Indoor Air Quality in a Museum Storage Room: Conservation Issues Induced in Plastic Objects

Abstract

This study focuses on assessing the indoor air quality in a storage room (SR) belonging to Museo Nazionale Scienza e Tecnologia Leonardo da Vinci in Milan (MUST), covering pollutants originating from outdoor sources and emissions from historical plastic objects made from cellulose acetate (CA), cellulose nitrate (CN), and urea–formaldehyde (UF) stored in metal cabinets. The concentrations of SO₂ (sulphur dioxide), NO₂ (nitrogen dioxide), NOx (nitrogen oxides), HONO (nitrous acid), HNO₃ (nitric acid), O₃ (ozone), NH₃ (ammonia), CH₃COOH (acetic acid), and HCOOH (formic acid) were determined. The concentrations of SO₂, O₃, and NOx measured inside the metal cabinets were consistently lower compared to the other sampling sites. This result was expected due to their reactivity and the lack of internal sources. The SR and metal cabinets showed similar concentrations of NO and NO₂, except for CA, where a high NO concentration was detected. The interaction between the CA surfaces and NO₂ altered the distribution of NO and NO₂, leading to a significant increase in NO. The presence of HNO₃ potentially led to the formation of ammonium nitrate, as confirmed by ER-FTIR measurements. High levels of HONO and HNO₃ in CN and NH₃ in the UF indicate object deterioration, while elevated concentrations of CH₃COOH in CA and HCOOH in the SR suggest specific degradation pathways for cellulose acetate and other organic materials, respectively. These results could direct conservators towards the most appropriate practical actions.

1. Introduction

Maintaining optimal museum indoor air quality (IAQ) is a priority for safeguarding delicate artworks and historical items from deterioration caused by pollutants, UV irradiation, and fluctuations in humidity and temperature [1,2]. Moreover, IAQ directly influences the comfort and health of museum staff and visitors. Implementing effective ventilation systems, pollutant control measures, and monitoring protocols builds up the essential strategies for sustaining a healthy indoor environment, leading to both the preservation and enjoyment of cultural items. A museum’s indoor environment constitutes a complex system, where IAQ is influenced by several factors: gaseous pollutants entering from outside, emissions from stored artefacts and display cases housing objects [3], the presence of visitors [4], and routine maintenance activities within the museum. Furthermore, chemical reactions on indoor surfaces depend on numerous factors and can lead to the formation of new compounds that may either remain on the surface or evolve into the environment [5].

This study monitored the most harmful pollutants, generated both indoors and outdoors, that are generally dangerous for any collection; in this paper, a special focus on plastic objects was considered. These included oxidant species such as ozone O₃, which is harmful to a wide range of artistic materials, NO₂, SO₂ [6,7], and acidic species such as HNO₃ and HONO; the latter are well known for their special role in the decay mechanism of cellulose nitrate and acetate artefacts [7].

A general overview of plastic object decay patterns is also reported to connect visual observations of decay with possible generating mechanisms. Plastic objects have for several decades been considered new collectibles and valued as cultural assets. The Victoria and Albert Museum in London, the Smithsonian (DC), MOMA, and Cooper Hewitt in NYC conserve important collections of plastic objects, just to name a few. The history of plastic had its starting point in the middle of the 19th century, looking for an imitation of natural materials such as tortoise shells, ivory, and horn, which are close to depletion or are forbidden in order to safeguard the animal species providing those luxury materials. From that point on, some artificial polymers were produced, starting from cellulose (cellulose nitrate and cellulose acetate), and then synthetic polymers were used to produce a vast range of items [8]. Bakelite, Perspex, polyethylene (PE), Nylon, and polyethylene terephthalate (PET) entered the world of industrial production during the course of the 20th century. Therefore, collections of plastic objects demonstrate the production of designs as well as many contemporary art items. They were designed to be very durable, although they have grown to become an environmental issue due to their accumulation. In this respect, on the one hand, their degradation is desirable, which constitutes a topic of great scientific interest in the field of environmental studies [9]; on the other hand, their decay processes should be considered a crucial issue for plastic museum curators, who are interested in their conservation [10,11]. In another respect, durability lasting several decades does not ensure the conservation of this group of items for future generations. A scarce number of studies available in the literature have dealt with sustainability issues in an effort to study the durability of microplastics in various environments, where plastic fragments have been studied alongside prolonged contact with gaseous atmospheric components, water, marine salts, different temperatures, UV irradiation, and pH levels. This body of literature considers only polymeric materials comprising current plastic production (mainly PP, PET, and both LDPE and HDPE); on the contrary, materials constituting historical plastic objects (for example, cellulose acetate and cellulose nitrate and Bakelite) are ignored [12,13].

