Forensic Science International 278 (2017) 338–350

Contents lists available at ScienceDirect

Forensic Science International

journal homepage: www.elsevier.com/locate/forsciint

Application of dye analysis in forensic fibre and textile examination:

Case examples
Tom G. Schotman, Xiaoma Xu, Nicole Rodewijk, Jaap van der Weerd*
Netherlands Forensic Institute, Division of Chemical and Physical Traces, P.O. Box 24044, 2490AA The Hague, The Netherlands

A R T I C L E I N F O A B S T R A C T

Article history:
Received 23 March 2017 Seven cases and a quality assurance test are presented. In these cases, fibres or textiles submitted for
Received in revised form 21 July 2017 investigation were analysed by HPLC–DAD–MS to identify the dyes present. The cases presented illustrate
Accepted 24 July 2017 that it is possible to identify textile dyes in fibre traces recovered for forensic analysis. The results show
Available online 31 July 2017 that a mixture of dyes is present in all textiles investigated, except one sample that was taken from a
manufacturer dye shade card. It is concluded that dye analyses improves the evidential value of forensic
Keywords: fibre examinations, as it becomes possible to distinguish textiles that are different in dye chemistry, but
Dye have a similar colour. In addition dye analysis makes the examination more robust, as it becomes possible
HPLC
to attribute colour differences between samples to identical dyes (mixed in different ratios) or to
Chromatography
chemically different dyes.
Mass spectrometry
Fibres and Textiles © 2017 Elsevier B.V. All rights reserved.

1. Introduction use additional techniques, such as Raman spectrometry, thin layer

Forensic investigation of trace evidence provides invaluable
information on the circumstances of and items involved in many
criminal activities. Trace examiners routinely deal with traces that
cannot or barely be seen by bare eye. Analytical instrumentation
used in the analysis of trace evidence must be very sensitive to
facilitate analyses of these minute traces. In addition such
instrumentation should be highly accurate: the more information
can be retrieved from a trace, the better it can be described,
provenanced and related to a known source.
Fibres are a commonly known instance of trace evidence. They
are very informative from a criminalistic point of view, as they can
be transferred from many textiles by only a slight contact. The main
instrument used in the examination of fibres is the microscope.
Experienced microscopists provide accurate descriptions of fibres,
while several colour block studies, such as those summarised by
Vooijs et al. [1], have shown that fibres from many sources can
effectively be discriminated. Other techniques used routinely in
most laboratories are UV/vis microspectrometry (MSP) and
Fourier-transform infrared spectrometry (FT-IR). Many laboratory
* Corresponding author.
E-mail addresses: tosch@nfi.minvenj.nl (T.G. Schotman), x.xu@nfi.minvenj.nl
(X. Xu), nirod@nfi.minvenj.nl (N. Rodewijk), jawee@nfi.minvenj.nl
(J. van der Weerd).
chromatography, chemical and melting tests.
Several groups have proposed chromatographic analyses to
chemically identify dyes used to colour textiles. Information on the
chemical nature of a dye or dye mixture may have pronounced
advantages, as it is complementary to colour information observed
by microscopists or analysed by MSP. It may e.g. (a) discriminate
between dyes with a similar colour, but different chemical
compositions; (b) reveal the presence of additional dyes (next to
the main dyes present) in quantities that are too small to induce a
noticeable colour difference in fibres; (c) relate fibres to a source
that contains very heterogeneous fibres; (d) provide information
on the original application of the investigated fibres. This is
possible as textile dyers use several criteria (stability/fastness,
price, application method) to select a dye for a specific application.
In a previous paper, the authors presented a method that
combines spectrometric separation by high pressure liquid
chromatography with analysis by diode array detection (DAD)
and mass spectrometry [2]. That contribution contains an
overview of earlier dye analyses, which is not be repeated here.
Later studies include the chromatographic studies on disperse dyes
by Kato et al. [3,4], studies on broader sets of fibres by Hoy [5] and
Morgan [6], the direct mass spectrometric study by Cochran et al.
[7], the microspectrometric [8–10] and Raman spectrometric
studies [9,11]. The main innovation of the method presented earlier
is the possibility to analyse a broad variety of dyes (disperse, direct,
reactive, basic and acid dyes) with a single chromatographic

