www.rsc.org/analystRegistered Charity Number 207890

Showcasing analysis of cultural heritage materials from the laboratory of Professor Ruth Ann Armitage at the Department of Chemistry, Eastern Michigan University, Ypsilanti, USA.

Title: Characterization of blood in an encrustation on an African

mask: spectroscopic and direct analysis in real time mass

spectrometric identifi cation of haem

A rapid, simple direct analysis in real time mass spectrometry

(DART-MS) method can be used to confi rm the presence of blood

in cultural heritage objects like this African mask from the Detroit

Institute of Arts.

As featured in:

See Daniel Fraser

Analyst, 2013, 138, 4470.

Analyst

PAPER

Publ

ishe

d on

11

June

201

3. D

ownl

oade

d on

26/

10/2

014

10:1

2:05

.

View Article OnlineView Journal | View Issue

aDepartment of Chemistry and Physical Sci

Boulevard, Sylvania, Ohio, 43560, US

+1-419-517-8936bConservation Department, Detroit Institut

Michigan, 48202, USA. E-mail: cselviusderocJEOL USA, 44 Dearborn Rd, Peabody, Massa

comdDepartment of Chemistry, Eastern Michiga

Complex, Ypsilanti, Michigan, 48197, USA

734-4870290

Cite this: Analyst, 2013, 138, 4470

Received 31st March 2013Accepted 11th June 2013

DOI: 10.1039/c3an00633f

www.rsc.org/analyst

4470 | Analyst, 2013, 138, 4470–447

Characterization of blood in an encrustation on anAfrican mask: spectroscopic and direct analysis in realtime mass spectrometric identification of haem

Daniel Fraser,*a Cathy Selvius DeRoo,b Robert B. Codyc and Ruth Ann Armitaged

Prior to exhibiting an African Komomask from the collections of the Detroit Institute of Arts, a multianalytical

approach was undertaken to characterize the flaking encrusted coating on the surface of the mask.

Preliminary XRF and FTIR examination of the coating on the Komo mask revealed the presence of

significant quantities of iron and protein, possibly indicating the presence of blood. Raman spectroscopy

showed evidence for the porphyrin structure of haem as well. To confirm that blood was indeed present

in the coating, we developed a novel method for identifying the haem moiety from blood by use of in situ

methylation and direct analysis in real time mass spectrometry (DART-MS). Following a denaturing step

with formic acid, the resulting solution was combined with an excess of phenyltrimethylammonium

hydroxide to promote desorption, applied to a melting point tube, and placed into the direct analysis in

real time ion source gas stream at 550 �C. The permethylated haem ion (m/z 644.208) from myoglobin,

haemoglobin, fresh blood, and blood aged in the laboratory for 10 years was readily observed above the

background. By the described DART-TOF-MS method, permethylated haem was positively identified in the

mask coating, confirming the presence of blood. This method has obvious utility in forensic science

beyond that for identifying blood incorporated in cultural heritage materials.

Introduction

The identication of blood, oen degraded and in limitedquantities, is a major consideration in forensic science. Estab-lished protocols include visual inspection of suspected bloodstains, followed by presumptive tests. Such tests establish onlythe likelihood that a substance is blood based on a colour changeor production of light when the haemmoiety catalyses a chemicalreaction. The most commonly applied of these presumptive testsare phenolphthalein, tetramethylbenzadine, and luminol.1 Thephenolphthalein test has been described for use in conservationscience,2 and clinical test strips for identifying blood in urine,which utilize the tetramethylbenzidine reaction have beenapplied in the eld of archaeology to identify blood as residues inrock paintings3 and on stone tools.4,5 Presumptive tests suffersignicantly from a high rate of false positives from othercontaminants which can also act as catalysts.6 Furthermore,

ences, Lourdes University, 6832 Convent

A. E-mail: dfraser@lourdes.edu; Tel:

e of Arts, 5200 Woodward Ave, Detroit,

o@dia.org

chusetts, 01960, USA. E-mail: cody@jeol.