Some plastic decay processes involve the transport phenomena of plasticisers, which move from the bulk of the material and exudate on the outer surface in the form of a viscous liquid; otherwise, gaseous components are released by some polymer decomposition reactions. These gaseous substances can be collected in a conservation environment, such as, for example, in closed-window cases or small metal cabinets where the items are on exhibition or stored. Over time, the conservation environment becomes unfit for preserving specific items in the overall collection. Hence, this challenge requires a classification of plastic objects into general categories based on the polymer type, based on the knowledge of the overall decay mechanism, including gas pollutant evolution and the separation of objects considered “malignant” due to their negative impact on the rest of the collection and their subsequent required isolation in separate storage. By doing this, best conservation practice is achieved, leading to enhanced plastic object durability.

This issue was considered in a recent project funded by the Italian Ministry of University and Research and involving the Institute of Heritage Science of the National Research Council (ISPC CNR Milan Unit), Museo Nazionale Scienza e Tecnologia Leonardo da Vinci di Milano—MUST (National Museum of Science and Technology Leonardo da Vinci in Milan), Cesmar7-Centro per lo Studio dei Materiali per il Restauro (Research Centre for Conservation Materials), and CASVA-Gli Archivi del Progetto del Comune di Milano; it was based on multipurpose tasks focused on the MUST plastic objects collection.

The main objective of the project “Storie di Plastica” (PANN20_00738) can be summarised as follows:

• Classify objects from a compositional point of view, organising their storage into homogeneous groups according to the molecular structure of their primary polymeric main component.
• Identify the decay patterns present on the objects and develop a specialised illustrated glossary containing the terminology related to the decay of plastic materials.
• Provide museum conservators with the best strategy for preserving the objects. These guidelines aim to prevent the release of pollutants caused by the decay mechanisms of the objects, which could negatively impact their proper conservation in the storage cabinets.

The MUST collection, named “Montedison”, consists of approximately 600 objects produced in a time span from the early 1900s to the 1950s. The collection includes combs, mirrors, boxes, handbags, frames, toys, and various other small items. Currently, the objects are conserved in a metal/glass cabinet in a storage room located about 9 metres below street level, and it is not directly connected to the outdoor environment, but it is linked to other storage rooms housing the museum’s study collections and the stairwell.

The objects are made from both semi-synthetic and synthetic plastics, cellulose nitrate (CN), cellulose acetate (CA), casein–formaldehyde (CS), phenol–formaldehyde (PF Bakelite and Cast Phenolics), amino plastics (UF urea- and thio-urea–formaldehyde), polystyrene (PS) and poly-methyl-methacrylate (PMMA).

The first phase of the project included the following steps:

• Identification of constituent materials and revision of the museum’s existing inventory;
• Environmental assessment of the storage condition;
• Detection of actively degrading objects, using indicators such as odour (e.g., strong acetic or acidic smell), observation of plasticisers separation and crystallisation, signs of mechanical damages, and the placement of self-prepared bromocresol green strips indicators prepared in-house near the objects.

2. Aims of the Paper

This paper reports some of the main results of the project “Storie di Plastica”, with a focus on investigating indoor air quality within a museum environment, specifically in a storage room. In particular, the focus is regarding the contribution to emissions of gaseous pollutants coming from historical plastic objects, conserved in metal cabinets (previously mentioned) that house the “Montedison” plastic objects collection. Additionally, the potential interactions between the gaseous pollutants emitted by these objects and those infiltrating from external sources were thoroughly investigated to assess their combined impact on the overall air quality. The decision to monitor those cabinets was informed by well-documented evidence that materials such as cellulose acetate, cellulose nitrate, and urea–formaldehyde release harmful substances, including acetic acid, nitric acid, and ammonia [14,15,16,17].

To achieve this goal, the authors addressed the following research questions:

• Does the emission of pollutants from the objects affect the overall air quality of the storage rooms?
• Based on the measurements of indoor air quality in both the storage room and the metal cabinets, can these conditions be deemed optimal for conserving plastic objects?
• Can the findings guide museum conservators in isolating high-risk malignant objects in separate cases? Would this approach represent the most effective preservation strategy [18]?

3. Materials and Methods

This study was conducted in the storage room of the MUST Museum housing the plastic object collection. Located in central Milan, the museum is one’s of Italy most prominent science and technology institutions, ranging from a submarine to reconstructions of several of the machines imagined by Leonardo Da Vinci.