http://dx.doi.org/10.1016/j.forsciint.2017.07.026
0379-0738/© 2017 Elsevier B.V. All rights reserved.
 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
method. Also, the inclusion of both DAD and MS analyses allows an
accurate and sensitive analysis of the dyes present in a single fibre.
In recent years, the authors have applied the developed method
in several cases. The present contribution presents a selection of
cases in order to provide insights into the data that can be obtained
and their evaluation in the framework of a case.
In three of the presented cases, analytical data is presented in
order to provide a general overview of the data that can be
acquired. The presented data is not complete, but comprises only a
small fraction of the acquired data sets. Inclusion of all relevant
data would make the article very long and would probably distract
from its main topic, i.e. the evaluation of the obtained results rather
than the results themselves.
Evaluation may vary from a simple confirmation of the
microscopic examination, to more relevant information on the
application of the investigated dye or textile or the heterogeneity
of a sample.
The case-descriptions presented here are based on the
information provided to us by the police or the prosecutor at
the time of the investigation. No additional information on the
backgrounds of the case or lawsuits were gathered, as the focus of
the current contribution is on laboratory procedures rather than on
case reconstruction.
2. Experimental
2.1. Experimental procedures
Presented MSP spectra were acquired using an J&M MSP800
(J&M Analytik AG, Essingen, Germany).
Three procedures have been developed to isolate dye molecules
from different types of fibres. The method is made for single fibre
analysis. The first procedure, developed to isolate basic dyes from
acrylic fibres, has not been used in the current study. The second
procedure has been developed to isolate reactive dyes from cotton
or viscose fibres. In this procedure, 10 mm of fibre is submerged in
10 mL NaOH (3 M, 4
C, 4 h). Afterwards, the fibre is rinsed twice in
acetic acid solution (0.5 M) and in cellulase solution (Trichoderma
Viride, 1.1 U/mg, Brunschwig Chemie, 0.01 g in 10 mL acetic acid
solution at pH 5), submerged in 10 mL cellulase solution, and mixed
in a thermo mixer (Eppendorf Comfort, 50
C, 550 rpm, 20 h).
Finally, the samples are centrifuged (5000 rpm, 5 min) and 10 mL
methanol (HPLC grade) is added. The third procedure is used to
extract direct, disperse and acid dyes. 10 mm of fibre is submerged
in 20 mL DMSO (z.a.) and heated to 100
C until the fibre is
discoloured or to a maximum of 2 h.
HPLC Analyses are carried out on a system consisting of an auto
sampler (Thermo Scientific Finnigan Surveyor Auto sampler Plus),
a pump (Thermo Scientific Finnigan Surveyor MS Pump Plus), a
pre-column (AJO-4286 and Guard cartridge holder KJ10-4282), a
column (Grom-sil 120 ODS-5 ST 150  2.0 mm i.d., 3 mm, Grace
Davison Discovery Sciences, Deerfield, USA). The column is kept at
22
C by a column oven Spark Holland Mistral. The column oven is
equipped with both a heating and a cooling system to accurately
maintain the set temperature. The injection volume amounted to
10 mL for DMSO extracts and 20 mL for cellulose extracts. A linear
gradient ammonium acetate 10 mM in water/methanol (95:5) and
ammonium acetate 25 mM in acetonitrile/methanol (50:50) was
used as mobile phase. The analysis time was 78 min. Eluents were
analysed by diode array detection (DAD, Thermo Scientific
Finnigan Surveyor PDA Plus Detector, spectral range 200–
800 nm) and mass spectrometry (Thermo Scientific LTQ Orbitrap,
scan range 150–2000 m/z). The obtained mass accuracy of the
Orbitrap system is better than 2 ppm with the use of a lock mass.
Entrance of the eluents into the mass spectrometer was enabled by
an ESI source at 4 kV, a discharge current of 20 mA, and a capillary
339
temperature of 300
C. A nitrogen flow was used as sheath gas. A
more elaborate description is provided elsewhere [2]. The only
alteration concerns the length of the HPLC run, which is now
78 min and a higher amount of injection volume for cellulose
extracts. Incompatibility of extraction solvents and eluents has
been reported [12]. Such issues were not encountered using the
procedures described above since the liquids are compatible with
MS systems.
All plots were prepared using Matlab (release 2015B, The
Mathworks).
2.2. Data interpretation
An analysis by HPLC–DAD–MS results in time series of DAD and
mass spectra. In these series, signals due to dyes can be recognised
by their absorbency of visible light (DAD spectrum). Corresponding
mass spectra can be selected on the basis of the derived retention
time. Often, mass spectra contain signals of other compounds, such
as adulterants and background ions as present in the eluent. Mass
peaks caused by the dye of interest can generally be recognised, as
they elute only at the retention time observed in DAD analysis. The
three relevant analysis parameters (retention time, DAD spectrum,
mass spectrum) can be related to a specific dye. An effective and
efficient way to achieve identification is by compilation of a
database with known dyes. In recent years, over 250 samples with
known dyes have been analysed. These were obtained from shade
cards or quality tests by dye suppliers. Known dyes present in the
database can easily be recognised by comparison of the relevant
parameters. Dyes that are not in the database require a more
elaborate search. This search is based on the molecular composi-
tion of the dye (as derived from the accurate mass) and comparison
of DAD spectra to those in the database. This last step is useful as
similar dye structures normally lead to similar DAD spectra. The
derived information on the molecular composition can then be
combined with information provided by the colour index or
information provided by dye suppliers. Sometimes, identification
is not possible, as the different sources of information do not yield
a consistent pattern, or no structure is registered to a dye name.
The dyes mentioned in the current paper are provided in Table 1.
No molecular structures are provided in this manuscript (except
those in Fig. 10), as these can easily be obtained by an internet
search.
The amount of data acquired for each sample is large and stored
in different software programs. As a result, the data collected from
reference samples cannot currently be shared in an easy format.
The interested reader is however invited to contact the corre-
sponding author for collaboration or data exchange.
2.3. Data extraction and images
An HPLC–DAD–MS run yields a large amount of information,
and only a small fraction of the acquired data can be shown. In the
current manuscript, analytical data are presented in Figs. 2, 5, 8 and
9. The structure of these figures is kept constant for clarity. The
current section explains this structure. All figures contain data on
both the known source (black lines) and the compared trace
material (grey lines). Sub plots (a) present the chromatograms
derived from DAD spectra (dotted line) and mass spectra (solid
lines). The mass chromatograms are prepared by extracting the
intensity at a specific mass or masses from the acquired data set.
The used mass and the corresponding retention time (along with
their elemental composition and identification) are printed as bold
lines in Table 1. The DAD chromatogram is prepared by extracting
the maximum absorbency in the visible range of the DAD spectrum
shown in sub-plot (b). Sub plots (b) present DAD spectra extracted
at the retention times shown in bold font in Table 1 (unless stated
340 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350

Table 1
Results and interpretation of HPLC–DAD–MS analyses obtained in cases 1–7. The 2nd column mentions the identity of the fibre matrix, as determined by microscopy or
infrared spectrometry. The 3rd column (‘Rt’) shows the retention time at which the dye under consideration eluted from the HPLC column. The 4th and 5th columns mention
the observed mass of this dye and the elemental composition as derived from this exact mass. The 6th column (‘identification’) shows the inferred identity of the dye.

Case Fibre Rt Mass Elemental comp. Identification Figure

1 Co 32.3 651.1127 C32H24N6O6S2 Direct Red 28 (congo Red) [M  H] 2
29.4 679.1075 C33H24N6O7S2 Not identified
29.7 663.1126 C33H24N6O6S2 DAD related to Direct Red 28
33.8 651.1127 C32H24N6O6S2 2
36.2 649.0972 C32H22N6O6S2
37.7 – – Unknown

2 Co 16.4 584.5629 C43H43N7O24S4 Reactive Orange 122 [M + 2glu  2H]2

17.5 746.616 C55H63N7O34S4 Reactive Orange 122 [M + 4glu  2H]2
19.9 492.0316 – – (red dye)

3 Co 16 624.5407 C43H44N7O27S5 Reactive Red 195 [M + 2glu  2H]2

14.2 – – – (Yellow dye)
17.5 – – – (Blue dye)

4 PES 50.1 504.1289 C22H24ClN5O7 Disperse Red 167 [M  H] 5

(car seat) 48 418.0922 C18H18ClN5O5 Disperse Red 167 [M  C4H6O2]
47.5 462.1184 C20H22ClN5O6 Disperse Red 167 [M  C2H2O]
44.6 420.108 C18H20ClN5O5 Disperse Red 167 [M  C4H4O2]

4 PES 51.4 331.0848 C20H13NO4 Disperse Red 60 [M] 5

(shirt suspect) 53.5 462.0564 C20H19Cl2N5O2S Disperse Red 153 [M + Ac  H]

