Analysis of the Naturally Aged Scented Components of Montien Boonma’s House of Hope

Featured Application

Techniques described were here employed to identify herbs and spices used create an artwork dating to 1997.

Abstract

Analysis of the scented components of the art installation House of Hope by Montien Boonma, including eight bags of unlabeled or poorly labeled powder and five strands of aromatic beads, was completed to facilitate its display at the Museum of Modern Art (MoMA). Though an olfactory experience is central to the piece, limited information was available concerning the origin of the object’s scent. Identification of the aromas from the powders and beads, some of which were nearly 30 years old, was accomplished through visual assessment, attenuated total reflectance–Fourier-transform infrared spectroscopy (ATR-FTIR), and static headspace solid-phase microextraction gas chromatography–mass spectrometry (HS-SPME-GCMS) and confirmed using controls. Challenges included spices that had lost some of their potency or become cross-contaminated. The contents of five of the eight bags of powder were successfully identified as black pepper (two bags), clove, turmeric, and white sandalwood (Santalum album). All beads contained nutmeg, peppermint, ginger, and turmeric, while licorice root, thyme, cardamom, and clove were noted in some. The beads were bound using pine honey, a unique type of honey produced by bees that feed on aphid excretions. Identifying the scented components informs current and future installations so that the artist’s original intent is more closely approximated.

1. Introduction

During his lifetime, renowned contemporary Thai artist Montien Boonma (1953–2000) showed his work internationally and nationally [1]. Influential as both a painter and sculptor, his work drew inspiration from various sources, including Buddhism and Buddhist practice, the arte povera movement, and the pre-industrialized culture of Thailand [2,3]. His works combined found objects, natural local materials, and scented organic components to explore the dualities of life, including grieving and healing, globalization and rural life, and suffering and peace [3,4]. The scale of his works was often large, encouraging the viewer to be enveloped by, or become part of, the piece.

Boonma conceived the monumental House of Hope (Figure 1) in 1996–1997 in the years following the death of his wife. Before MoMA acquired it in 2023, the artwork had been displayed at several other venues, including the Museum of Contemporary Art in Tokyo, Japan (1996), Deitch Projects in New York City (1997), the DESTE Foundation for Contemporary Art in Athens, Greece (1998), and was included in the 2005 Venice Biennale [5]. The work is composed of two major elements: a central construction that evokes a house or temple [4] and a mural that surrounds it. The house is created from 1385 strands of fragrant, hand-strung beads suspended from a metal grid. Nestled beneath are 440 red-painted stools stacked and organized into a rectangular shape. The surrounding mural, painted with glutinous rice starch paste and powdered herbs and spices, evokes the candle soot and smoke that settles on the walls of Buddhist temples. In 1998, writer Frances Richard noted that the combined scent from the beads and mural created a “… sharp, musty, penetrating [and] pleasurable tickle high up in the delicate nose-hairs” [6].

When MoMA accessioned House of Hope in 2023, it arrived in 13 separate crates. Figure 2a–h shows images of eight bags of powders as they appeared when removed from the crates. In some instances, the bags had been packed together in one container. This possibly led to cross-contamination of the volatile organic compounds (VOCs) emitting from them. A representative selection of five strands of differing-color beads, chosen from more than 1385 shipped with the object, was sequestered for visual and chemical analysis (Figure 2i–m).

House of Hope was examined one year before its display at MoMA. There was not enough extant powder present to create the mural or enough complete strands of bead to execute the hanging gabled shape. Written reports generated during previous installations disclosed somewhat vague—by scientific standards—accounts of the scented materials used to make both elements. One reported that the beads came “… in a variety of colors … composed of medicinal herbs and honey …,” while the other reported that “… about 18 medicinal herbs and spices …” were used to create the wall painting [7,8]. The objects conservators requested identification of the spices and herbs used to create the mural and the beads so that the installation could proceed. The results would inform the current installation (2023–2024) and future presentation of the work.

