paints are shown in Table1.
Next, they will be discussed in detail and compared
with the reference paints in our database [19, 20, 27]. It
should be noted that most of the molecular fluorescence
spectra obtained in medieval manuscripts belong to
one of these two main groups, Lac 1 and Lac 3, or have
some characteristics of one or both as will be discussed
below, Fig.4.
In a previous publication, and based on medieval
reproductions of lac dye paints, we proposed that lac
dye reds can be produced as “free lac dyes” and Al3+-lac
complexes [20]. For this reason, fourteen recipes for lac
dye were selected from eight technical sources dating
from the tenth century to the end of the sixteenth cen-
tury, including four that do not use an aluminium salt4
[32]. ese reproductions will support our analysis of the
original medieval paints.
Analysis oftheemission andexcitation spectra oflac dye
paints
“Free lac” paints are not complexed with Al3+ and display
the spectral features of Lac 1 group; (i) a broad excita-
tion spectrum with a maximum at 473nm (and possibly
also a band at 430nm, in the same region as we get an
instrumental artefact), Fig.4; (ii) an emission spectrum
with a maximum at 588nm and shoulders at 563nm and
600 nm, Fig. 4. From eight original manuscripts (from
a total of seventeen), twelve paints are included in this
Table 1 A total of 70 emission and 70 excitation spectra on 37 lac dye paints were collected from 17 illuminated
manuscripts produced inPortuguese monasteries (12th–13th c.). Based ontheir spectral features and intensity these
spectra were assembled in4 groups, Lac 1 to4
The values for the emission and excitation maxima are in italics
sh shoulder
Lac 1 Lac 2 Lac 3 Lac 4
Excitation maxima 473 nm
Band at 430 nm 535–555 nm 525 nm
sh 500 nm (resin chromo-
phores)
sh 550–555 nm
554 nm
Intensity
Max 4.2 × 1054.8 × 1056.5 × 1051.0 × 106
Min 1.3 × 1052.6 × 1052.2 × 1054.6 × 105
Emission maxima 588–593 nm
sh 562 nm sh 615 nm 585–590 nm
sh 563 nm
sh 610–616 nm
585–590 nm
sh 610–616 nm 600–610 nm
592–594 nm
sh 610 nm
Intensity
Max 8.1 × 1049.2 × 1041.3 × 1059.3 × 104
Min 2.1 × 1043.4 × 1042.0 × 1043.6 × 104
4 Ms. of Ibn Bādīs (c.1025), Chapter6; Ms. O Livro de como se fazem as cores
(fifteenth century), Chapter13; Ms. Paduan (late 16th–seventeenth century),
Recipe 90 and 113 [7, 55–57].
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 7 of 18
Nabaisetal. Herit Sci (2021) 9:32
group, Lac 1,5 Additional file1: TableS1. eir spectra
can be compared to those obtained from the colours
described in Ibn Bādīs (c.1025), chapter 6; the Ms O
Livro de como se fazem as cores (15th c.), chapter13; and
manuscript Paduan (late sixteenth to seventeenth cen-
tury), recipes 90 and 113 [7, 27, 32, 55–60], Fig.5.
e excitation spectrum for Al3+-lac complexes is dif-
ferent from “free lac”, being characterized by two bands at
ca. 554–56nm and 523–25nm as well as by a shoulder at
ca. 500nm which is characteristic of the yellow chromo-
phores present in sticklac, Fig.4. Lac 3, combines both
features in its excitation spectrum, the Al3+-lac complex,
and the resin signature [20]. e emission spectra for this
group is characterized by a maximum at 588 nm, like
Lac 1 group, but is easily distinguished from the latter
because the shoulder at 563nm is absent, Fig.4. From six
original manuscripts (from a total of 17), seven paints are
included in this group Lac 3,6 Additional file1: TableS3.
Lac 3 spectra compare well with those obtained from
the colours described in Ms Mappae Clavicula (twelfth
century), recipe 253; Ms Bolognese (fifteenth century),
recipes 129, 131, 137 and 140; Strasbourg manuscript
(fifteenth century), recipe for Bright Paris Red; Montpel-
lier Ms (fifteenth century), recipe 1.9; and Jean le Begue
(1431), recipe 36, Figs.5 and 6 [20, 27, 32, 55–60]. e
spectra from the reconstructions, generally, have a better
signal-to-noise ratio, Fig.6.
In group Lac 2 (Additional file1: TableS2), the emis-
sion spectra are similar to group Lac 1, but the excitation
Fig. 4 Lac dye paints, excitation and emission spectra from: 12th c.–13th c. Portuguese manuscripts that are represented as groups Lac 1 ( ), Lac
3 ( ), and Lac 4 ( ); sticklac (raw material, resin + chromophores); Al3+-laccaid acid A complex, applied on filter paper, pH = 3.5. For more details,
please, see the text
5 Lac 1 group—SC 20, ff. 191 and 197v; Lv 13, ff. 21, 30 and 44v; ALC 412, ff.
10v and 12; ALC 421, f. 202v; ALC 446, f. 96v, Lv 12, ff. 17 and 94; and SC 21,
f.19.
6 Lac 3 group—Lv 5, ff. 6 and 73v; Lv 13, ff. 44v; Lv 15, f. 26; SC 21, f. 2; ALC
247, f. 21v, and ALC 421, f. 193v.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 8 of 18
Nabaisetal. Herit Sci (2021) 9:32
spectra are different, Fig.5. ese excitation spectra are
still characterized by a very broad band, but the maxima
fall at higher wavelengths, between 535 and 555nm (the
region in which the first band of the complex Al3+ -lac-
caic acid A emits). It resembles, to a certain degree, the
excitation spectra obtained for brazilwood lake pigments
(see Sect.2). From the original manuscripts, nine paints
are included in group Lac 2.7 In this case, we could not
find a medieval recipe for a lac dye colour with these spe-
cific spectral features.
e paints from the monumental bible Santa Cruz 1
(SC1) are possibly the most complex in the collection. In
ff. 2v, 37 and 77, the excitation spectra can be compared
with the Lac 2 group, but the emission spectra show a
well-resolved envelope, with two maxima at ca 560 and
between 580 and 586nm (closest to the Lac 1 spectra,
where the shoulder has turned into a band), Additional
file1: TableS5. We could not find a match with any of our
historical reconstructions. Different spectra, from those
described above, were acquired in two other folios: that
of folio 14v can be compared to that of Lac 1 group, with
an excitation maximum at 474nm and an emission maxi-
mum at 593nm, while that of folio 24 has the same exci-
tation maximum as the latter, but an emission maximum
shifted to 560nm with a shoulder at 580nm, similar to
the emission spectra obtained in folio 37.