5 (2a) PES 43.8 392.0594 C21H15NO5S Disperse Red 60 (variant) [M  H]

53.3 376.0649 C21H15NO4S Disperse Red 60 (variant) [M  H]

5 (2B) PES 50.4 457.1763 C27H24N2O4 Disperse Red 60 (variant) [M + H]+

5 (3) Co 18.2 533.5795 C39H43N9O21S3 – (Yellow dye) [M + 2glu  2H]2

23.1 494.6033 C40H45N7O19S2 – (Orange dye) [M + 2glu  2H]2

6 PES 46.6 360.0505 C15H12ClN5O4 Disperse Yellow 211 [M  H]

49 349.0265 C15H12Cl2N4O2 Disperse Yellow 211 [M  NO2 + Cl  H]
49.6 444.1677 C24H23N5O4 Disperse Orange 288 [M + Ac  H] 8
52.1 434.1111 C18H19ClN6O5 Blue dye, not identified 8

6 PA 39.8 807.0707 C32H28CoN8O10S2 Solvent yellow 151 [M] 9

44.4 796.0321 C36H20CrN6O11S Solvent Black 47 (variant) [M] 9
47.4 926.0046 C40H23CrN6O14S2 Solvent Black 47 [M  H]

PES 45.9 498.1627 C23H24O8N5 Disperse Red 82 [M + Ac  H]

46.6 360.0504 C15H12ClN5O4 Disperse Yellow 211 [M  H]
49.6 444.1677 C24H23N5O4 Disperse Orange 31
(variant Ac ! benzene) [M + Ac]
52 434.1108 C18H19ClN6O5 Disperse Violet 93 [M  Br + Cl]
54.5 532.071 C21H21BrN6O6 Disperse Blue 291G [M]
54.1 488.1216 C21H21ClN6O6 Disperse Blue 291G [M  Br + Cl]

Abbreviations: Rt retention time; glu glucose; M molecular structure; Ac Acetate; benz benzene;  not observed or identified, Co cotton, PES polyester, PA polyamide (nylon).
The parameters emphasised by bold print serve as the input for Figs. 2, 5, 8 and 9 (see Section 3.3 for details). The lines were only the mass is printed bold serve as the input for
the lower inserts in Figs. 2c, 5c, 8c and 9c.

otherwise). Sub plots (c) present parts of the acquired mass spectra
extracted at the retention times shown in bold font in Table 1. The
spectra of the trace material are shown as negative signals
(multiplied by 1) to enhance readability of this figure. Sub plots
(c) contain two inserts. The top insert is a plain magnification of a
part of the main spectrum shown. The bottom insert shows a detail
of a mass spectrum extracted from a different mass spectrum. The
displayed mass is shown as a bold number in Table 1 (only mass
shown bold).
The DAD and mass spectra shown have been normalised on the
highest value in the shown part of the spectrum. Some of the
chromatograms were given an offset to improve readability of the
figures.
straightforward as the dye analyses merely confirm and extend
the results of other methods used. Later cases have been included
paper as they provide some challenges, such as a complementary
evaluation with microscopic data (case 3), a trace that was altered
due to heat (case 4), a very small difference between dyes in the
known source and trace (QA test, case 5), a very high heterogeneity
(QA test, case 5), small traces (case 6), the combination of different
fibre types (case 7) and a very complex mix of dyes in samples with
a complicated evaluation (case 8).
For every case, a short description of the circumstances of the
case is provided, followed by description of microscopic and
analytical results. Finally, dye analyses are discussed within the
framework of the case.

3. Results 3.1. Case 1 — robbery (including analytical results)

In the present section, seven cases and a quality assurance test Description: A number of people entered a house and forced the
(‘case’ 5) are presented. Cases 1 and 2 are considered occupants to open their safe. They bound one of the occupants with
 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350 341

Fig. 1. Materials investigated in case 1. (a) Cord found at the crime scene; (b) tuft of fibres recovered from the suspects coat.

the red cord shown in Fig. 1a. This cord belonged to the victims and
had been found in their house. Shortly after the perpetrators left
the premises, policed chased a car with people they suspected of
involvement in this incident. During a chase, police saw that
several items were thrown out of the window of the car, such as a
(fake) gun. After the chase, one suspect was arrested, while the
others escaped. Police observed the presence of red tufts of fibres
on his gloves, and asked to compare these to fibres of the red cord
found at the crime scene.
Results: The traces recovered from the gloves consisted of small
threads. One of these is shown in Fig. 1b. During microscopic
examination, the fibres of the cord were classified as red cotton
fibres. They were compared to the red fibres found on the gloves
and were found to match. During microscopy, bleeding of the dye
from the fibre to glycerol (used as a mounting medium) was
observed.
Results of HPLC–DAD–MS are presented in Fig. 2. A general
description of this plot is provided in Section 3.3. Fig. 2b shows
DAD spectra of the cord (black line) and the tuft of fibres (grey line)
acquired at a retention time of 32.3 min. Chromatograms of the
maximum absorbency in this plot are shown as dotted lines in
Fig. 2a. Fig. 2c shows mass spectra, also acquired at a retention time