The body of literature describing the use of SPME-GCMS to identify the scent markers of herbs and spices is voluminous and spans many fields, including food science, fragrance and cosmetics development, and medical research [9,10]. Headspace SPME in particular has successfully been used to identify the origin of the aroma of fresh flowers and spices [11,12,13]. The same technique is used in cultural heritage; however, it is typically used to look for evidence of volatile organic compounds (VOCs) evolving from construction materials and commercial products or to predict the impact of VOCs on the stability of cultural artifacts [14,15,16,17]. While there is some literature discussing the positive impact scents and odors have on the experience of a museum visitor [18,19,20], only one article was found that cited the analysis of spices slated for use in a museum environment [21]. The goal there was to determine whether off-gassing from the spices would damage art nearby. Other articles relevant to the field of cultural heritage have tended to focus on characterizing the VOCs from traditional artists’ materials like wax or varnish [22,23]. From a literature review, there were seemingly no published articles from the cultural heritage sphere on the use of HS-SPME-GCMS techniques to reverse engineer the scented components of an artwork to procure new materials meant to reproduce its intended scent.

The goal of this project was to identify the composition of eight bags of powders and handmade beads in order to faithfully recreate House of Hope, including its aroma, at MoMA. Visual assessment of the powders and beads was accompanied by the use of ATR-FTIR and HS-SPME-GCMS to identify the composition of both sets of materials. Due to the incomplete documentation of the artwork before it entered the collection at MoMA, this information was deemed crucial to MoMA’s stewardship of the artwork and enables its presentation to current and future audiences. This project highlights the success of applying analytical techniques familiar to culture heritage professionals in gaining an improved understanding of a unique, monumental installation artwork.

2. Materials and Methods

2.1. Materials

2.1.1. Control Spices

To confirm the identity of the herbs and spices in the bags of powders or beads, a series of control herbs and spices were analyzed using ATR-FTIR and HS-SPME-GCMS. Included in the investigation were andrographis, anise seed, black pepper, cardamom, chamomile, cinnamon, clove, coriander, cumin, eucalyptus, ginger, licorice root, “mint” (spearmint), East Indies nutmeg, West Indies nutmeg, peppermint, red sandalwood (Pterocarpus soyauxii), Indian red sandalwood (Pterocarpus santalinus), “sandalwood,” thyme, and turmeric. Spices indicated in quotation marks are identified as such because that was the identity written on the label of the spice jar. Spices were chosen based on their proposed use in one of the previous reports [8] or their known use in Asia. Control spices were purchased in New York City.

2.1.2. Unknown Powders

Eight bags of unknown powder were removed from the crates used to ship the art and analyzed. These bags were intended for use in painting the mural. Not enough powder was present to complete the mural in the galleries at MoMA. Analysis was requested to identify the composition of each powder so that more could be purchased.

2.1.3. Beads

More than 1385 strands of beads were shipped with the artwork. It was noted that the beads were not homogeneous in color across all strands, so five different color strands were selected for analysis to identify the spices and binder used to make them. The five beads were identified as black, red-orange, yellow-orange, red-brown, and yellow-brown, respectively.

2.1.4. Control Binders

To identify the binder used to make the beads, the following controls were analyzed using ATR-FTIR and HS-SPME-GCMS: pine honey (honeydew honey), honey I (flower honey), honey II (flower honey), agave syrup, corn syrup, glycerin, and maple syrup. The materials were purchased locally in New York except for the pine honey, which was purchased online from Amazon.

2.2. Methods

2.2.1. Visual Assessment

The unknown powders and beads were assessed by visual examination after weighing out 5 g of material and adding it to a 60 mL vial, then 10 mL of hot water was added to the vial to dissolve and disperse the solids. After sitting for 10 min, the color of the liquid and the texture of the solids at the bottom of each vial were both used to attempt to identify the composition of the powder or bead.

2.2.2. Attenuated Total Reflectance–Fourier-Transform Infrared Spectroscopy (ATR-FTIR)

ATR-FTIR was carried out using a Nicolet iS50-FTIR bench with a Nicolet Smart Orbit ATR module (Madison, WI, USA) and a diamond plate installed. A DTGS detector and Thermo Scientific OMNIC 9.0 software collected data at 4 cm⁻¹ resolution from 4000–600 cm⁻¹. A total of 64 scans were collected for each sample, and data were interpreted using Thermo Scientific OMNIC Specta 2.0 software.

Control and unknown powders were analyzed by depositing and compressing them onto the surface of the diamond plate using the pressure clamp. The beads were soft enough to be compressed by hand and small sections were pressure-clamped to the surface of the diamond plate. Droplets of the potential binders were deposited directly onto the diamond surface and analyzed as is. The accessory was cleaned with 70% ethanol between experiments.