n University, 541 Mark Jefferson Science

. E-mail: rarmitage@emich.edu; Tel: +1-

4

these tests were never intended to stand alone for identicationof blood under other than clinical laboratory conditions.7 Positivepresumptive tests must legally be followed up with conrmatorytests in forensic science applications.1

Ideally, conrmation of the presence of blood should also benecessary in applications to art and archaeological materials.Instrumental methods, including various immunoassays8 andamino acid analysis with HPLC or GC-MS methods9,10 are themost common conrmatory methods in conservation science.These methods have been less successful for characterization ofpurported blood on archaeological materials.11 All chromato-graphic techniques require signicant sample preparation,including extraction and derivatisation for gas chromato-graphic analysis. The preparation procedure is time-consumingand requires the use of signicant quantities of solvent andoen toxic or hazardous reagents. Immunoassays12–17 andMALDI-TOF or tandem mass spectrometry of peptides orproteins via proteomics18–21 have both proven useful in deter-mining the species of origin of blood and proteins in and oncultural heritage materials.

With the growing access to mass spectrometric methods inforensic science laboratories, more specic methods for iden-tication of haem from blood based on mass have becomepossible. Mazel et al.22 reported the use of time of ightsecondary ion mass spectrometry to identify haem in bloodapplied to ritual artifacts from Africa. We report here ourinvestigations of the encrusted coating on an African Komo

This journal is ª The Royal Society of Chemistry 2013

Paper Analyst

Publ

ishe

d on

11

June

201

3. D

ownl

oade

d on

26/

10/2

014

10:1

2:05

. View Article Online

mask. This study has led us to develop a novel method for directand rapid identication of the haem functionality in blood andrelated proteins, as a conrmatory test for blood, with directanalysis in real time ionization23 coupled to high resolutiontime of ight mass spectrometry.

The Komo mask from the Detroit Institute ofArts

The Komo mask (DIA74.229) in the collection of the DetroitInstitute of Arts is the creation of the Bamana people of Mali,West Africa (Fig. 1). While the mask has been in the collectionsince 1974, the date of attribution is very broad, possibly fromthe late 19th into the 20th century. The mask is a compositeanimal head constructed from animal horns, skulls, quills,feathers, fur, cloth, and encrusted with a mud-like material. Allof these materials confer power to the mask which is dancedonly by Bamana men who are the members of the Komo powersociety. Ritual objects associated with this group are known tobe empowered by the ritual application of blood.24,25

This Komo mask was brought to the Conservation Depart-ment for examination and treatment in preparation for exhibi-tion. Concerns about the fragility of the aking mud-like

Fig. 1 The Detroit Institute of Arts' Komomask (74.229). Image courtesy of the DIA.

Fig. 2 FTIR spectrum of the Komo mask coating, showing significant overlapwith both albumin and dried blood (reference spectra from IRUG database ofartists' materials).

This journal is ª The Royal Society of Chemistry 2013

encrustation on the surface of the mask and the possible exis-tence of pesticide residues prompted questions about itscomposition in order to inform the conservation process. Theimportance of conserving the coating on the artefact requires anunderstanding of the nontangible aspects of the artifact.26 Mazeland his colleagues report extensively on characterization of so-called “patinas” and coatings onWest African ritual objects fromthe Dogon culture.27,28 A comprehensive approach to identifyingthe nature of the Komo mask coating was chosen, to combineseveral spectroscopic and mass spectrometric analyses.

ExperimentalSpectroscopic analyses at the Detroit Institute of Arts

Spectroscopic analyses were carried out in the ConservationDepartment at the DIA. X-ray uorescence spectra were acquiredwith a Bruker Artax XRF spectrometer under the following condi-tions: 50 kV, 700 mA,He purge,Mo tube, nolter, 240 s acquisitiontime.FTIRspectrawereacquired intransmissionmode (256scans)with the sample mounted in a diamond compression cell on aBrukerTensor27/Hyperion2000micro-FTIRspectrometer.Ramanspectra were acquired on a Bruker Senterra Raman spectrometerunder the following conditions: 532 nm laser, 50� objective,0.2 mW, 25� 1000 mm aperture, 180 s acquisition.