Finally, the spectra of the Lac 4 group (Additional file1:
TableS4), were found only in four paints of three folios
from two manuscripts (ALC 238 and ALC 347), display-
ing a different spectral distribution, with maxima shifted
to higher wavelengths when compared to the other
groups. In Fig.5, two representative spectra are depicted,
being characterized by an excitation maximum at 554nm
and emission at ca 592–610nm. In this case, it was pos-
sible to find a very good match with a lac dye paint repro-
duced following the instructions of the recipe 309 from
De diversis coloribus, which is included in the manuscript
compiled by Jean Le Begue (1431), employing both lac
Fig. 5 Lac dye paints, representative excitation and emission spectra for each group compared with reference paints applied, using glair as binding
medium, on filter paper; for Lac 2 group it was not possible to find a match with a medieval reconstruction. Lac 1, ALC 412 f. 10v with an ibn Bādīs
reconstruction. Lac 3, ALC 247 f. 21v with a Mappae Clavicula reconstruction. Lac 4, ALC 238 f. 206v Jean le Begue, recipe 309, reconstruction
7 Lac 2 group—ALC 249, f. 109v and ALC 419, f. 98; ALC 427, f. 115v; Lv 15,
ff. 16 and 50; Lv 50, ff. 1v and 64v, and SC 20, ff. 78 and 86.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 9 of 18
Nabaisetal. Herit Sci (2021) 9:32
dye and brazilwood [32, 56]. e excitation spectrum is
what best distinguishes this recipe from all other medi-
eval lac dye paints, and the spectral envelope acquired
in the original manuscripts compares very well with the
Jean le Begue 309 recipe, both in spectral distribution
and intensity. is paint combines spectral features from
two different chromophores, which are medium to weak
emitters, so, depending on the raw materials, their quan-
tities, and processing as well as on the acquisition con-
ditions, one can dominate over the other. It should be
noted that, of the fourteen recipes selected to be repro-
duced, two included a mixture of lac with brazilwood; in
one of the recipes, the quantity of both materials is indi-
cated,8 while in the second, the recipe Jean le Begue 309,
only a vague description is given: "Take an ounce of lake,
and rasp finely a little Brazil wood" [32, 56]. e spectra
of the historical reconstructions are included in Addi-
tional file1: TableS6.
In our previous publications, of these medieval lac dye
paints, Surface-enhanced Raman spectroscopy (SERS)
identified only lac dye chromophores [20, 27], including
in ALC 238 and ALC 347. Although brasilein was not
identified by SERS in these manuscripts, it is likely that
in ALC 238 and ALC 347 the red colour was obtained
admixing lac dye and brazilwood colorants, as indicated
by molecular fluorescence. By SERS it would have been
very difficult to detect brasilein in a mixture with lac dye.
e intensity of the emission is also characteristic of
a molecular fluorescence spectrum, and for historical
reconstructions, it has been observed that it increases
with the amount of alum used in the recipe. us, under
certain conditions, the signal intensity can be corre-
lated with the presence of an Al3+-complex and the use
of alum. However, emission intensity can be strongly
affected by the morphology of the analysed surface. For
this reason, meaningful conclusions are only possible
when relevant changes are observed. Based on our expe-
rience of applying microspectrofluorimetry to the study
of works of art, for an experienced user, it is possible to
acquire spectra with a consistent S/N for the same col-
our/micro-sample, so it is possible to include the signal
intensity in the spectral characteristics. e effect of sur-
face morphology is minimized by selecting the same type
Fig. 6 Excitation and emission spectra of: shellac (processed resin without lac dye chromophores), sticklac (raw material, resin + chromophores),
and Al3+-lac complex, pH = 3.5; lac dye reconstructions from recipes in Bolognese 130, Mappae Clavicula and Strasbourg, applied, using glair as a
binding medium, on filter paper
8 Ms. Bolognese (fifteenth century), Recipe 130, To make lake as before in
another manner: Take of gum lac 5 lbs., (…) take 6oz of verzino in very fine
powder [56].
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 10 of 18
Nabaisetal. Herit Sci (2021) 9:32
of surface (usually smooth/homogeneous surfaces) and
optimizing the S/N before acquiring the signal.
In the medieval paints the differences in intensity
observed are small and, for the time being, do not allow
us to conclude if alum was not used in these medieval
colours. e greatest variations are detected in the exci-
tation spectra, with the lowest intensities found in Lac
1 and Lac 2 groups, 4.2 × 105 and 4.8 × 105, respectively;
while for Lac 3 and Lac 4 groups, they are 6.5 × 105 and
1 × 106, respectively. Interestingly, in the SC1 group, the
excitation spectra close to Lac 1 group, acquired in ff. 14v
and 24, also present the lowest intensities of this group.
Correlating thegroups oflac dye paints based onmolecular
uorescence spectra withthepaint formulation
givenbyinfrared spectroscopy
As will be detailed below, for lac dye paints, the diver-
sity observed in the paint formulations by infrared spec-
troscopy is not directly reflected in the photophysical
properties of the main chromophores that allowed us to
distinguish four groups for medieval lac dye paints. On
the other hand, the two analytical techniques offer rele-
vant complementary information. In these paints, infra-
red spectroscopy detects the presence of shellac resin,
proteinaceous binder, and fillers as gypsum and calcium
carbonate. e latter may be used as fillers or as opaci-
fiers and/or to create lighter colours, such as when add-
ing lead white (rarely applied to create lighter tones, but
extensively used to “heighten the colours” [40]). It is also
clear, from the infrared spectra, that their relative con-
centration varies. However, neither specific charges nor
relative proportions could be associated with the four
groups, created based on the analysis of the molecular
fluorescence spectra.
Representative infrared and molecular fluorescence
spectra are available as Additional file1: Tables S1 to S5.
Most of the infrared spectra of the paints in group Lac
1 show the presence of the shellac resin through its dis-
tinctive C-H stretching bands, but in some, its presence
is barely visible. All spectra are dominated by the pro-
teinaceous binder. Fillers such as calcium carbonate and
gypsum are also clearly visible, Additional file1: TableS7.
When compared with the binder, they are usually present
in low concentrations, with one exception, the paints in
folios 30 and 44v in Lv 13, as it is visible in Additional
file 1: Table S8, where the CaCO3 percentage is more
than 6 times as higher as the rest of the manuscripts.
Most excitation spectra displayed consistently a broad
band with a maximum at ca 472–474nm and shoulders
at 508–514 and at ca. 550nm. On the other hand, the
emission spectra display a higher diversity, being char-
acterized by a maximum at ca 590nm or in the interval
605–615nm, and a more or less pronounced shoulder at
563 nm. However, no meaningful correlation between
the fillers and the fluorescence emission spectra could be
detected.
e infrared spectra in group Lac 2 are characterized
by a lower amount of shellac resin and a higher propor-
tion of proteinaceous binder and fillers when compared
with group Lac 1; ALC 249 and 419 (ff. 109v and 98,
respectively) display a high proportion of calcium car-
bonate and gypsum; Lv 15 and Lv 50 of calcium carbon-
ate and lead white. For the latter, it is possible that the
difference in the excitation maxima may be due to the
presence of lead white in a relatively high amount [27].