Fig. 2. Results of dye analysis on fibres in case 1. (a) Chromatograms, extracted from the acquired data by plotting the maximum absorbance (dotted lines) and or mass (solid
lines) of the dye involved of both the cord (known source, printed in black) and the tuft (trace, printed in grey); (b) DAD spectrum at a retention time of 32.3 min; (c) mass
spectrum at 32.3 min. The top insert in the spectrum is a magnification of the main spectrum and shows the high mass resolution obtained by the used instrumentation. The
bottom insert shows a detail of a mass spectrum acquired at retention time 33.8 min.
342 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
of 32.3 min. Several sharp bands are seen in these spectra. Many of
the peaks shown in the mass spectrum are considered background
signals, as they are observed fairly continuous at different
retention times. The band at m/z 651.1127 is not a background
signal. This is clear from the chromatogram of this band shown as
solid line in Fig. 2b, which shows that signals with this mass are
mainly observed at a retention time of 32.3 min.
Fig. 2 shows that the band at 32.3 min is visible in both
references fibres taken from the known source (cord) and the tuft
of fibres recovered from the glove. Also, it is observed both by DAD
and MS analysis, which indicates that the observed mass signal is
to be attributed to a dye rather than to a background signal. The
DAD spectrum in Fig. 2b shows a detailed spectrum containing
three peaks. A near perfect agreement between sample and known
source is observed. Agreement is also clear from the MS spectra,
presented in Fig. 2c. The spectra shown in Figs. 2b and c are
attributed to Direct Red 28 (Congo Red).
The top insert contains a magnification of the main spectrum in
Fig. 2c and shows the narrowness of the peak. In addition to Congo
red, small quantities of five other dyes are shown in both the
known source and the fibre traces (see Table 1). The mass spectrum
of one of these dyes is partly shown in the lower insert of Fig. 2c. It
has a mass equal to that of Congo Red (m/z 651,1127) but a different
retention time (33.8 min instead of 32.2 min, a minor signal at this
retention time is seen in the mass chromatograms (solid lines) in
Fig. 2a). This dye is not identified in detail, but might be an isomer
introduced during the synthesis of Congo Red.
Evaluation: The good agreement between the sample and the
known source confirms the agreement observed by microscopy
and MSP. In addition, the presence of 5 different additional
materials improves the evidential value, as the chance that an
arbitrary textile contains a similar mixture is considered to be
small. Finally, the identification of the main dye, Congo Red,
provides additional information on the textile on which it used.
Congo red has a very low wet fastness. This explains the bleeding
observed in microscopy. The known source, a cord mostly used for
decorative purposes, is a logical source for this dye, as it is not
meant to be washed or to be used in outside environments.
However, the use of Congo red would be most remarkable for most
clothing, domestic textiles and technical textiles. Also this
improves the value of the results.
3.2. Case 2 — car accident
Description: A person walking with her family on a road side was
hit by an object and died as a result of the impact. Two of her family
members were injured. Eye witnesses indicated the involvement of
a car trailer used to transport scrap metal. They had noticed that
one piece of metal protruded from the trailer. Police trace a trailer
with scrap metal nearby. No irregularities in the packing of its
contents were observed, but a reordering of the items after the
incident was suspected. The presence of coloured material on one
of the pieces of scrap metal is observed. This material is recovered
on a tape lift and sent to the laboratory along with the shirt of the
deceased.
Results: The tape lift from the scrap metal contained around
20 fibres, which were compared to the red cotton fibres of which
the investigated shirt was made. Microscopic and spectrometric
analysis did not reveal any relevant differences between the fibres.
Analysis of the dyes showed that both samples were dyed using
Reactive Orange 122. The results showed a number of bands due to
this dye. This is attributed to the pre-processing method: reactive
dyes are bound covalently to the cotton and cannot be extracted by
simple solvent extraction. Instead the procedure hydrolyses the
cotton chains. In practise, it is found that the method followed does
not hydrolyse cotton to monomers, but that dimers or oligomers
are also formed. It appears that the number of glucose units
attached to a reactive dye molecule does not alter its DAD
spectrum. However, the retention time and mass of the analysed
molecule are obviously affected. Based on the mass, it is possible to
derive the number of glucose units connected to the dye molecule.
Table 1 only shows the parameters derived for Reactive Orange
122 with 2 and 4 glucose units attached. In addition two Reactive
Orange 122, traces of a second dye were observed in the known
source and few of the analysed trace fibres. Based on the DAD
spectrum, this dye is probably red. However, the signal of this dye
were at the LOD (limit of detection) level for the mass
spectrometer. Since there are multiple peaks for a single dye,
the signal is also of lower concentration and the ionisation of
reactive dye extracts is unfortunately relatively low. It appeared
not possible to identify this dye.
Evaluation: The dye analysis confirms the results of microscopy.
The observed match is consistent with the hypothesis that the
victim was hit by the scrap metal on the trailer. However, the
evidential vale is limited, as red cotton fibres are among the most
frequently occurring fibres. No additional evidential value was
assigned to the presence of the second dye, due to its low
concentration.
3.3. Case 3 — rape
Description: During a sexual assault, the pink shirt of the victim
was damaged. An eye witness reported that he had observed a fight
between the victim and the offender. Later, a suspect was arrested.
The investigating police officer noticed a small thread on the coat
of the suspect. This thread was sent to the laboratory with the
clothing of the victim.
Results: The shirt received is made of pink cotton fibres. It is
damaged near the button fastening, as shown in Fig. 3a. The
damaged area consists of three layers of folded cloth, one of which
was not exposed before damaging of the shirt, as shown in Fig. 3b.
The fluorescence of the fibres in this non-exposed layer was found
to be different from the fluorescence of the exposed cloth. The
differences between the exposed and non-exposed parts were not
investigated in detail, but preliminary attributed to a reduced
exposure towards light and/or detergents.
Dye analysis indicated the presence of a mixture of dyes. The
main dye present was identified as Reactive Red 195. In addition,
small amount of two other dyes were found. These could not be
identified, but their DAD spectra indicated that one of them was
yellow, while the other was blue. Dye analysis also indicated the
presence of the same dyes in both the exposed and non-exposed
parts. Hence, discrimination of these parts on the basis of dye
analysis is not feasible.
The material recovered from the clothing of the suspect is
shown in Fig. 3c. Colour and fluorescence of the cotton fibres
present in this thread match the fibres in the non-exposed part of
the shirt. The mixture of dyes present in the shirt was also found in
the thread, including the additional (yellow and blue) dyes.
Evaluation: In this case, the results of the dye analysis were not
yet available when the case report for the court needed to be
delivered. Nevertheless, the analysis was finished so that the court
could be informed in case the analysis invalidated the results of
microscopic and spectrometric analysis. The results presented
show the complementarity of the described techniques. Fibres
from different parts of the shirt could be discriminated by
fluorescence microscopy. This discrimination was not possible
by dye analysis. Probably, the differences are not due to a different
dye content, but rather to additional components, such as optical
brighteners (It may be possible to analyse optical brighteners by
the same method, but this has not been explored yet). The origin of
the two additional dyes, as shown by dye analysis, is not clear. They
 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350 343

Fig. 3. Materials investigated in case 2. (a) Button fastening of the shirt of the victim with visible damage; (b) the seam of the damaged part consists of three layers of folded
cloth (top layer lifted by tweezers for visibility); (c) thread recovered from the suspects clothing.