2.2.3. Static Headspace Solid-Phase Microextraction Gas Chromatography–Mass Spectrometry (HS-SPME-GCMS)

Control spices, unknown powders, beads, and control binders were prepared for analysis using a Gerstel MultiPurpose Sampler (MPS) (Linthicum, MD, USA). This system allowed for samples to be incubated either in an oven (30–300 °C) or at room temperature to drive volatile organic compounds (VOCs) into the headspace of the vial. Preliminary experiments were conducted for all four material types to determine how much material needed to be weighed out, the appropriate incubation temperature, the ideal time to extract VOCs from the sample vials, and the best SPME fiber to use to optimize the signal in the mass spectrometer. Older materials required more material be added to a sample vial, higher incubation temperatures, and longer extraction times. A Supelco carboxen–divinylbenzene–polydimethylsiloxane (car–DVB–PDMS) fiber (Bellelonte, PA, USA) was used to analyze all materials [24].

Control spices were analyzed by adding 0.01 g of material to a 20 mL headspace vial, incubating for 1 min at room temperature, and sampling for 1 min. For the unknown powders, 0.01 g of material was added to a 20 mL headspace sample vial, incubated for 5 min at 60 °C, and sampled for 5 min. To analyze the beads, 0.1 g of material was weighed out into a 20 mL headspace vial, incubated for 20 min at 60 °C, and sampled for 20 min. For the control binders, 1 mL of liquid was pipetted into the 20 mL vial, incubated for 5 min at 60 °C, and sampled for 5 min.

Automated injections were administered by the Gerstel MPS arm into an Agilent multi-mode inlet (MMI) attached to an Agilent 8890 Gas Chromatograph and Agilent 5977B (Santa Clara, CA, USA) single quadrupole mass spectrometer. An Agilent DB-5MS UI column (Santa Clara, CA, USA), 30 m long, 0.25 mm in diameter, film thickness of 0.50 μm was used to analyze the volatile compounds that were emitted by the control spices, unknown powders, beads, and control binders. Data were collected using Agilent’s Masshunter Workstation 10.0 software. Samples were held at 35 °C for 1 min, then heated from 35–250 °C at 10 °C/min. Helium was the carrier gas and flowed at 1.2 mL/min. Masses analyzed ranged from 25 to 600 m/z. Data were deconvoluted and peaks were identified using Agilent’s Unknowns Analysis 10.2 software. Compounds matching a NIST20.0 library at 90% or higher were identified. Three blanks were run after each sample was analyzed to guarantee no carryover between experiments. The retention times of the unknown powders and beads were compared to those of the control spices to help confirm the assignment of any spice.

3. Results

3.1. Control Spices

Figure 3 shows the results from HS-SPME-GCMS analysis of a select group of control spices: cardamom, clove, ginger, licorice root, nutmeg, peppermint, thyme, and turmeric. Though often more than one peak was used to positively identify a spice, the marker compound peak, identified in

Table 1. Control spices analyzed with HS-SPME-GCMS, their identifying chemical markers using the analytical approach described here, the chemical formulae of the identifying chemical markers, and the GCMS elution times.

Spice Analyzed	Marker Compound(s) from HS-SPME-GCMS	Chemical Formula	GCMS Elution Time (Minutes)
Andrographis	1-Butanol	C4H10O	5.2
Anise seed	Anethole	C10H12O	16.1
Black pepper	Caryophyllene/d-Limonene	C15H24/C10H16	18.3/12.1
Cardamom	α-Terpinyl acetate	C12H20O2	17.0
Chamomile	Farnesene	C15H24	18.3
Cinnamon	Cinnamaldehyde	C9H8O	16.1
Clove	Eugenol	C10H12O2	17.1
Coriander	Linalool	C10H18O	13.1
Cumin	Cuminaldehyde	C10H12O	15.6
Eucalyptus	Eucalyptol	C10H18O	12.1
Ginger	α-Curcumene	C15H22	18.8
Licorice root	Isoborneol	C10H18O	14.4
“Mint” (spearmint)	Carvone	C10H12O	15.5
Nutmeg	Terpinen-4-ol	C10H18O	14.6
Peppermint	dl-Menthol	C10H20O	14.5
Red sandalwood (Pterocarpus soyauxii)	β-Eudesmol	C15H26O	21.3
Red sandalwood (Pterocarpus santalinus)	β-Eudesmol	C15H26O	21.3
“Sandalwood”	Amberonne	C16H26O	21.4
Thyme	Thymol	C10H14O	16.0
Turmeric	aR-Turmerone	C15H22	21.2