Mass spectrometric analysis

Myoglobin (from equine cardiac muscle, Sigma) and haemoglo-bin (Acros Organics) were used as reference materials containingthe haem moiety. Biological samples were prepared up to 10years ago at Eastern Michigan University (EMU) for long-termstudies of organic binders in rock paintings. Included amongstthe binders were blood (both reconstituted defribrinated beefblood from ICN Biomedical and clotted pig blood obtained fromHua Xing Chinese Market, Ypsilanti, MI), bone marrow, and egg(from local groceries) as negative controls. These materials wereapplied to clean glass microscope slides and placed undervarious conditions to age: stored in a climate controlled lab,stored outside sheltered from the elements, and stored outsideexposed to the elements. Small sub-milligram samples have beenremoved over the intervening years for other analyses.

Subsamples consisting of 2–4 mm2 were removed with asterile syringe needle and transferred to a clean glass vial. Formicacid (10 mL, 88%, Fisher Scientic), a well known protein dena-turant (Bradbury and King 1969), was added to the sample andallowed to briey act to free the haem moiety from the hemo-globin molecule. Aer approximately one minute, an excess(10 mL) of 20–25% phenyltrimethylammonium hydroxide (inmethanol, TCI America) was added to the resulting solution.Permethylation signicantly improves the detection limit of themethod. By protecting the two carboxylic acid groups on thehaem, the thermal desorption of the free haem occurs moreeffectively. This is analogous to derivatization in GC/MS and isdone for the same reason, to increase the volatility of the analyte.No signal for haem was observed without the combined additionof the acid denaturant and the derivatisation reagent. Theresulting solution was analysed directly.

Analyst, 2013, 138, 4470–4474 | 4471

Analyst Paper

Publ

ishe

d on

11

June

201

3. D

ownl

oade

d on

26/

10/2

014

10:1

2:05

. View Article Online

The closed end of a melting point capillary was used to pickup a small portion of the solution, and placed in the gapbetween Orice 1 of the AccuToFmass spectrometer ( JEOL USA,Peabody MA) and the DART ionization source (IonSense, Sau-gus, MA). The DART was used in positive ionization mode withhelium at 550 �C. Temperatures from 100–550 �C were tested,and no signal for permethylated haem was observed at less than550 �C. At this temperature, pyrolytic decomposition of theprotein matrix was also maximized. Orice 1 was set to 90 V toimprove fragmentation of the interfering protein decomposi-tion products and 120 �C; Orice 2 was held at 10 V, and the ringlens voltage at 8 V. The DART grid voltage was kept at the defaultvalue of +240 V. The mass spectrometer RF ion guide (“peaksvoltage”) was set to 2500 V to maximize sensitivity at the highmass range. Mass calibration was carried out with PEG-600(from m/z 239.150 to 1031.631) in methanol during eachacquisition. The mass resolving power was approximately 6000over this region.

ResultsXRF analysis

Iron dominated the XRF spectrum of the mask coating, whichalso showed minor to trace levels of Ca, K, Zn, Br, Si, Ti, Cl, Al,and Mn. The detection of Br is possibly indicative of the use ofmethyl bromide as a pesticide, applied at some point since themask's manufacture. The high Fe content appeared to supportthe idea that blood was likely part of the coating. Also, little tono particulate earthenmaterial was apparent upon examinationof mask samples by optical microscopy.