might be due to small colour adjustments during dyeing of the
cloth from which the shirt was made. This case shows that
microscopy and dye analysis provide complementary information.
Dye analysis, and especially presence of the two minor dyes, form a
strong extra indication that the shirt of the victim was indeed the
source of the trace found on the suspects coat. If it is assumed that
the shirt is indeed the source of the recovered trace, microscopy
can indicate from which part of the shirt the trace was derived. In
this case, localisation is also informative, as the trace was related to
a part of the shirt that was not exposed (and thus not available to
transfer material) before the shirt was damaged.
3.4. Case 4 — explosion (including analytical results)
Description: In 2012, an explosion took place in car parked in a
residential area. Police suspected involvement of a man who was
treated in a nearby hospital for burning wounds. On the coat of this
man, policemen observe small red particles. The coat was
submitted to the laboratory for investigation of these particles.
As a known material, remnants of a red child seat recovered from
the damaged car was submitted. In addition, glass analysis was
requested, as several windows of the car were broken.
Results: The submitted part of the car seat, shown in Fig. 4a,
consisted of red polyester fibres. The traces, recovered from the
suspects coat, consisted of small particles that had clearly been
molten. Some of these are shown in Fig. 4b. Microscopic
comparison of these materials was basically uninformative by
the large difference in physical appearance. In addition, the
particles were too thick to allow acquisition of MSP spectra. The
particles were not squeezed or microtomed to reduce the
thickness. Instead, dye analysis was carried out, as it was reasoned
that the dye content (indicated by the red colour of the traces) was
not affected to a large extent by heat to which the material was
exposed. Results are shown in Fig. 5. The DAD spectra of the main
dyes present in the samples are shown in Fig. 5b. These spectra are
not similar and are extracted at different retention times, namely
50.1 min (known source, black curve) versus 51.4 min (trace, grey
curve). The chromatograms in Fig. 5a confirm the different elution
times of these dyes and also indicate the presence of different
types of dyes. Also the mass spectra shown in Fig. 5c and the inserts
in this figure show large differences. The red particles recovered
from the suspects shirts were found to contain Disperse Red 60 and
Disperse Red 153. Instead, the dye extracted from the car seat was
identified as Disperse Red 167. In addition, three other compounds
were found, probably related to Disperse Red 167. It seems unlikely
that the molecular structures of Disperse Red 60 and 153 are
formed due to degradation of Disperse Red 167, as their dye
structures are markedly different. Nevertheless, some hardened
parts of the car seat that had visibly been affected by heat were also
analysed to rule out the possibility of a change of dye content due
the exposure to heat. The dyes extracted from these hardened parts
were found to be identical to those in the intact material. Fig. 5a–c
shows clear differences between the dyes obtained from the car
seat and from the suspects coat.
Evaluation: The large difference between the dye content was
considered incompatible with the changes expected on heating a
sample and excluded the (received part of the) car seat as a
possible source for the traces found on the traces found on the
suspects coat. As the colour differences were relatively small and
the changes in the morphology of the sample could be explained by

Fig. 4. Materials investigated in case 3. (a) Part of the textile cover of the child seat; (b) particles of molten material recovered from the suspects coat.
344 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350

Fig. 5. Results of dye analysis on samples in case 4. (a) Chromatograms, extracted from the acquired data by plotting the maximum absorbency (dotted lines) and or mass
(solid lines) of the dye involved of both the child seat (known source, printed in black) and the thread (trace, printed in grey); (b) DAD spectrum obtained from material from
the child seat at a retention time of 50.1 min (black line). The DAD spectrum obtained from the particles on the suspects shirt did not yield a DAD spectrum at this retention
time. The grey spectrum shown is acquired at a retention time of 51.4 min; (c) mass spectra acquired at retention time of 50.1 min. The top insert shows a magnification of this
spectrum. The bottom insert shows a magnification of mass spectra acquired at 51.4 min.