In the infrared spectra of the paints in group Lac 3,
it is possible to assess the presence of the shellac resin
through its distinctive C–H stretching bands, Additional
file1: TableS7. Lv 5, f. 6 displays a spectrum dominated
by the shellac resin whereas in the other spectra the pro-
teinaceous binder is clearly visible. Gypsum was detected
in three paints (Lv 5, f. 6; SC 21 and ALC 247); whereas,
in Lv 13, ALC 421 and Lv 15, calcium carbonate was
found, the latter with lead white. Both gypsum and cal-
cium carbonate were employed as fillers (improving the
paint mechanical performance) and, depending on the
concentration, as opacifiers. So, although the paint for-
mulations differ, the recipe used to produce the lac dye
chromophore was probably similar. In this case, micro-
spectrofluorimetry could group the recipes according to
the chromophore and infrared spectroscopy was able to
discriminate the paints’ formulation (SC 21 and ALC 247
are very similar).
In SC1 the infrared spectra are dominated by the pro-
tein fingerprint and the shellac resin is not visible, except
for folio 14v; a small amount of calcium carbonate was
also detected. ere is no information in these spectra
that explains the differences observed in the molecular
fluorescence spectra.
e two paints in group Lac 4 were found in manu-
scripts from the collection of the Monastery of Alcobaça,
ALC 238 (Book of birds) and ALC 347 (Sermones de ver-
bis Domini); in ALC 238 paint, the shellac resin and the
proteinaceous binding medium are clearly visible (C–H
and amide stretching, respectively), and a high amount
of gypsum as a filler (when compared with the binder) is
also observed. On the other hand, the infrared spectrum
of the ALC 347 paint is dominated by the protein finger-
print and a high amount of calcium carbonate. So, again,
the diversity found in the paint formulations, detected by
the infrared spectra, did not leave a visible mark in the
chromophore emission. In this case, as already discussed,
the excitation spectrum points to the presence of a bra-
zilwood lake pigment (please, see next section and Fig.6).
Considering that the presence of shellac is very clear in
the infrared spectra, we propose this paint to be included
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 11 of 18
Nabaisetal. Herit Sci (2021) 9:32
in the recipes that ask for a mixture of lac and brazil-
wood, in the proportion used to reproduce the Jean le
Begue 309 recipe, Fig.6.
e presence of a brazilwood lake pigment is not visible
in the infrared spectrum, although the high gypsum con-
centration is indicative of brazilwood recipes [24]. e
use of this mixture of lac dye and brazilwood is an inter-
esting discovery concerning the production and experi-
mentation in the Alcobaça scriptorium. It provides a link
from an earlier period in which lac dye paints were used
to produce dark reds, pinks and carmine hues to its sub-
stitution by brazilwood lake pigments as observed in the
winter Breviary ALC 54 dated from the fourteenth cen-
tury [46]. A missing link, but which does not explain why
a glossy lac dye paint, was substituted by a matte colour:
the same hues are obtained with the two chromophores,
but not the same brightness. is can open new perspec-
tives for the accurate dating of medieval manuscripts.
The chemometrics approach forlac dye paints
e molecular fluorescence spectra (excitation and emis-
sion) were paired side-by-side before analysis. Before
alignment, each spectrum was processed using the nor-
malization by unit area. Paired pre-processed spectra
encompassing all samples for analysis were mean-cen-
tered and subjected to PCA. PCA results show that the
first three components encompassed over 95% of the
total original data variance, thus a sound basis for the
variability observed in the spectra. ese three compo-
nents were stored and used for further analysis with the
HCA method (Ward’s algorithm and Euclidean distance),
see Additional file1: Figure S2. e results of the cluster
analysis using both the excitation and emission spectra
are plotted in Fig.7. is figure offers a general view of
the separation of the groups for lac dye (Lac) and brazil-
wood (BW). For a close-up of the groups of each colorant
please see Figs. 8 and 11. Except for the SC1 set, the
grouping followed closely our proposal based on the dis-
cussion of the fluorescence emission and excitation spec-
tra. Although most of the folia (and colours) are within
the expected groups, the clustering method was unable
to differentiate group Lac 2 and Lac 3, since they share
common spectral features already discussed, and hence
are represented together in Fig.7.
Lac 1 group is represented at the extremity of the
cluster, being considered the most different among the
groups, which is in accordance with our previous analy-
sis, Table1. Lac 4 group is the nearest to the brazilwood
cluster, which agrees with the hypothesis of a possible
mixture of the two chromophores. Because of the use of
both excitation and emission spectra, the obtained den-
drogram correctly positions this group within the lac
dye cluster. Such a result would not have been possible
using only the excitation spectra, due to similarity with
brazilwood signals. Moreover, groups Lac 2 and Lac 3,
although being in a separate cluster, are close to group
Lac 4, while SC1 is a separate cluster close to groups Lac
1 and Lac 2 & 3; SC1 is named Lac 5 group in Fig.7. is
may be due to the excitation spectra of folios 2v, 37, and
77 resembling group Lac 2, with an emission signal like
what was found for group Lac 1, as discussed previously.
Some misplacements can be observed in Fig.8 when
compared with our previous grouping based on molecu-
lar fluorescence spectra, Additional file1: Tables S1–S5.
In group Lac 1, the following spectra are misplaced: SC
21, f. 2v, and ALC 421, f. 193. e first might be due to
a higher intensity of the band at ca. 483nm, when com-
pared to the rest of group Lac 3, where the sample sup-
posedly belong. e placement of ALC 421 in group
Lac 1 instead of group Lac 3, may be explained by poor
resolved spectral features, with a higher predominance of
the band at 475nm.
In group SC1, named group Lac 5 in Fig.7, the follow-
ing spectra are misplaced: Lv 50, f. 64v; Lv 15, f. 26; Lv
13, f. 21; Lv 12, f. 94; and ALC 412, ff. 10v and 12. Both
the folios misplaced of Lv 15, Lv 50, Lv 12, and ALC 412
show a shoulder ca 560nm in the emission spectra, simi-
lar to what was found in SC1, which caused the incorrect
positioning within the cluster. is was expected as this is
the manuscript displaying the more complex and diverse
features.
In groups Lac 2 & 3, the only misplaced paint is that of
folio 2v of SC 1. It is placed on the cluster Lac 2, indicat-
ing some similarity between excitation spectra.