FTIR analysis

The FTIR spectrum of the material was an excellent match to anaged blood spectrum from the IRUG (Infrared and RamanUsers' Group) database of artists' materials, though it was notpossible to rule out other proteinaceous materials. Fig. 2 showsthat both aged blood and albumin (and other proteinaceousmaterials) correlate well with the spectrum obtained from the

Fig. 3 Raman spectrum showing the porphyrin and pyrrole stretches observed fo

4472 | Analyst, 2013, 138, 4470–4474

mask material. In order to obtain denitive verication of thepresence of blood in the sample, it would be necessary toconrm the presence of haemoglobin.

Raman analysis

While assessing specic changes in the haemoglobin chemistryof the dried blood on the mask is difficult, the Raman spectraacquired from samples yielded spectra with two strong bands(Fig. 3): 1588 cm�1, characteristic of haem porphyrin ringstretching, and a broader band at 1336 cm�1, falling within thepyrrole ring stretching range of 1300–1400 cm�1.29,30 PublishedRaman studies of haemoglobin feature considerable variabilityin the spectra obtained. Most Raman haemoglobin studiesreport the analysis of fresh blood; rare studies report forensicexaminations of recently dried blood; and still others report theanalysis of long frozen red blood cells from the 5300 year oldIceman.29–33

Presumptive test for blood in the Komo mask coating

An approximately 1 mm2ake of the mask coating was placed

into a conical vial, to which was added 10 mL of deionized water.Aer mixing, the resulting solution was placed onto a Hemastixtest strip, where any intact haem would undergo the tetrame-thylbenzadine test. Aer 1 min, the strip was compared to theprovided chart. The Komo mask coating sample gave a weakpositive presumptive result for blood. Due to the problems ofcross-reactivity with other peroxidases,1 such presumptive testsare insufficient to state that the coating contained blood.Conrmation with a second method is required to say thatblood is indeed present in the mask coating.

DART-TOF-MS analysis

Fig. 4 shows the complete mass spectra obtained for myoglobin,haemoglobin, and dried blood by use of the described proce-dure. The most intense peak in the spectrum is that of the M+ ofpermethylated haem (m/z 644.2085), measured with a massaccuracy of within 3–5 mmu. No ion was observed for

r the Komo mask coating.

This journal is ª The Royal Society of Chemistry 2013

Fig. 4 DART-TOF spectra for haemoglobin, myoglobin, and dried blood,showing permethylated haem ion.

Table 1 Summary of DART-MS results for positive and negative control samples

Sample m/z 644.209 intensity

Haemoglobin 39 149Myoglobin 20 508Fresh blood 14 779Debrinated beef blood 8613Dry clotted blood 14 846

Table 2 Haem signal intensity for dried blood as related to time and storageconditions

SampleStorageconditions Time, years

m/z 644.209intensity

Without pigment Outdoors 4 3286Without pigment Sheltered

outdoors4 571

Without pigment Laboratory 11 336With pigment Outdoors 4 1172With pigment onlimestone

Laboratory 4 7455

With pigment Laboratory 11 3482With pigment Laboratory 4 2225

Fig. 5 DART-TOF spectrum for Komo mask coating.

Paper Analyst

Publ

ishe

d on

11

June

201

3. D

ownl

oade

d on

26/

10/2

014

10:1

2:05

. View Article Online

unmodied haem (M+, m/z 616.177 or MH+, m/z 617.185) underany of the tested conditions. Because positive ion DART typi-cally results in the formation of protonated molecules, MH+, itis unclear at this time how the haem is being ionized, thoughdirect Penning ionization is one possibility. The harsh condi-tions used – acid, high temperature, derivatisation – makepredicting the ionization pathway particularly difficult. Similarresults were obtained for dry blood (either debrinated beefblood or clotted pig blood), but no signal at m/z 644.209 wasobserved in bone marrow or egg, other proteinaceous materialsthat might also be present in cultural heritage materials,particularly rock paintings (Table 1).