the heat to which the suspect had been exposed, exclusion on the
basis of microscopy would be ambiguous. The source of the red
particles recovered from the suspects shirt was not discovered
since no additional samples were available. Glass particles
recovered from the same shirt, related the suspect to the explosion.
3.5. Case 5 — quality assurance test
Description: Every year, samples are sent to members of the
European Fibres and Hair Group (ETHG), a working group of
European forensic examiners. The provenance of these samples is
known by the organisers. The member laboratories can use these
samples as quality assurance tests, necessary for the laboratory
accreditation scheme, such as ISO17025. In 2014, three sets of
samples (numbered 1–3) were distributed for comparison. Two of
these sets (2a and b, 3a and b) are described here.
Results: Samples 2a and b consisted of polyester fibres (PET).
They could not be discriminated microscopically. MSP results of
these fibres, presented in Fig. 6a, provided a small but reproducible
difference of the absorption around 325 nm. This absorption
appears as a small shoulder in the spectra of sample 2a. Dye
analysis showed that the colour of sample 2a can be attributed to a
mixture of two dyes. The DAD spectra of these dyes both resemble
that of Disperse Red 60, which is present in our database. However,
the retention times and the molecular masses found in the
corresponding mass spectra did not correspond to Disperse Red 60.
A single dye signal was observed in sample 2b. The DAD spectrum
of this dye was also similar to Disperse Red 60, but the retention
time and mass indicated yet another variety. It was concluded that
the dyes in samples 2a and 2b are distinctly different, but that they
are all related Disperse Red 60.
Samples 3a and b consisted of red cotton fibres. Microscopy of
these samples was of limited value due to their very heterogeneous
colouration. This heterogeneity is also clear from the highly
variable absorption values in the absorbency spectra shown in
Fig. 6b. Microscopic comparisons of these samples was experi-
enced as difficult by the examiners due to the observed variation.
Dye analysis of a number of fibres from both samples indicated the
presence of two dyes in both samples 3a and 3b. Based on their
DAD spectra, it was concluded that these dyes are probably orange
and yellow respectively. The dyes could not be identified. However,
it was found that the ratio between the two dyes was similar in
both samples.
Evaluation: Samples 2a and b were discriminated by MSP.
However, the differences are small and in the UV region of the
spectrum. Such small differences may sometimes be explained by
the alteration of either the trace or the known material. If a suspect
flees from the crime scene through water or washes his clothing to
remove stains, small differences in the MSP spectra might become
visible. In many cases, such changes can be reproduced using
laboratory tests, but these require extra testing and evaluation as
well as assumptions on the history of the textile. Instead, the dye
analysis revealed that the dye used to colour the samples are
chemically different. It is highly unlikely that such differences are
due to alterations to one of the samples. Afterwards, the organisers
of the test revealed that samples 2a and 2b were taken from a
sample book. Sample 2a was dyed with Resolinbrilliantrot BS (i.e.
Disperse Red 132), Sample 2b with Resolinbrilliantrot BLS. (During
evaluation of the results, disperse Red 132 was considered as the
dye used to colour sample 2a. However, this assumption could not
be checked and was not reported, as reference material of this dye
was not available at that time.)
 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
Fig. 6. Absorption spectra obtained in case 5. (a) Samples 2ab; (b) samples 3ab.
Samples 3a and 3b were very heterogeneous. Dye analysis
showed the presence of the same dye mixture in both samples in a
similar relative amount. This implies that most of the heterogene-
ity must be explained by a irregular dye uptake rather than by the
presence of a different dye or a different mixture of dyes. This
strengthened our conclusion that the samples might be derived
from the same source. This conclusion appeared to be correct.
3.6. Case 6 — robbery (including analytical results)
Description: During a robbery in a jewellery store, a window and
glass displays were smashed using a hammer. The perpetrators
took jewellery from these displays and left the premises, leaving
behind the hammer. This hammer is shown in Fig. 7a. After a short
time, two men were arrested based on eye witness statements. A
pair of gloves was found near the place where one of the suspects
was arrested. These gloves and the hammer were submitted for
analysis.
Results: The gloves, shown in Fig. 7b, consist of grey and black
parts. The black parts contain knitted polyester. These were
analysed by HPLC–DAD–MS. Results of this analysis are shown in
Fig. 8. These parts were found to contain a complex, but identical
dye mixture, comprising of Disperse Yellow 211, a variation of
Disperse Yellow 211, Disperse Orange 288, and an unidentified blue
dye. Based on the retention time, this unidentified blue dye is
probably also a disperse dye. The lack of a UV–vis spectrum for the
trace is due the very low concentration which hinders extraction of
345
the spectrum. However the MS signal is strong enough for the
identification of the dyes.
The grey parts of the glove are made of artificial suede,
containing different types of polyamide fibres (including micro-
fibers). These microfibers were not analysed by HPLC, due to their
small sizes and their rather pale colour. It was estimated that the
the chances to obtain useful information on the dyes in these fibres
was small. The other type of polyamide fibres were found to
contain three major dyes, which were identified as coming from
Solvent Yellow 151 and two variations of Solvent Black 47 (results
shown in Fig. 9). These two variations are consistent with an earlier
analysis of Solvent Black 47 reference material which had been
stored in our database. The difference between the proposed
structures, shown in Fig. 10 is an additional sulphonate group in
one of the structures and the replacement of a naphthalene group
by a benzene group. The colours of the two, as indicated by DAD, is
almost equal.
Figs. 8 and 9 show rather large differences between the
chromatograms extracted from DAD and from MS: the major peak
in the MS chromatogram is relatively small in the DAD chromato-
gram and vice versa. This is attributed to a different sensitivity of
different dyes towards DAD and MS detection. Large differences in
ionisation sensitivity were also noted by Cochran [7] and ionisation
is often related to the chemical structure. The hammer was
sampled using a cotton swab to recover DNA traces. Fibres present
on these swabs were isolated and used for comparison with fibres
used in the construction of the gloves. Microscopic comparison
346 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
Fig. 7. Materials investigated in case 6. (a) Hammer; (b) gloves.
yielded three types of matching fibres, namely dark polyester
fibres, polyamide fibres, and polyamide microfibers.
The polyester and polyamide traces were analysed using HPLC–
DAD–MS. The dyes present in the fibres and their relative ratios
were found to match those found in the different parts of the
gloves accurately. Additional analyses also investigated a relation
between one of the suspects and the glove (DNA analysis), between
the suspects and the hammer (DNA) and between glass particles
found on the gloves and glass recovered from the crime scene.
Evaluation: In this case, the evidence is very strong, as matches
were found between different types of synthetic fibres used in
construction of the gloves and traces from the hammer. The
mixture of dyes found in the polyester fibres confirmed this match
and high evidential value.
In this case, it was found that a mixture of two varieties of
Solvent Black 47 were extracted from polyamide fibre processed in
the glove. The proposed structures only show small differences.
The background behind the presence of these two related forms is
unknown. It can be hypothesised that both forms may have been
synthesised during production of the dye, or that one form has
degraded to the other. The DAD spectra obtained during HPLC–
DAD–MS analysis indicate that the absorption of both forms is
similar or identical (data not shown). It is thus not possible to
reveal the presence of both forms on the basis of microscopy or
MSP. Independent of the origin of the two forms of Solvent black 47,
the discrimination of the two dye forms and the result that both
form are present in both the known source and the fibre traces
makes the evidence stronger.
3.7. Case 7 — robbery
Description: After a ring of her door bell, a woman opened her
front door. Directly after, she was beaten several times by two
unknown persons, fell on the ground, was threatened by a gun,
taped, and forced to reveal the code of her safe. Jewellery and
money were taken from this safe. Later, a suspect was identified,
and two different gloves were sent to the laboratory, along with the
tape used to bind the victim.
Results: One of the obtained gloves is shown in Fig. 11. This glove
consists of fleece material. The fibre types used for the base layer
and the fleece layer can be discriminated, as can be expected for
fleece material [13].
HPLC–DAD–MS of the fleece fibres showed the presence of at
least 6 types of dyes. These are Disperse Red 82, Disperse Yellow
211, a variation of Disperse Orange 31, Disperse Violet 93, Disperse
Blue 291G, and a variation of Disperse Blue 291G. Several fibres
(>70) recovered from the tape were found to match to the fleece.
Some of these were analysed using HPLC–DAD–MS and were found
to contain the same mixture of dyes.
Evaluation: The most striking result of the dye analysis carried
out in this case was the presence of at least six different dyes. Two
of these structures are related (Disperse Blue 291G and the
 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350 347

Fig. 8. Results of dye analysis on polyester fibres in case 6. The structure of this figure is identical to that of Figs. 2 and 5. (a) Chromatograms of the maximum absorbance of the
DAD spectrum shown in (b), and the mass peak shown in the upper insert in (c); (b) DAD spectrum at Rt = 52.1 min. No useful DAD spectrum could be acquired of the polyester
traces recovered from the hammer; (c) mass spectrum acquired at Rt = 52.1 min. The top insert shows a detail of this spectrum. The bottom insert shows a detail of the mass
spectrum acquired at 49.6 min.