Finally, in group Lac 4 the misplaced spectra are: ALC
427, f. 115v, and Lv 15, f. 50. e predominance of the
band at ca. 550nm, with a well-resolved emission spectra
with bands at 586 and 614nm have placed them next to
folio 3, lac of ALC 347.
Fig. 7 Dendrogram generated by HCA applied to excitation and
emission spectra of all the historical samples for lac dye (Lac) and
brazilwood (BW) based paints
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 12 of 18
Nabaisetal. Herit Sci (2021) 9:32
Brazilwood inmedieval manuscripts
e differences observed in the molecular fluorescence
spectra of brazilwood medieval paints are much smaller
than those found for lac dye paints, so compared to them,
brazilwood colours can be considered a single group,
Fig. 9. Likewise, for the four recipes of brazilwood in
the O Livro de como se fazem as cores, Fig.10. However,
based on our previous experience applying chemometrics
to historical reconstructions [19], we think that it is pos-
sible to extract more information from the brazilwood
cluster. Besides the position and intensity of the maxima
in a spectrum, band broadening and other less visible
features will be processed in the chemometrics analysis.
In fact, a spectrum of molecular fluorescence is very rich
in information, and with the support of chemometrics,
the way is open for an in-depth analysis of the complexity
of medieval painting.
Contrarily to what was observed for lac dye paints, for
brazilwood pigment lakes, it is the spectral distribution
of the emission spectra that most differentiates them. It
is important to note that the differences are in the range
of a few nanometres, being the intensities observed very
similar and in the range to what observed for lac dye
paints, Table2. Based on the emission and excitation
Fig. 8 Close-up for the lac dye cluster of the dendrogram generated by HCA applied to excitation and emission spectra of all the historical samples
of lac dye-based paints from all three Portuguese Monasteries: Alcobaça (black), Lorvão (grey), and Santa Cruz (italic): Lac 1 (green), Lac 2 & 3 (pink),
Lac 4 (light blue) and Lac 5 (yellow)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 13 of 18
Nabaisetal. Herit Sci (2021) 9:32
Fig. 9 Brazilwood paints, excitation and emission spectra representative for BW 1 group, Ajuda Songbook f. 59); BW 2 group, IL 42 f. 113; BW 3
group, ms 24 f. 60; and brazilein 5 × 10–5 M in MeOH:H2O (70:30, v/v) with Al3+ (× 1000) at pH 3.2, applied on filter paper
Fig. 10 Excitation and emission spectra of brazilwood lake pigment reconstructions from the Livro de como se fazem as cores: recipe 8, 9, 27, and 44
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 14 of 18
Nabaisetal. Herit Sci (2021) 9:32
spectral distribution, we propose three groups that will
be next discussed.
Analysis oftheemission andexcitation spectra ofbrazilwood
paints
In group BW 19 are included all the spectra collected
from Ajuda Songbook, which are characterized by exci-
tation maxima within 552–556 nm and emission max-
ima at ca. 585nm, Fig.9, Table2 and Additional file1:
TableS9. ese spectral features compare well with those
obtained from the four recipes described in the O Livro
de como se fazem as cores, in particular with recipe 8, as
well as the Al3+–brazilein reference sample, Figs.9 and
10 and Additional file1: TableS13.
e spectra frombrazilwood paints applied in the
Books of hours of French or Flemish production may be
assembled in two groups. In group BW 2,10 the excitation
maxima are shifted to slightly higher wavelengths, 555–
562nm, and the band is not so broad when compared
with group BW 1, Additional file1: TableS10. Emission
maxima are also shifted to higher wavelengths, being
found in the interval 594–602nm. e excitation spec-
tra that represent group BW 311 are found between 553
and 560nm, falling in the interval of group BW 2, being
differentiated by their emission maxima that is shifted
to slightly lower wavelengths, 589–600 nm, Additional
file1: TableS11. Usually, also a lower S/N was obtained
for BW 3 signals. As already pointed out, the differences
between these two groups are smaller than those found
for lac dye. However, they can reveal unique informa-
tion that can be linked to a different original recipe or the
aging of the paint. We will further discuss this issue when
analysing the infrared data.
Finally, the paints from the winter breviary (ALC 54, f.
92) may be compared to the French book of hours ms 22,
f. 76v that is included in group BW 3, Additional file1:
Tables S11 and S12.
Correlating thegroups ofbrazilwood paints based
onmolecular uorescence spectra withthepaint formulation
givenbyinfrared spectroscopy
For medieval brazilwood paints, it is possible to correlate
the differences observed in the molecular fluorescence
spectra with the paint formulation. We will start by ana-
lysing the pigment reconstructions to support our dis-
cussion of the medieval paints. For both, reconstructions
and original paints, infrared analysis showed extenders
and binders and, in certain cases, the pigments’ alumi-
nate substrate, Table2 and Additional file1: TableS13.
e infrared spectrum of recipe 8 is mainly characterized
by the presence of gypsum, while both recipes 9 and 27
present a mixture of gypsum and calcium carbonate. e
infrared spectrum of recipe 44 however, is characteristic
of the pigments’ substrate, showing the aluminate com-
pound [24], Additional file1: TableS13. All these data are
relevant as we show that added extenders can change the
colour of the paint [24]. Pigments may appear redder or
more violet, darker, or lighter depending on the complex-
ing metal ion, Al3+, Ca2+, or Pb2+, with the latter two dis-
playing a bluer shade.
Group BW 1 is characterized by the presence of both
calcium carbonate and lead white. is is similar to what
was found in recipe 8 of the Livro de como se fazem as
cores, where the extraction is done in the presence of
alum and white lead. Calcium carbonate may be found
in the pigment’s substrate if the filtration is done over
a chalk bow [24]. However, as indicated in the recipe, a
gypsum bowl can also be used, in which case, no calcium
carbonate would be present, Additional file1: TableS13.
is is the only group to which lead white is added and,
considering that Pb ions can quench the fluorescence
emission [18], this would explain the low signal-to-noise
ratio found in fluorimetric data.