, needed to be present to positively identify that material in an unknown. The marker compound peak was either the most intense peak by area for some spices, including cardamom, nutmeg, peppermint, and thyme, or the most unique peak in the chromatogram, as seen for clove, ginger, licorice root, and turmeric. The control spices, marker identification peaks from HS-SPME-GCMS, chemical formulae, and elution times are summarized in Table 1.

Table 2. Index number, retention time, compound name, and presence (X) of that chemical in eight of the control spices analyzed using static HS-SPME-GC for cardamom, clove, ginger, licorice root, nutmeg, peppermint, thyme, and turmeric.

Index Number	Retention Time (min)	Compound	Cardamom	Clove	Ginger	Licorice	Nutmeg	Peppermint	Thyme	Turmeric
1	10.10	α-Thugene	X				X	X
2	11.17	β-Myrcene	X		X
3	11.60	α-Phellandrene	X		X		X	X
4	11.91	γ-Cymene				X	X		X
5	12.01	d-Limonene	X	X		X	X	X
6	12.09	β-Phellandrene					X		X
7	12.11	Eucalyptol						X
8	12.17	β-Ocimene	X
9	12.47	γ-Terpinene	X	X			X	X
10	13.06	Linalool	X	X	X		X	X
11	13.63	Fenchol			X
12	13.78	trans-2-Pinanol			X
13	14.08	Isopulegol						X
14	14.17	(+)-2-Bornanone			X
15	14.23	l-Menthone						X
16	14.41	Isoborneol			X	X
17	14.53	endo-Borneol			X	X
18	14.56	dl-Menthol						X
19	14.61	Terpinen-4-ol	X	X			X	X
20	14.80	α-Terpineol	X		X		X
21	15.40	Menthyl formate						X
22	15.43	Linalyl acetate	X				X
23	15.50	Pulegone						X
24	15.56	(−)-Carvone						X
25	16.06	Thymol							X
26	16.13	(−)-Neomenthyl acetate						X
27	16.15	Anethole				X
28	16.16	(+)-Borneol acetate			X		X
29	16.28	Safrole					X
30	16.99	α-Terpinyl acetate	X		X		X
31	17.10	α-Cubebene			X		X	X
32	17.14	Eugenol		X
33	17.59	Copaene		X	X	X	X			X
34	18.29	Caryophyllene	X	X			X	X		X
35	18.80	α-Curcumene			X	X				X
36	19.04	Zingiberene			X
37	19.20	α-Muurolene				X	X
38	19.21	β-Bisabolene			X					X
39	19.55	Calamenene		X		X
40	21.17	aR-Turmerone								X
41	21.41	α-Santalol
42	21.66	Curlone								X

summarizes a list of other compounds noted in each of the eight control spices.

3.2. Unknown Powders

Table 3. Written labels, results from visual assessment, ATR-FTIR and HS-SPME-GCMS analyses, and conclusions regarding the composition of the eight bags of powder analyzed during this study.

Powder Bag Number	Label Written on Bag	Visual Assessment 1	FTIR Assessment	HS-SPME-GCMS Assessment	Conclusion
1	Talc	Thyme	Not talc; Carbohydrate	Not thyme	Unknown; not talc or thyme
2		Cinnamon	Black Pepper	Black Pepper	Black Pepper
3	Cinnamon	Clove	Clove	Clove	Clove
4	ผงขมิ้น	Turmeric	Carbohydrate	Turmeric	Turmeric
5	Cardimum [sic]	Ginger	Carbohydrate	White Sandalwood 1	White Sandalwood 1
6			Black Pepper	Black Pepper	Black Pepper
7	Talc	Thyme	Not talc; Carbohydrate	Not thyme	Unknown; not talc or thyme
8			Carbohydrate		Unknown

¹ Identified with assistance from staff at IFF (see Acknowledgments).

summarizes the results from the application of the three techniques to analyze the unknown powders: visual assessment, ATR-FTIR, and HS-SPME-GCMS. Visual assessment was hindered by fading of some of the powders and varying grinds of the powders. Still, two bags were successfully identified using that approach, bags 3 and 6, and two other bags were able to have their labels deemed incorrect, bags 1 and 7. ATR-FTIR (Figure 4) analysis was more useful, confirming the identity of three of the bags of powders, bags 2, 3 and 5, and confirmed the findings from visual assessment that bags 1 and 7.