Fig. 9. Results of dye analysis on polyamide fibres in case 6. The structure of this figure is identical to that of Figs. 2, 5 and 8. (a) Chromatograms of the maximum absorbency of
the DAD spectrum shown in (b), and the mass peak shown in the upper insert in (c); (b) DAD spectrum at Rt = 39.8 min. No useful DAD spectrum could be acquired of the
polyamide traces recovered from the hammer; (c) Mass spectrum acquired at Rt = 39.8 min. The top insert shows a detail of this spectrum. The bottom insert shows a detail of
the mass spectrum acquired at 44.4 min.
348 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350

Fig. 10. Proposed structures of two varieties of Solvent Black 47 extracted from the polyamide fibres of the gloves in case 6.

variation thereof) and might have been formed during a single
synthesis (see discussion in case 6). The other dyes are not related
and it is likely that this mixture was deliberately chosen to obtain a
specific colour shade. This complex mixture adds to the evidential
value of the results substantially.
3.8. Case 8 — fishing pods
Description: A man was arrested while attending illegal fishing
pods. These fishing pods were recovered. Nearby the spot, some
other pods were recovered. Police noticed the reddish colour of the
net material used. At the house of the suspect, a dye bath
containing a consumer type dye mixture was found, as well as
some other pods. Police obtained (unconfirmed) information that
dyeing of pods reduces their degradation due to fungal or other
biological activity. The different pods, as well as a sample from the
dye bath were submitted to the lab to establish a possible relations
between the items of the suspect and the illegal pods found in the
water.
Results: Microscopy and infrared microspectrometry of the
different pods showed that the different pods were constructed of
polyamide fibres with an identical morphology. The colour of fibres
in all pods was pale red. Because of this pale colour, we extracted
dyes from a small piece of thread rather than from a single fibre.
The colour intensity and fluorescence of fibres varied within and
between the different samples. MSP spectra of fibres from the
different pods had a similar shape, but varied in intensity. HPLC–
DAD–MS data were acquired of a number of the obtained pods. The
pods recovered during the arrest of the suspect were found to
contain a complex of tens of dyes. A chromatogram based on the
full DAD spectra are shown in Fig. 12. Some of the observed dyes
are acid dyes (e.g. Acid Red 82, Acid Blue 25) which are to be
expected for polyamide fibres. In addition, several other dyes were
identified, including direct and disperse dyes. These dyes are not

Fig. 11. Glove investigated in case 7.