Table 2 A total of 80 emission and 80 excitation spectra
on18 brazilwood paints were collected from8 illuminated
manuscripts produced during the 13th—15th c. Based
on their spectral features these spectra were assembled
in3 groups
The values for the emission and excitation maxima are in italics
Group BW 1 Group BW 2 Group BW 3
Excitation maxima 552–561 nm 555–562 nm 553–564 nm
Intensity
Max 2.2 × 1053.2 × 1051.7 × 105
Min 9.1 × 1041.0 × 1059.9 × 104
Emission maxima 582–586 nm 594–602 nm 590–600 nm
Intensity
Max 9.5 × 1048.6 × 1046.2 × 104
Min 3.5 × 1043.0 × 1042.4 × 104
Infrared CaCO3
Lead White CaCO3
CaSO4·2H2OCaCO3
» CaSO4·2H2O
Reconstruction LKFK 8 LKFK 9
9 BW 1 group – Ajuda Songbook, ff.4, 17, 21 and 59; IL 15, f. 60.
10 BW 2 group –IL 15, f. 15; IL 19, f. 91, IL 21, f. 88, IL 42, ff. 9, 85 and 113.
11 BW 3 group – ms 22, f. 76v; ms 24, f. 60; IL 15, f. 66; IL 19, f. 21; IL 42, ff.
23 and 133.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 15 of 18
Nabaisetal. Herit Sci (2021) 9:32
Groups BW 2 and BW 3 are characterized by the pres-
ence of calcium carbonate and gypsum. e main dif-
ference is the higher amount of gypsum found in the
latter, which may be responsible for the lower intensities
observed with our acquisition set-up (less light will be
absorbed by the chromophore, which will lead to a lower
fluorescence emission), Fig. 9. e infrared spectra of
group BW 3 are comparable with recipe 9 of the Livro de
como se fazem as cores, in which is noticeable the pres-
ence of both gypsum and calcium carbonate.
Again, the winter Breviary, ALC 54, does not fit in any
of the proposed groups. e infrared spectrum does not
show the presence of lead white, calcium carbonate, or
gypsum.
Fig. 11 Close-up for the brazilwood cluster of the dendrogram generated by HCA applied to excitation and emission spectra of all the historical
samples of brazilwood based paints: group BW1 (green), BW 2 (blue) and BW 3 (red)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 16 of 18
Nabaisetal. Herit Sci (2021) 9:32
The chemometrics approach forbrazilwood paints
e results of the hierarchical cluster analysis, using both
the excitation and the emission spectra, are plotted in
Fig.11. e grouping followed closely our proposal based
on the discussion for the fluorescence emission and exci-
tation spectra, combined with infrared spectroscopy. e
presence of specific extenders, proved by infrared spec-
tra, influences the molecular fluorescence spectra and
is detected by the statistical approach applied, Table2.
e exception is the paint in IL 19 f. 21; due to the pres-
ence of lead white it could be included in BW1, however,
its emission maximum is shifted towards longer wave-
lengths, closer to BW2 & BW3.
ALC 54 was placed within the BW 3 cluster, essen-
tially due to the emission spectrum shifted to longer
wavelengths. However, ALC 54 presents a lower signal-
to-noise ratio when compared with the other samples
in group BW 2. Nevertheless, the clustering method
identifies some differences, placing it on the limit of the
dendrogram. Another challenging case is the book of
hours IL 15, the only one of Flemish production. e
three folios analysed are spread through all three groups.
is can indicate that three different illuminators were
at work, an aspect that we would like to explore in the
future.
Conclusions
Molecular fluorescence data embodies the rich com-
plexity that characterizes paints prepared with organic
chromophores. Based on our knowledge of the recon-
struction of medieval paints, we anticipated that lac dye
formulations would be more complex than brazilwood
lake pigments. On the one hand, from the source mate-
rial several chromophores are extracted, although we
have always found laccaic acid A as a main compound in
our paints, Fig.1 [32]. On the other, in lac dye paints we
find variable amounts of shellac resin, a complex mate-
rial that also emits in the visible, partially overlapping
the molecular fluorescence spectra of laccaic acid A, the
main chromophore. Possibly, due to this greater com-
plexity, it has not yet been possible to fully correlate the
influence of fillers and other additives, in the formulation
of the paint, with its spectral signature.
e complete evaluation of the paints’ composition,
in particular, the data obtained in the semi-quantitative
analyses based on the infrared spectra, allowed an in-
depth discussion of the molecular fluorescence spectra.
In general, the grouping based on the molecular fluo-
rescence signature of the chromophores was correctly
obtained from the chemometric methods, with few mis-
placed samples, Fig.7. ese deviations are explained by
a low signal-to-noise ratio or because the sample was dif-
ferent from all the others.
Notably, in the spectral signature of brazilwood paints,
it was possible to probe the presence of fillers and to dif-
ferentiate between paints’ formulation. Future work will
systematically study these effects to provide a general
rationale. It is important to note that molecular fluores-
cence pinpointed, for the first time, a colour in which
both lac and brazilwood chromophores are present (Lac
4 group), in manuscripts from the Alcobaça scripto-
rium (De avibus and Sermones de verbis Domini, ALC
238 and 347, respectively). Being this one of the "differ-
ent paints". e presence of the two chromophores could
not be detected either in the infrared or Raman spectra.
Remarkably, the use of these two chromophores to pro-
duce a medieval paint had been described in the text by
Jean Le Begue and, indeed, these paints compare very
well with Le Begue’s recipe 309. is allows us to hypoth-
esize that, with further developments, molecular fluo-
rescence could be used as a tool to provide geographic
information on the place of the manufacture of an illumi-
nated manuscript as well as on its dating.
Supplementary Information
The online version contains supplementary material available at https ://doi.
org/10.1186/s4049 4-021-00490 -8.
Additional le1: Figure S1. Typologies of the manuscripts found in Por-
tuguese collections. TableS1–S5. Microspectrofluorimetry and infrared
data of samples from lac dye in medieval manuscript illuminations (12th–
13th c.). TableS6. Microspectrofluorimetry and infrared data of lac dye
reconstructions. TableS7. Infrared spectra. TableS8. Percentage of cal-
cium carbonate. TableS9–S12. Microspectrofluorimetry and infrared data
of samples from brazilwood in medieval manuscript illuminations (end
of 13th–15th c.). TableS13. Microspectrofluorimetry and infrared data of
brazilwood reconstructions. Figure S2. Scatter plot representing the first
three scores from the PCA model used to produce the dendrogram for lac
dye (Lac) and brazilwood (BW) based paints. Link to video in: https ://www.
dropb ox.com/s/4vfyc 63cjs vn7j1 /Herit age%20Sci %20Nab ais%20P_%20Sca
tter%20plo t.mp4?dl=0.
Abbreviations
SC: Santa Cruz Monastery/Mosteiro de Santa Cruz; Lv: Lorvão Monastery/
Mosteiro do Lorvão; ALC: Alcobaça Monastery/Mosteiro de Alcobaça; SERS:
Surface-enhanced Raman spectroscopy.
Acknowledgements
The authors would like to thank the staff and directory board of the Arquivo
Nacional da Torre do Tombo (ANTT), Biblioteca Nacional de Portugal (BNP),
Biblioteca Pública Municipal do Porto (BPMP), Biblioteca do Palácio Nacional
da Ajuda and Palácio Nacional de Mafra (PNM) for their generous support and
collaboration. We gratefully thank T. Vitorino for her research in brazilwood
paints and historical reconstructions.