Even in aged blood samples exposed to weathering, thepermethylated haem was readily observed in the mass spec-trum, as shown in Table 2. Environmental conditions signi-cantly affected the amount of haem present. While we did notattempt to quantify the amount of haem, Table 2 shows theintensity of the peak at m/z 644.209 clearly decreases withboth time and exposure to weathering. Variations in thesignal intensity are also likely related to the size of the sampletested.

This journal is ª The Royal Society of Chemistry 2013

Because we are interested in using this application to iden-tify blood in rock paintings, samples with and without addedpigment were included in the study. The presence of pigment insamples exposed to the outdoors appears to signicantly reducethe relative intensity of the haem signal observed with thismethod. One possible explanation is that salts may form aprotective layer around the organic material, preventing theformic acid and methylating agent from reaching the haem inthese samples, or may interfere in the desorption process in theDART source. In the case of samples stored under laboratoryconditions, the differences in intensity are more likely related tothe amount of sample used in the analysis.

A small ake of the coating from the Komo mask was sub-jected to the same preparation procedure and characterized byDART-TOF-MS. A clear signal (Fig. 5) for the permethylatedhaem was observed at m/z 644.212, conrming the presence ofhaem, and thus blood, in the coating on the mask. Only a small

Analyst, 2013, 138, 4470–4474 | 4473

Analyst Paper

Publ

ishe

d on

11

June

201

3. D

ownl

oade

d on

26/

10/2

014

10:1

2:05

. View Article Online

ake of the coating – approximately 2 mm2 – was necessary forthis conrmatory test. The analysis was complete in less than 2minutes and carried out under atmospheric pressure withoutsignicant preparation of the sample aside from the addition offormic acid and the methylating reagent. This is a rapid, reli-able, and simple method for conrming the presence of bloodin cultural heritage materials.

Conclusions

The DART-MS method described herein is simple, rapid, andconrmatory for the presence of haem-containing proteins suchas myoglobin and haemoglobin. Because the presence of thehaem moiety is generally considered to conrm the presence ofblood in forensic analyses, this method has potential as a new,mass spectrometric conrmatory test for blood. Our primaryinterest in this method was to develop a reliable mass spec-trometric test for the presence of blood in and on archaeologicalmaterials, including lithic tools and rock paintings. Future workwill focus on these applications. This method was used toconrm the presence of haem, and therefore blood, in theceremonial Komo mask from the DIA.

Acknowledgements

The authors gratefully acknowledge nancial support from theNational Science Foundation (Award MRI-R2 #0959621), as wellas the EMU Chemistry Department and the Lourdes UniversityDepartment of Chemistry and Physical Sciences. Photos of theKomo mask were provided courtesy of the Detroit Institute ofArts. We thank Dr Nii Quarcoopome, DIA Curator of African Art,for providing the opportunity to characterize this material.

Notes and references

1 R. P. Spalding, in Forensic Science: An Introduction to Scienticand Investigative Techniques, ed. S. H. James and J. J. Nordby,CRC Press, Boca Raton, 2003, pp. 181–201.

2 D. Stulik and H. Florsheim, J. Am. Inst. Conserv., 1992, 31,275–288.

3 B. S. Williamson, J. Archaeol. Sci., 2000, 27, 755–762.4 V. Rots and B. S. Williamson, J. Archaeol. Sci., 2004, 31, 1287–1299.

5 T. H. Loy and B. L. Hardy, Antiquity, 1992, 66, 24–35.6 P. Manning, J. Archaeol. Sci., 1994, 21, 159–162.7 D. Fraser and R. A. Armitage, in Proceedings of the 37thInternational Symposium on Archaeometry, ed. I. Turbanti-Memmi, Springer, Berlin, 2011, pp. 205–210.

8 J. D. Reuther, J. M. Lowenstein, S. C. Gerlach, D. Hood,G. Scheuenstuhl and D. H. Ubelaker, J. Archaeol. Sci., 2006,33, 531–537.