HS-SPME-GCMS (Figure 5) was able to confirm the contents of five of the bags: bags 2, 3, 4, 5, and 6. Analysis of the powders proved to be challenging due to a strong disparity in the intensity of chromatographic peaks, with some very large and others small. Figure 5 shows that all eight bags of powder off-gassed large amounts of isoborneol (14.4 min) and menthol (14.5 min), characteristic of licorice root and peppermint, respectively, relative to the other peaks in the chromatogram. As each unknown powder was ostensibly composed of a single material, it was not clear why these two compounds were present in all of them. In the 2018 report written about the artwork, it was stated that the artwork had suffered a bug infestation [8], so natural insecticides such as peppermint and licorice root might have been added to the bags to prevent further insect attack [25,26]. Regardless of the source, it was concluded that these two compounds were a form of contamination and were disregarded when attempts to identify the composition of the powders were made. Detailed descriptions of the findings for each unknown powder are found below.

Bags 1 and 7 were both manually labeled as talc, but it was not clear who identified these bags or how that assessment was made. Upon visual assessment, it was concluded that the two bags contained the same material, but neither contained talc. Instead, it was suggested that the bags contained a ground leafy material, perhaps thyme. ATR-FTIR analysis confirmed that the powders were not talc; however, it could not identify the powders apart from them being carbohydrate-based. HS-SPME-GCMS spectra for these two bags were very weak, but did not show a peak for thymol at 16.0 min, indicating thyme was not present. Following analysis, the powders in bags 1 and 7 remained unidentified.

Neither bag 2 nor bag 6 was labeled. Upon visual assessment of bag 2, it was found it to be a finely ground brownish powder, and it was proposed that the bag contained cinnamon. By contrast, bag 6 was composed of a speckled material with a sand-like texture, and no assessment was made. ATR-FTIR showed that both powders contained piperidine (3014, 2941, 2922, 1636, 1612, 1583, 1495, 1447, 1371, 1319, 1226, 1195, 1136, 996, 931, 846, 831, 805, 787, 717, and 701 cm⁻¹), which is associated with black pepper. Though the spectra of the two bags were slightly different, this was attributed to the difference in the grind of the pepper.

HS-SPME-GCMS of the contents of bag 2 was relatively straightforward. Excluding the contamination peaks, the largest peak in the chromatogram was caryophyllene (18.3 min), a marker compound for black pepper. For bag 6, a eugenol peak (17.1 min) was larger than that of the caryophyllene. However, after combining these results with those from ATR-FTIR, it was concluded that bag 6 contained black pepper and that the eugenol peak was due to the close contact of bag 6 with bag 3 during storage.

Bag 3 was labeled as cinnamon, but there was unanimous agreement among all three techniques that the bag contained clove. This conclusion was based on the odor and color, a comparison of the FTIR spectrum of the powder from the bag to the spectrum of the clove control powder (2922, 2851, 1711, 1686, 1636, 1604, 1513, 1450, 1430, 1363, 1319, 1267, 1232, 1205, 1149, 1120, 1030, 913, 849, 815, 794, and 744 cm⁻¹), and an enormous eugenol peak (17.1 min) observed in the GCMS data (Figure 5c).

Bag 4 was labeled in Thai, ผงขมิ้น, which translates to turmeric in English. Visual assessment of the yellow powder confirmed this identification. The ATR-FTIR spectrum was too similar to other materials to confidently identify the powder as turmeric, and the most confident assessment that could be made was that it was carbohydrate-based. HS-SPME-GCMS showed the telltale turmerone peak at 21.2 min, confirming the powder was turmeric.