Authors’ contributions
PN contributed with the acquisition of emission and excitation spectra;
conception of the models and treatments applied to the spectral data, and
with the writing and revision of the version to be published. MJM contributed
with the conception and design of the research work; acquisition, analysis and
interpretation of data; writing and revision of the version to be published. JAL
contributed with the conception of the data treatment and calculations, as
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 17 of 18
Nabaisetal. Herit Sci (2021) 9:32
well as the revision of the version to be published. MV contributed with the
acquisition and spectral interpretation of infrared spectra of brazilwood recon-
structions and the revision of the version to be published. RC contributed
with the reconstructions of lac dye and as well as with the acquisition of the
data related to lac dye medieval paints. AR contributed with the conception
of the work, data interpretation and revision of the version to be published. All
authors read and approved the final manuscript.
Funding
This research was funded by the Portuguese Science Foundation [Fundação
para a Ciência e Tecnologia, Ministério da Educação e da Ciência (FCT/
MCTES)], through doctoral programme CORES-PD/00253/2012, and PhD
grants awarded to Rita Castro [SFRH/BD/76789/2011], Paula Nabais [PD/
BD/105895/2014] and Márcia Vieira [SFRH/BD/148729/2019]; Project STEMMA
(“From singing to writing—survey on material production and routes of
Galician-Portuguese Lyric”, PTDC/LLT-EGL/30984/2017); Associate Labora-
tory for Green Chemistry- LAQV financed by national funds from FCT/MCTES
(UIDB/50006/2020 and UIDP/50006/2020).
Availability of data and materials
Most of the data on which the conclusions of the manuscript rely is published
in this paper, and the full data is available for consultation on request.
Competing interests
The authors declare that they have no competing interests.
Author details
1 LAQV-REQUIMTE and DCR, Faculty of Sciences and Technology, NOVA
University of Lisbon, 2829-516 Caparica, Portugal. 2 Faculty of Pharmacy, iMed.
ULisboa-Research Institute for Medicines, University of Lisbon, Av. Prof. Gama
Pinto, 1649-003 Lisbon, Portugal. 3 SMAArt Centre and Department of Chem-
istry Biology and Biotechnology, University of Perugia, via Elce di Sotto 8,
06123 Perugia, Italy.
Received: 8 September 2020 Accepted: 25 January 2021
References
1. Miliani C, Monico L, Fantacci S, Romani A, Melo MJ, Angelin EM, Janssens
K. Photochemistry of artists’ dyes and pigments: towards better under-
standing and prevention of colour change in works of art. Angew Chem
Int Ed. 2018;57:7324–34. https ://doi.org/10.1002/anie.20180 2801.
2. Seixas de Melo J, Moura AP, Melo MJ. Photophysical and Spectroscopic
Studies of Indigo Derivatives in Their Keto and Leuco Forms. J Phys Chem
A. 2004;108:6975–81. https ://doi.org/10.1021/jp049 076y.
3. Sousa MM, Miguel C, Rodrigues I, Parola AJ, Pina F, Seixas de Melo JS,
Melo MJ. A photochemical study on the blue dye indigo: from solution to
ancient Andean textiles. Photochem Photobiol Sci. 2008;7:1353–9. https
://doi.org/10.1039/b8095 78g.
4. Rondão R, Seixas de Melo J, Melo MJ, Vitorino T, Parola AJ. Brazilwood
reds: the (Photo)chemistry of Brazilin and Brazilein. J Phys Chem A.
2013;117:10650–60.
5. Melo MJ, Claro A. Bright light: microspectrofluorimetry for the char-
acterization of lake pigments and dyes in works of art. Acc Chem Res.
2010;43(6):857–66. https ://doi.org/10.1021/ar900 1894.
6. Nabais P, Oliveira J, Pina F, Teixeira N, de Freitas V, Brás NF, Clemente A,
Rangel M, Silva AMS, Melo MJ. A 1000-year-old mystery solved: Unlocking
the molecular structure for the medieval blue from Chrozophora tinctoria,
also known as folium. Sci Adv. 2020. https ://doi.org/10.1126/sciad v.aaz77
72.
7. Melo MJ, Castro R, Nabais P, Vitorino T. The book on how to make all the
colour paints for illuminating books: unravelling a Portuguese Hebrew
illuminators’ manual. Herit Sci. 2018;6:44. https ://doi.org/10.1186/s4049
4-018-0208-z.
8. Derrick M, Newman R, Wright J. Characterization of yellow and red
natural organic colorants on Japanese woodblock prints by EEM fluo-
rescence spectroscopy. J Am Inst Conserv. 2017;56:171–93. https ://doi.
org/10.1080/01971 360.2016.12754 38.
9. Mounier A, Le Bourdon G, Aupetit C, Lazare S, Biron C, Pérez-Arantegui
J, Almazán D, Aramendia J, Prieto-Taboada N, Fdez-Ortiz de Vallejuelo S,
Daniel F. Red and blue colours on 18th–19th century Japanese wood-
block prints: in situ analyses by spectrofluorimetry and complementary
non-invasive spectroscopic methods. Microchem J. 2018;140:129–41.
https ://doi.org/10.1016/j.micro c.2018.04.023.
10. Picollo M, Aceto M, Vitorino T. UV-Vis spectroscopy. In: Giamberini M,
Jastrzab R, Liou JJ, Luque R, Nawab Y, Saha B, Tylkowski B, Xu CP, Cerruti
P, Ambrogi V, Marturano V, Gulaczyk I, editors. Physical sciences reviews.
Berlin: de Gruyter GmbH; 2018. https ://doi.org/10.1515/psr-2018-0008.
11. Brunetti B, Miliani C, Rosi F, Doherty B, Monico L, Romani A, Sgamellotti A.
Non-invasive investigations of paintings by portable instrumentation: the
MOLAB experience. Top Curr Chem. 2016. https ://doi.org/10.1007/s4106
1-015-0008-9.
12. Thoury M, Echard JP, Réfrégiers M, Berrie B, Nevin A, Jamme F, Bertrand L.
Sinchrotron UV-visible multispectral luminescence microimaging of his-
torical samples. Anal Chem. 2011;83(5):1737–45. https ://doi.org/10.1021/
ac102 986h.
13. Nevin A, Echard JP, Thoury M, Comelli D, Valentini G, Cubeddu R. Excita-
tion emission and time-resolved fluorescence spectroscopy of selected
varnishes used in historical musical instruments. Talanta. 2009;80(1):286–
93. https ://doi.org/10.1016/j.talan ta.2009.06.063.