9 M. R. Schilling, H. P. Khanjian and L. A. C. Souza, J. Am. Inst.Conserv., 1996, 35, 45–59.

10 D. A. Scott, M. Newman, M. Schilling, M. Derrick andH. P. Khanjian, Archaeometry, 1996, 38, 103–112.

4474 | Analyst, 2013, 138, 4470–4474

11 I. Joosten and M. R. van Bommel,Microchim. Acta, 2008, 162,433–446.

12 M. Lowenstein, J. D. Reuther, D. G. Hood, G. Scheuenstuhl,S. C. Gerlach and D. H. Ubelaker, Forensic Sci. Int., 2006, 159,182–188.

13 A. Heginbotham, V. Millay and M. Quick, J. Am. Inst.Conserv., 2006, 45, 89–105.

14 J. Arslanoglu and J. Schultz, Metrop. Mus. Art Bull., 2009, 67,40–45.

15 J. Arslanoglu, J. Schultz, J. Loike and K. Peterson, J. Biosci.,2010, 35, 3–10.

16 J. Arslanoglu, S. Zaleski and J. Loike, Anal. Bioanal. Chem.,2011, 399, 2997–3010.

17 G. Sciutto, L. S. Dolci, A. Buragina, S. Prati, M. Guardigli,R. Mazzeo and A. Roda, Anal. Bioanal. Chem., 2011, 399,2889–2897.

18 S. M. Stevens, S. Wolverton, B. Venables, A. Barker,K. W. Seeley and P. Adhikari, Anal. Bioanal. Chem., 2010,396, 1491–1499.

19 W. Fremout, M. Dhaenens, S. Saverwyns, J. Sanyvova,P. Vandenabeele, D. Deforce and L. Moens, Anal. Chim.Acta, 2010, 658, 156–162.

20 W. Fremout, M. Dhaenens, S. Saverwyns, J. Sanyova,P. Vandenabeele, D. Deforce and L. Moens, Anal. Chim.Acta, 2012, 728, 39–48.

21 I. D. van der Werf, C. D. Calvano, F. Palmisano andL. Sabbatini, Anal. Chim. Acta, 2012, 718, 1–10.

22 V. Mazel, P. Richardin, D. Debois, D. Touboul, M. Cotte,A. Brunelle, P. Walter and O. Laprevote, Anal. Chem., 2007,79, 9253–9260.

23 R. B. Cody, J. A. Laramee and H. D. Durst, Anal. Chem., 2005,77, 2297–2302.

24 S. C. Brett-Smith, Res, 2001, 102–136.25 P. J. Imperato, in Surfaces: Color, Substances, and Ritual

Applications on African Sculpture, ed. L. Kahan, D. Page andP. J. Imperato, Indiana University Press, Bloomington andIndianapolis, 2009, pp. 151–196.

26 S. P. Mellor, J. Am. Inst. Conserv., 1992, 31, 3–16.27 V. Mazel, P. Richardin, D. Touboul, A. Brunelle, P. Walter

and O. Laprevote, Anal. Chim. Acta, 2006, 570, 34–40.

28 V. Mazel, P. Richardin, D. Debois, D. Touboul, M. Cotte,A. Brunelle, P. Walter and O. Laprevote, J. Cult. Herit.,2008, 9, 347–353.

29 G. Rusciano, A. C. De Luca, G. Pesce and A. Sasso, Sensors,2008, 8, 7818–7832.

30 M. Janko, R. W. Stark and A. Zink, J. R. Soc., Interface, 2012, 9,2581–2590.

31 T. G. Spiro and T. C. Strekas, J. Am. Chem. Soc., 1973, 96, 338–345.

32 K. Virkler and I. K. Lednev, Anal. Bioanal. Chem., 2010, 396,525–534.

33 H. Sato, H. Chiba, H. Tashiro and Y. Ozaki, J. Biomed. Opt.,2001, 6, 366–370.

This journal is ª The Royal Society of Chemistry 2013