Bag 5 proved to be the most challenging bag to identify. Labeled “cardimum” [sic], it was assessed visually to be ginger. ATR-FTIR was not able to provide more specific information than that the powder was carbohydrate-based. HS-SPME-GCMS showed a small peak at 21.4 min, and the powder was identified as α-santalol. This is a marker compound for white sandalwood (Santalum album). This aromatic wood is increasingly difficult to purchase due to overharvesting in its native India [27,28]. Counterfeit products proclaiming to be sandalwood are rampant. As a result, standards have been developed for quality control purposes to confirm whether a product is truly sandalwood [29]. Three materials were purchased as control sandalwoods for this study: one labeled “sandalwood” and two labeled “red sandalwood.” The first powder contained amberonne (21.4 min), a compound used to create a woody scent [30], and patchouli hexanol (15.2 min), another chemical known to emit a woody odor, but lacked an α-santalol peak. Hence, this material was not sandalwood. The marker compound for both materials identified as “red sandalwood” was β-eudesmol (21.3 min), a compound affiliated with the Pterocarpus genus of trees [31] and not the Santalum genus (Santalum album). These three controls proved useless for identifying the powder. Experts at International Fragrances and Flavors, Inc. (IFF) confirmed that the powder was white sandalwood.

Bag 8 could not be identified using any of the techniques applied here. However, the brown color of the powder, coupled with the carbohydrate-based assessment from ATR-FTIR and a peak from GCMS data at 19.8 min, identified it as hedycaryol, a marker for leaf degradation [32], leading to the conclusion that the powder was an aged herb that had lost its odor over time.

3.3. Beads

An attempt was made to use visual assessment, ATR-FTIR, and HS-SPME-GCMS to identify the composition of the five bead types studied here. However, it quickly became clear that the beads were complex mixtures of myriad materials. Visual assessment was thus not used. Figure 6 shows the ATR-FTIR of the beads. It can be seen that the spectra for all five beads were similar to one another, with each showing broad -OH stretching around 3300 cm⁻¹, C-H stretching between 2800 and 3000 cm⁻¹, and C-O stretching between 1060 and 1170 cm⁻¹, indicating that they were carbohydrate-based. Library matches were poor, further implying that the beads were mixtures and not pure materials, and it was impossible to identify any particular spice or herb from this data. These results highlighted the usefulness and necessity of the SPME technique for analyzing mixtures.

When sniffed, each of the five colored beads had a unique odor. Hence, it was hoped that VOCs emanating from the beads could be used to identify their composition. Figure 7 shows the HS-SPME-GCMS chromatograms for all five bead types This technique permitted identification of some—possibly all—of the spices used to make the beads as well as the binder.

Table 4. The five bead types studied here, the spices identified in each bead using HS-SPME-GCMS, and chemical markers not positively identified.

Bead	Herbs and Spices Positively Identified [Chemical, (Elution Time [min])]	Unassigned Compounds [Chemical, (Elution Time [min])]
Black	Licorice root (14.4) Peppermint (14.5) Nutmeg (14.6) Thyme (16.2) Cardamom (17.0) Clove (17.1) Ginger (18.8) Turmeric (21.2)	γ-cymene (11.9) Maltol (13.4) β-bisabolene (19.2)
Red-orange	Licorice root (14.4) Peppermint (14.5) Nutmeg (14.6) Ginger (18.8) Turmeric (21.2)	Maltol (13.4) β-bisabolene (19.2)
Yellow-orange	Licorice root (14.4) Peppermint (14.5) Nutmeg (14.6) Thyme (16.2) Clove (17.1) Ginger (18.8) Turmeric (21.2)	Maltol (13.4)
Red-brown	Peppermint (14.5) Nutmeg (14.6) Ginger (18.8) Turmeric (21.2)
Yellow-brown	Peppermint (14.5) Nutmeg (14.6) Thyme (16.2) Ginger (18.8) Turmeric (21.2)	Maltol (13.4) β-bisabolene (19.2)

summarizes the findings from HS-SPME-GCMS analysis.

The black beads showed the presence of isoborneol, or licorice root (14.4 min); menthol, or peppermint (14.5 min); terpinen-4-ol, or nutmeg (14.6 min); thymol, or thyme (16.2 min); α-terpinyl acetate, or cardamom (17.0 min); eugenol, or clove (17.1 min); α-curcumene, or ginger (18.8 min); and turmerone, or turmeric (21.2 min).