14. Nevin A, Comelli D, Valentini G, Cubeddu R. Total synchronous fluores-
cence spectroscopy combined with multivariate analysis: method for
the classification of selected resins, oils, and protein-based media used in
paintings. Anal Chem. 2009;81(5):1784–91. https ://doi.org/10.1021/ac801
9152.
15. Comelli D, D’Andrea C, Valentini G, Cubeddu R, Colombo C, Toniolo L.
Fluorescence lifetime imaging and spectroscopy as tools for nondestruc-
tive analysis of works of art. Appl Opt. 2004;43(10):2175–83. https ://doi.
org/10.1364/ao.43.00217 5.
16. Delaney JK, Zeibel JG, Thoury M, Littleton R, Palmer M, Morales KM,
René de la Rie E, Hoenigswald A. Visible and infrared imaging spectros-
copy of Picasso’s Harlequin Musician: mapping and identification of
artist materials in situ. App Spectrosc. 2010;64(4):584–94. https ://doi.
org/10.1366/00037 02107 91414 443.
17. Claro A, Melo MJ, Seixas de Melo JS, van den Berg KJ, Burnstock A,
Montague M, Newman R. Identification of red colourants in van Gogh
paintings and ancient Andean textiles by Microspectrofluorimetry. J Cult
Herit. 2010;11:27–34. https ://doi.org/10.1016/j.culhe r.2009.03.006.
18. Claro A, Melo MJ, Schäfer S, Seixas de Melo JS, Pina F, van den Berg KJ,
Burnstock A. The use of microspectrofluorimetry for the characterization
of lake pigments. Talanta. 2008;74(4):922–9. https ://doi.org/10.1016/j.
talan ta.2007.07.036.
19. Nabais P, Melo MJ, Lopes JA, Vitorino T, Neves A, Castro R. Microspectro-
fluorimetry and chemometrics for the identification of medieval lake pig-
ments. Herit Sci. 2018;6:13. https ://doi.org/10.1186/s4049 4-018-0178-1.
20. Castro R, Pozzi F, Leona M, Melo MJ. Combining SERS and microspectro-
fluorimetry with historically accurate reconstructions for the characteri-
zation of lac dye paints in medieval manuscript illuminations. J Raman
Spectrosc. 2014;45:1172–9. https ://doi.org/10.1002/jrs.4608.
21. Valeur B, Berberan-Santos MN. Molecular fluorescence: principles and
applications. 2nd ed. Weinheim: Wiley; 2012. (ISBN: 978‑3‑527‑32837‑6).
22. Melo MJ, Ferreira JL, Parola AJ, Seixas de Melo JS. Photochemistry for
Cultural Heritage. In: Bergamini G, Silvi S, editors. Applied photochem-
istry. Berlin: Springer International Publishing; 2016. p. 499–530 (ISBN:
978‑3‑319‑31669‑7).
23. Douglas P, Burrows H, Evas R. Foundations of photochemistry: a back-
ground on the interaction between light and molecules. In: Evans RC,
Douglas P, Burrows H, editors. Applied photochemistry. Berlin: Springer
Science; 2013. p. 1–88 (ISBN: 978‑90‑481‑3829‑6).
24. Vitorino T, Melo MJ, Carlyle L, Otero V. New insights into brazilwood
lake pigments manufacture through the use of historically accu-
rate reconstructions. Stud Conserv. 2016;61(5):255–73. https ://doi.
org/10.1179/20470 58415 Y.00000 00006 .
25. Cardon D. Natural dyes: sources, tradition, technology and science. Lon-
don: Archetype Publications; 2007.
26. Melo MJ. History of natural dyes in the Ancient Mediterranean world. In:
Buchtold T, Mussak R, editors. Handbook of natural colorants. Chichester:
John Wiley & Sons; 2009. p. 6–7 (ISBN: 978‑04‑705‑1199‑2).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 18 of 18
Nabaisetal. Herit Sci (2021) 9:32
27. Castro R. The book of birds in Portuguese scriptoria: preservation and
access. Doctoral dissertation. Lisbon: Universidade Nova de Lisboa, Fac-
uldade de Ciências e Tecnologia; 2016. https ://run.unl.pt/handl e/10362
/21481 .
28. Donkin RA. The insect dyes of Western and West-Central Asia. Anthropos.
1977;72:847–80.
29. Bose PK, Sank aranarayanan Y, Sen Gupta SC. Chemistry of lac. Namkum:
Indian Lac Research Institute; 1963.
30. Kirby J. Some aspects of medieval and renaissance lake pigment technol-
ogy. In: Kirby Jo, editor. Dyes in history and archaeology 21. Proceedings
of the 21st meeting of dyes in history and archaeology. 2002; Avignon
and Lauris, France. London: Archetype Publications; 2008. p. 89–108.
31. Tamburini D, Dyer J, Bonaduce I. The characterisation of shellac resin by
flow injection and liquid chromatography coupled with electrospray
ionisation and mass spectrometry. Sci Rep. 2017;7:1–15.
32. Castro R, Miranda A, Melo MJ. Interpreting lac dye in medieval written
sources: new knowledge from the reconstruction of recipes relating to
illuminations in Portuguese manuscripts. In: Eyb-Green S, Townsend JH,
Atkinson JK, Kroustallis S, Pilz K, van Leeuwen I, editors. Sources on art
technology: back to basics. London: Archetype publications; 2016. p.
88–99 (ISBN: 978‑1‑909492‑33‑2).
33. Melo MJ, Otero V, Vitorino T, Araújo R, Muralha VSF, Lemos A, Picollo M.
A spectroscopic study of Brazilwood paints in medieval books of hours.
App Spectrosc. 2014;68(4):434–44. https ://doi.org/10.1366/13-07253 .
34. Melo MJ, Miranda MA, Castro R, Lopes J, Sarraguça J. Between tradition
and innovation: the art of colour in the Lorvão Beatus. In: Ricciardi P,
Panayotova S, editors. Manuscripts in the making: art and science, vol.
II. London and Turnhout: Harvey Miller/Brepols; 2018. p. 132–43 (ISBN:
978‑1‑909400‑10‑8).
35. Melo MJ, Nabais P, Guimarães M, Araújo R, Castro R, Oliveira MC, Whit-
worth I. Organic dyes in illuminated manuscripts: a unique cultural and
historic record. Phil Trans R Soc A. 2016;374(20160050):1–20. https ://doi.
org/10.1098/rsta.2016.0050.
36. Roger P, Villela-Petit I, Vandroy S. Les laques de brésil dans l’enluminure
médiévale: reconstitution à partir de recettes anciennes. Stud Conserv.
2003;48(3):155–70.
37. Kirby J. Paints, pigments, dyes. In: Glick T, Livesey SJ, Wallis F, edi-
tors. Medieval science, technology, and medicine: an encyclopedia.