 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
Fig. 12. Chromatogram obtained from one of the pods investigated in case 8 by integration of the visible range of the DAD (400–700 nm). (a) Full chromatogram; (b) detail of
the same chromatogram, in which all signals above 1% of the signal shown in Fig. 11a have been cut off. This view shows the presence of small signals due to several additional
peaks.
normally applied in polyamide textiles. The dye content of these
pods were compared to that of the other pods recovered from open
water. A number of dyes were found in all pods from the suspects
and pods recovered from open water. Other dyes were found only
in one or a number of pods.
In addition, a sample from the dye bath was analysed. Also the
dye bath contained a large number of different dyes. However, the
overlap between the dye bath and the pods was limited.
Evaluation: Processing of the data acquired in this case
appeared challenging due to the high number of dyes in each of
the samples. It became clear however that mixtures of tens of dyes
can be processed without major problems. Several dyes were
identified in both the pods found in open water and those
recovered from the suspect. However, the dye mixture was not
identical for all pods. Differences were found between pods from
different locations. However, pods from the same location were
also found to have a different dye content. It was hypothesised that
these differences can be explained by a dye process that is not
accurately controlled and the washing out of dyes with low
substantivity for the used polyamide material. Summarising, the
observed similarities with respect to morphology, material use and
dye content provided an indication for a relation between the
different pods. It was also concluded that the dye content indicated
that the pods were dyed in a consumer type dye bath.
A relation between the dye bath and the pods and the dye bath
would indicate that the suspect was not only involved in
(occasional) illegal fishing, but also made more advanced
preparations. However, the pods could not be related to the dye
bath easily (i.e. to the specific dye bath belonging to the suspect
and sampled by the police). The analysed dye bath sample
contained several types of dye classes. This is to be expected for
consumer type dye baths, as these need to be able to colour a
variety of textile fibres. Also, a mixture of dyes was found in the
samples from the pods found in open water and those from the
suspect. The presence of this mixture of dyes indicates that the
pods were indeed dyed using a consumer type dye bath. However,
only a few of the dyes identified in the dye bath corresponded to
349
those in the pods. This can be attributed to (a combination of)
different possible causes (a) there is no direct relation between the
dye bath and the pods (other than that some of the pods and the
dye bath were related to the suspect); (b) some dyes were already
present in the pods before they were dyed in the dye bath; (c) some
dyes in the dye bath are exhausted due to their affinity/
substantivity to the textile being dyed, while dyes with a lower
substantivity are still present in the dye bath; (d) masking effects
due to the varying sensitivity of different dyes and the large
number of dyes present.
4. Discussion and conclusion
In the current study, results obtained in seven cases as well
those obtained in a quality assurance test care presented. In most
cases, it appeared possible to identify one or more of the dyes in the
obtained textiles. In other cases, identification appeared impossi-
ble, as the dye under consideration was not in our database, and the
structure could not be derived from the obtained results. Even if
identification is not feasible, the obtained results (retention time,
DAD spectrum, mass spectrum) provide a useful ‘fingerprint’ of the
dye under consideration which is at least useful to compare fibres
from a known source to recovered traces.
The most important conclusion of the current contribution is
that it is possible to identify dyes in single textile fibres in case
work. It appears possible to obtain useful DAD and mass spectra
from minute samples, which makes the technique suitable for
forensic analysis. To our knowledge, these are the first reports in
which successful dye analyses have been carried out as a routine
technique in forensic fibre examination.
Our previous contribution [2] was based on a small set of
reference materials. In the current studies, identification of several
additional dyes is presented. In most of the presented cases, only a
single textile fibre or a part thereof was used for each analysis. In
one of the described cases (case 6, Figs. 8 and 9) the DAD spectrum
appeared too low to be useful for the traces. Nevertheless, even in
that case, the nature of the dye could be inferred from the obtained
350 T.G. Schotman et al. / Forensic Science International 278 (2017) 338–350
mass. In our experience, a single textile fibre of a few millimetres
length is sufficient for darker and intensely coloured fibres.
Samples of lighter fibres need to be somewhat bigger, while very
pale colours may not be suitable for analysis. The sensitivity of the
mass spectrometric analyses is dependent on the chemical nature
of the dyes. For example, reactive dyes are more readily analysed
than disperse dyes.
Mixtures of dyes were identified in all textiles, except for one
reference sample obtained from a dye shade card (case 5, sample
2b), The number of dyes ranges from a few to around 50 in case 8.
This illustrates the high added value that can be derived from
independent identification of dyes present as a mixture. Several
possible reasons exist for the presence of mixtures of dyes. The
most obvious reason is the use of a dye mixture during colouration
to achieve the desired shade during production of the textile.
Alternatively, the dyes present might be side products or
degradation products of a main dye. Finally, dyes may be
transferred from other textiles by laundry procedures or, as shown
in case 8, by additional colouration.
The possibility to analyse dyes in a mixture independently has
two major advantages. First, it increases the evidential value that
can be obtained. This is especially true if the relative concentration
of the different dyes can also be quantified. Full quantification
would require standardisation of many variables, e.g. the exact
amount of sample inserted. This would be very challenging and is
not intended. However, a relative quantification can be based on
the ratio of signals obtained by DAD and/or mass spectrometry. A
relative quantification can help to include or exclude a known
source as the donor of a specific trace. In addition, some dyes might
be present in very small amounts, e.g. due to dye bleeding during
cleaning procedures. Changes in dye mixtures might be visible
microscopically as a colour difference. However, they might be
very well be unnoticeable, as dyes in small quantities can be
masked by the more intense signals of the main dyes.
The advantage of dye analysis by HPLC–DAD–MS over analysis
by Raman and microspectrometry [9] is paramount. Future work
on direct high resolution mass spectrometry, as proposed by
Cochran [7] might also provide independent identification of
different dyes, though it might become challenging to recognise
adulterants, breakdown products, and independent dyes since
these are not separated by chromatography.
The high discrimination of HPLC–DAD–MS comes at a cost. The
time required to extract and analyse the dyes, as well as the time
required to evaluate results can be considerable, especially in cases
such as case 8, where a variety of dyes is present. In addition, the
methodology is destructive, though the required amount of
material is only a few mm for an analysis. Nevertheless, it is
planned to continue these analysis, especially in cases were more
than one or two fibre traces were recovered, due to the relevance of
the derived information.
Secondly, independent identification of dyes in a mixture make
the analysis more robust. Fibres that contain different amounts of
dye, or a mixture of dyes in different ratios, will show colour
differences in microscopy. Comparison of such heterogeneous
samples is often awkward, and the examiner may have to compare
a trace with several samples from the known source before a
matching fibre is found, or before he concludes that the trace fits
within the range of variation of the known source. In contrast,
analysis by HPLC–DAD–MS might show that the same dye or
mixture of dyes is available in varying quantities. The second
advantage is illustrated by case 5. The red fibres investigated were
inhomogeneous leading to a large variation in MSP spectra
(Fig. 6b). Dye analysis showed both samples obtained were
coloured with the same dyes. These could not be identified, but still
formed an useful matching property. Another example of the
robustness of the analysis was shown in case 4. In this case, the
investigated traces were affected by heat, which made a
microscopic analysis of fibre morphology irrelevant. Nevertheless,
a clear exclusion was provided on the basis of dye analysis.
Together, these factors make dye analysis by HPLC–DAD–MS a
very robust and discriminating technique in the analysis of fibre
evidence.
References
[1] C. Vooijs, P. Vergeer, J. van der Weerd, Towards source level evaluation of the
evidential value of fibre examinations, Forensic Sci. Int. 250 (2015) 57–67, doi:
http://dx.doi.org/10.1016/j.forsciint.2015.02.009.
[2] A. Carey, N. Rodewijk, X. Xu, J. van der Weerd, Identification of dyes on single
textile fibers by HPLC–DAD–MS, Anal. Chem. 85 (2013) 11335–11343, doi:
http://dx.doi.org/10.1021/ac402173e.
[3] T. Kato, T. Ikeue, Y. Suzuki, M. Handa, Analysis of disperse dyes using liquid
chromatography/linear ion trap mass spectrometry (LC/LIT-MSn) and database
construction, Anal. Sci. 32 (2016) 147–152, doi:http://dx.doi.org/10.2116/
analsci.32.147.
[4] T. Kato, Y. Suzuki, M. Handa, Extraction and analysis of disperse dyes from
colored polyester single fibers using liquid chromatography/linear ion trap
tandem mass spectrometry, Anal. Sci. 32 (2016) 1019–1022, doi:http://dx.doi.
org/10.2116/analsci.32.1019.
[5] S.J. Hoy, Development and Figures of Merit of Microextraction and Ultra-
Performance Liquid Chromatography for Forensic Characterization of Dye
Profiles on Trace Acrylic, Nylon, Polyester, and Cotton Textile Fibers, University
of South Carolina, 2013 http://scholarcommons.sc.edu/etd/2391 (Accessed
4 February 2016).
[6] S.L. Morgan, 2015. Validation of Forensic Characterization and Chemical
Identification of Dyes Extracted from Millimeter—Length Fibers, https://www.
ncjrs.gov/pdffiles1/nij/grants/248579.pdf.
[7] K.H. Cochran, J.A. Barry, D.C. Muddiman, D. Hinks, Direct analysis of textile
fabrics and dyes using infrared matrix-assisted laser desorption electrospray
ionization mass spectrometry, Anal. Chem. 85 (2013) 831–836, doi:http://dx.
doi.org/10.1021/ac302519n.
[8] E.J. Reichard, E.G. Bartick, S.L. Morgan, J.V. Goodpaster, Microspectrophoto-
metric analysis of yellow polyester fiber dye loadings with chemometric
techniques, Forensic Chem. 3 (2017) 21–27, doi:http://dx.doi.org/10.1016/j.
forc.2016.11.001.
[9] J. Was-Gubala, R. Starczak, Nondestructive identification of dye mixtures in
polyester and cotton fibers using Raman spectroscopy and ultraviolet–visible
(UV–vis) microspectrophotometry, Appl. Spectrosc. 69 (2015) 296–303, doi:
http://dx.doi.org/10.1366/14-07567.
[10] J. Was-Gubala, R. Starczak, UV–vis microspectrophotometry as a method of
differentiation between cotton fibre evidence coloured with reactive dyes,
Spectrochim. Acta A Mol. Biomol. Spectrosc. 142 (2015) 118–125, doi:http://dx.
doi.org/10.1016/j.saa.2015.01.116.
[11] J. Was-Gubala, W. Machnowski, Application of Raman spectroscopy for
differentiation among cotton and viscose fibers dyed with several dye classes,
Spectrosc. Lett. 47 (2014) 527–535, doi:http://dx.doi.org/10.1080/
00387010.2013.820760.
[12] R. Griffin, J. Speers, High performace liquid chromatography, Forensic
Examination of Fibres, CRC Press, 1999.
[13] K. Lunstroot, D. Ziernicki, T. Vanden Driessche, A study of black fleece
garments: can fleece fibres be recognized and how variable are they? Sci.
Justice 56 (2016) 157–164, doi:http://dx.doi.org/10.1016/j.scijus.2015.12.006.