The red-orange beads showed the presence of the following spices: licorice root (isoborneol), peppermint (menthol), nutmeg (terpinen-4-ol), ginger (α-curcumene), and turmeric (turmerone). Compounds that were not affiliated with a spice analyzed during this study included maltol and β-bisabolene. The yellow-orange beads were found to contain licorice root (isoborneol), peppermint (menthol), nutmeg (terpinen-4-ol), thyme (thymol), clove (eugenol), ginger (α-curcumene), and maltol. The red-brown beads were found to contain the fewest spices, including peppermint (menthol), nutmeg (terpinen-4-ol), ginger (α-curcumene), and turmeric (turmerone). Finally, the yellow-brown beads were found to contain peppermint (menthol), nutmeg (terpinen-4-ol), thyme (thymol), ginger (α-curcumene), turmeric (turmerone), maltol, and β-bisabolene. While it was not clear why different mixtures were used when making the beads and it was not possible to determine quantitatively the ratio of spices used to make each color of bead, the analysis showed that the different combinations led to beads with different color and scent profiles. When inevitably it becomes necessary to make additional beads in the future, these combinations can be used as a starting point to approximate the original beads’ color and scent.

Several peaks were affiliated with more than one spice (Table 2), and therefore were not assigned to a specific one. For instance, γ-cymene (11.9 min), maltol (13.4 min), and β-bisabolene (19.2 min) were noted. γ-cymene was observed in several of the control spices studied here, including andrographis, black pepper, licorice root, nutmeg, and thyme, but since it was not the identifying marker compound, it was not assigned to any particular spice. Maltol was found in any of the control spices analyzed here, but is found in spices such as ginseng [33]; further work with a broader variety of control spices may facilitate identification. β-bisabolene was noted in both ginger and turmeric, but was too small to be considered as a marker compound.

3.4. Binders

Figure 8 shows the ATR-FTIR spectra of the seven potential binders analyzed in this study. It was reported that Boonma used honey to make the beads [7], so two types were analyzed: two made from flower nectar, Honey I and Honey II, and one made from “honeydew”, or aphid secretions: pine honey. Several other potential binders, including agave, maple syrup, corn syrup, and glycerin were added to see if they could be differentiated from one another and whether any compared well with the spectra of the beads. Figure 8 highlights that the spectra for these liquids had similarities to each other, though there were distinct differences in the spectral region between 1500 and 750 cm⁻¹, which was associated with the absorption of monosaccharides (e.g., fructose and glucose) and disaccharides (e.g., sucrose) [34]. Apart from differences in the peak profile between 1100 and 1050 cm⁻¹, attributed to the C-O stretching vibration in carbohydrates indicating varied sugar composition [35,36,37], the vibration area of 950–750 cm⁻¹, associated with the anomeric region of carbohydrates, can be used to differentiate between certain natural sweeteners [38]. The three honey types and agave syrup had characteristic peaks at 920, 870, 820, and 780 cm⁻¹, which was attributed to C-H and C-C deformation vibrations in carbohydrates [38]. These four peaks were absent in the corn syrup, glycerin, and maple syrup controls. The five bead types showed all four of these characteristic peaks, indicating that the binder was likely a honey or agave syrup. However, due to the inclusion of other carbohydrate-based materials within the beads (i.e., herbs and spices), ATR-FTIR could not determine specifically which material was used as a binder.

Figure 9 shows the HS-SPME-GCMS spectra for potential binders. A comparison between the chromatograms from the beads with Honey I and Honey II led to the conclusion that nectar honey was not used to make the beads: the largest peak by area for Honey I was at 12.7 min, which was interpreted to be trans-linalool oxide, a common honey compound [39] that was not seen in the beads. Honey II also showed the presence of trans-linalool oxide; however, its largest peak by area at 14.5 min was interpreted to be endo-borneol, which was likewise not observed in the chromatograms for the beads. VOCs from the agave, corn, and maple syrups were not similar to the VOCs observed in the beads either.

Chromatograms showed that all the beads contained glycerin (11.4 min). This initially led to the conclusion that the beads were bound with glycerin. However, the control glycerin chromatogram only exhibited a large glycerin peak, and the remaining peaks in the chromatogram, at 13.3, 15.9, and 18.2 min, belonged to siloxanes from the GC column. By contrast, the beads showed the presence of other compounds, including esters 3-hydroxy-22,4-trimethylpentyl isobutyrate and 2,2,4-trimethyl-1,3 pentanediol diisobutyrate, at 17.1 and 17.4 min, respectively, that were not seen in the glycerin and were not associated with any spice.