New York: Routledge, Taylor & Francis Group; 2005. p. 379–83 (ISBN:
978‑0‑415‑96930‑7).
38. Claro A. An interdisciplinary approach to the study of colour in Portu-
guese manuscript illuminations. Doctoral dissertation. Lisbon: Universi-
dade Nova de Lisboa, Faculdade de Ciências e Tecnologia; 2009. https ://
run.unl.pt/handl e/10362 /11506 .
39. Miguélez Cavero A, Melo MJ, Miranda MA, Castro R, Casanova C. Beatus
manuscripts under the microscope: the Alcobaça Beatus and the Iberian
Cistercian tradition revisited. J Mediev Iber Stud. 2016;8(2):217–51. https
://doi.org/10.1080/17546 559.2016.12211 16.
40. Melo MJ, Castro R, Miranda A. Colour in medieval portuguese manu-
scripts: between beauty and meaning. In: Sgamellotti A, Brunetti BG,
Miliani C, editors. Science and art: the painted surface. London: Royal
Society of Chemistry; 2014. p. 170–92 (ISBN: 978‑1‑84973‑818‑7).
41. Miguel C, Claro A, Gonçalves AP, Muralha VS, Melo MJ. A study on red
lead degradation in a medieval manuscript Lorvão Apocalypse (1189). J
Raman Spectrosc. 2009;40(12):1966–73. https ://doi.org/10.1002/jrs.2350.
42. Muralha VS, Miguel C, Melo MJ. Micro-Raman study of Medieval Cister-
cian 12–13th century manuscripts: Santa Maria de Alcobaça Portugal. J
Raman Spectrosc. 2012;43(11):1737–46. https ://doi.org/10.1002/jrs.4065.
43. Melo MJ, Araújo R, Castro R, Casanova C. Colour degradation in medieval
manuscripts. Microchem J. 2016;124:837–44. https ://doi.org/10.1016/j.
micro c.2015.10.014.
44. Miguel C. Le Vert et le rouge: A study on the materials, techniques and
meaning of the green and red colours in medieval Portuguese illumina-
tions. Doctoral dissertation. Lisbon: Universidade Nova de Lisboa, Fac-
uldade de Ciências e Tecnologia; 2012. https ://run.unl.pt/handl e/10362
/9304.
45. Nabais P, Castro R, Lopes GV, de Sousa LC, Melo MJ. Singing with light: an
interdisciplinary study on the medieval Ajuda Songbook. J Mediev Iber
Stud. 2016;8(2):283–312. https ://doi.org/10.1080/17546 559.2016.12340
61.
46. Barreira CF, Melo MJ, Araújo R, Casanova C. Through the eyes of science
and art: a fourteenth-century winter breviary from Alcobaça scriptorium.
J Mediev Iber Stud. 2016;8(2):252–82. https ://doi.org/10.1080/17546
559.2016.12211 19.
47. Lemos A, Araújo AR, Casanova C, Melo MJ, Muralha VSF. Regards croisés
des historiens de l’art et des chimistes sur deux livres d’heures de la
Bibliothèque Nationale Du Portugal, Les Mss. IL15 et IL19. In: Miguélez
Cavero A, Miranda MA, editors. Portuguese Studies on Medieval Illumi-
nated Manuscripts. Turnhout: Brepols Publishers; 2015. p. 145–68 (ISBN:
978‑2‑503‑55473‑0).
48. Araújo R. Os Livros de Horas (séc. XV ) na colecção do Palácio Nacional de
Mafra: estudo e caracterização. Master dissertation. Lisbon: Universidade
Nova de Lisboa; 2012. https ://run.unl.pt/handl e/10362 /60955
49. Melo MJ, Araújo AR, Casanova C, Muralha V, Lemos A. Segredos desc-
obertas nos Livros de Horas: história, materiais e técnicas. In: Custódio D,
Miranda MA, editors. Livros de Horas: o imaginário da devoção privada.
Lisboa: Instituto de Estudos Medievais/Biblioteca Nacional de Portugal;
2015. p. 193–226.
50. Melo MJ, Araújo AR, Muralha V, Lemos A. O que nos dizem os materiais
da cor sobre os livros de horas do Palácio Nacional de Mafra. In: Lemos
A, editor. Catálogo dos Livros de Horas Iluminados do Palácio Nacional
de Mafra. Lisboa: Instituto de Estudos Medievais da FCSH-UNL/Palácio
Nacional de Mafra; 2012.
51. Vitorino T. A Closer Look at Brazilwood and Its Lake Pigments. Master dis-
sertation. Lisbon: Universidade Nova de Lisboa; 2012. https ://run.unl.pt/
handl e/10362 /10179 ?local e=en.
52. Miguel C, Lopes JA, Clarke M, Melo MJ. Combining infrared spectroscopy
with chemometric analysis for the characterization of proteinaceous
binders in medieval paints. Chemometr Intell Lab. 2012. https ://doi.
org/10.1016/j.chemo lab.2012.09.003.
53. Mannino MR, Orecchio S, Gennaro G. Microanalytical method for study-
ing paintings by use of fluorescence spectroscopy combined with
principal component analysis. Microchem J. 2013;110:407–16. https ://doi.
org/10.1016/j.micro c.2013.05.014.
54. Rinnan Å, van den Berg F, Engelsen SB. Review of the most common
pre-processing techniques for near-infrared spectra. Trends Anal Chem.
2009;28(10):1201–22. https ://doi.org/10.1016/j.trac.2009.07.007.
55. Levey M. Mediaeval Arabic bookmaking and its relation to early chem-
istry and pharmacology. Trans Am Phil Soc. 1962;52(4):5–57. https ://doi.
org/10.2307/10059 32.
56. Merrifield MM. Medieval and Renaissance treatises on the arts of paint-
ing: original texts with English translations. New York: Dover Publications;
1999. (ISBN: 978‑0486404400).
57. Strolovitch DL. O libro de komo se fazen as kores das tintas todas (Trans-
literation). In: Afonso LU, editor. As Materias da Imagem. Lisboa: Campo
da Comunicação; 2010. p. 213–36 (ISBN: 978–9898465009).
58. Smith CS, Hawthorne JG. Mappae clavicula: a little key to the world of
medieval techniques. Trans Am Phil Soc. 1974;64(4):1–128. https ://doi.
org/10.2307/10063 17.
59. Neven S. Les recettes artistiques du Manuscrit de Strasbourg et leur
tradition dans les réceptaires allemands des XVe et XVIe siècles (Étude
historique, édition, traduction et commentaires technologiques). Doc-
toral dissertation, Université de Liège; 2011.
60. Clarke M. Mediaeval Painters’ materials and techniques: the Mont-
pellier Liber Diversarum Arcium. Londres: Archetype; 2011. (ISBN:
978‑1904982).
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