A comparison of the bead chromatograms with pine honey was conducted next. This honey is different from nectar honey, as bees are prevented from feeding on flowers. Instead, they are sequestered in pine forests, where they feed off the “honeydew,” or excrement of tree aphids [40,41,42]. When analyzed using HS-SPME-GCMS, the chromatogram of pine honey showed both the glycerin peak as well as the two esters. It was concluded that the beads were likely made using pine honey as the binder.

4. Discussion

House of Hope is a complex artwork containing an ephemeral mural that is recreated each time it is displayed. The more permanent elements are the metal ceiling grid, the stools, and the strands of bead, which are rehung at each installation. The combined scent, central to the multi-sensory experience of the piece, works together to form the whole. The conflict of combining one component made using fresh materials with one that has continuously aged over the last 28 years posed a dilemma to the team at MoMA, namely, how to construct the artwork in a way that respected the intention of the artist. The analytical effort completed here took on greater urgency and importance when it became clear that communications with the artist’s assistant and archival research at the artist’s archive in Bangkok would not be realized in time for the museum’s planned display of the work in late 2023.

The unassumingly simple question of what this artwork is made of and how it should smell turned out to be difficult to answer. The goal became the use of chemical analysis to identify the powders and beads and then use that information to approximate the intended scent of the artwork. Results from the analysis of the powders led to the inclusion of clove and turmeric powders in the 2023 mural component of the installation. Licorice root and peppermint, found in the beads, were also incorporated into the mural, both for their scent and their color profile. Results related to the composition of the beads were written down and will be instrumental when additional beads need to be fabricated in the future.

The analysis also put into perspective the challenges associated with visual and scent analysis of aged scented materials. For instance, when experts from International Flavors & Fragrances, Inc. (New York, NY, USA) were asked to assess the powders based on their appearance, e.g., bag 5, its color led to the conclusion that the powder was ginger because it had no scent. This assessment was generally accepted until HS-SPME-GCMS was completed and it was revealed to be sandalwood. The age of the material also impacted the interpretation of the HS-SPME-GCMS data on the beads. At first, the myriad small peaks, which were about 1% the size of the most intense turmeric peak, were ignored, and only the larger peaks were cataloged. However, with the realization that the beads were between 25 and 30 years old and the scent profiles had diminished with time, the smaller peaks were re-examined, leading to the final assessment of the composition of the beads.

Though the analyses successfully identified specific spices in the powders and beads, the analysis is only partially complete. Further work, including the analysis of more controls, might allow for the remaining unidentified components to be identified. The volatiles in some components may also have largely off-gassed over the decades [43,44], and it may be that their identity will never be revealed.

5. Conclusions

The goal of this work was to facilitate the installation of Montien Boonma’s House of Hope by identifying the spices used to make the artwork. The spices arrived in two formats, as bags of unknown powder and as strands of hand-strung beads. It was inferred that the bags were intended to be used to create a mural that is painted on the walls surrounding the structure at the center of the piece. It was not possible to discern definitively whether the mural and beads were intended to be made using identical materials. This project provided an opportunity to explore the applicability of HS-SPME-GCMS to detect aromas in art objects and explore its advantages and limitations.

Though partly hindered due to the age of the materials, five of the powders and myriad spices in the beads were successfully identified. The binder used to make the beads was also determined. Visual assessment and ATR-FTIR were used to examine the unknown powders and beads in their solid state, while HS-SPME-GCMS was used to identify spices based on VOCs identified in the chromatograms and comparison with controls also collected here. The powders identified included black pepper, clove, turmeric, and white sandalwood. The remaining powders were found to be comprised of carbohydrates. A summary of the spices identified in the various beads included cardamom, clove, ginger, licorice root, nutmeg, peppermint, thyme, and turmeric. Potential future work includes identifying the composition of any remaining unidentified powders using alternate techniques, including pyrolysis GCMS, determining if it is possible to calculate the ratio of the spices relative to one another, and researching the artist’s archive to determine whether other spices were reported to have been used by the artist. As no information had previously been recorded by previous gallerists about the materials used to create the mural or beads, the results found here aided the 2023–2024 installation of the work at MoMA and will inform future installations when new murals or strands of beads must be made.