The identification of materials was first performed via observation and examination (naked eye, digital microscope), merging details deriving from sensation (odour, appearance, weight, mechanical properties, and colour). The second step consisted of a comparison with the literature (images of the object, trademarks, and brands [19,20,21] and online resources (MODIP, PHS) [22,23]. Limited to some types of plastic (CN and CA), it was possible to confirm the previous identification with simple spot tests. This approach proved particularly useful for identifying two of the most hazardous plastics, thus reducing the number of instrumental analyses and overall costs, and improving project sustainability. The objects that could not be identified with the previous methods were investigated using FTIR spectroscopy. This initial phase focused on separating malignant plastics (CN and CA) into homogeneous groups. Subsequently, the objects under the active decay process were identified, to implement specific tailored preventive conservation strategies, using specific absorbers and setting the environmental parameters.

The biological attacks were investigated on different plastic objects sampling with transparent Fungi-Tape^®, sterile swabs, and cellulose nitrate membrane [24]. Swabs and membranes were processed through cultivation techniques, while the Fungi-Tape^® were stained with trypan blue and observed under high magnification using an optical microscope: these analyses revealed the presence of a dominant Aspergillus species in all samples. Biological activity was determined using a luminescence assay, a portable bioluminometer (ENSURE–Hygiena) and specific swabs (Ultrasnap–Hygienia) containing luciferin–luciferase substrate complex for ATP determination, were employed. The results indicated significant biological activity in most cases.

To evaluate the air quality risk related to the conservation of the stored objects, a 3-week monitoring campaign was carried out during the summer of July 2022.

The campaign was aimed at the determination of SO₂, NO₂, NO_x, HONO, HNO₃, O₃, NH₃, CH₃COOH, and HCOOH.

Air quality assessments were performed in the following three specific locations:

• Within metal cabinets housing plastic objects (Figure 1a);
• In the storage room (Figure 1b);
• In the surrounding environment adjacent to the storage area.

Each metal cabinet contained plastic items composed of different polymeric materials (

Table 1. Indoor pollutant measurement points in the MUST.

Site	Description
MCA	Metal cabinet containing objects made of cellulose acetate
MCN	Metal cabinet containing objects made of cellulose nitrate
MUF	Metal cabinet containing objects made of urea–formaldehyde
SR	Storage room (hosting the metal cabinet)
OSR	A room connecting the storage room with the rest of the museum’s exhibition on the ground floor, through a door

). This musealisation, before the project, did not consider a proper conservation practice, but it was addressed at optimising the use of the space.

During the sampling period, six types of passive samplers were used. For each type, four replica samplers were deployed simultaneously, with one sampler kept sealed to serve as a field blank.

The sampling procedure began by removing the polyethylene cap from each sampler and replacing it with a stainless steel grid, which remained in place throughout the sampling period. The samplers were placed directly inside the same metal cabinets housing the plastic objects, while SR and OSR samplers were mounted at a height of 2 m using a passive sampler rack. As previously mentioned, the storage room is located 9 m below ground level, and it is equipped with air conditioning and an HVAC system (heating, ventilation) to control temperature and relative humidity. The room has a volume of approximately 1700 m³ and a surface-to-volume ratio of 0.73 m⁻¹, featuring a highly complex plan (see the Supplementary Materials). The lights are switched off most of the time when the room is not in use. During the campaign, the lights were switched on for a total of 66 h. Temperature (T °C) and relative humidity (RH%) data, including average, maximum, and minimum values during the campaign and throughout the year are reported in

Table 2. Temperature and relative humidity during the monitoring campaign and the entire year.

	Monitoring Campaign—July 2022		Whole Year—2022
	T °C	RH%	T °C	RH%
Min	20.7	45	15.6	37
Max	21.3	56	25.8	66
Average	21.0	52	20.4	55

; these measurements indicate that T °C and RH% remained relatively stable over the course of the year.

The area outside the storage room is connected to the museum via a staircase located approximately 15 m from one of the museum’s entrances.

The two metal cabinets containing objects made of cellulose acetate and nitrate have a combined volume of 0.50 m³ and a surface-to-volume ratio of 7.8 m⁻¹. In comparison, the metal cabinet holding objects made of urea–formaldehyde has a volume of 0.41 m³ and a surface-to-volume ratio of 12 m⁻¹.

To prevent the accumulation of pollutants emitted by the plastic objects, the metal cabinets were left partially open with an opening area of approximately 0.03 m².

The monitoring campaign employed a passive sampling technique [25,26,27,28] based on the diffusion of gas molecules, as described by Fick’s first law [29]. Pollutants are absorbed onto a specific substrate, typically an impregnated filter. The equivalent sampling rate F can be calculated using the following equation F = D(T) A/L, where D(T) is the diffusion coefficient of the species (cm² s⁻¹), A is the collecting surface area (cm²), and L is the length of the diffusion path (cm). Three passive samplers, designed for the collection of samples from each species, were exposed, plus one sampler, which was kept closed and used as a field blank.

All the filters were extracted in water (in the case of SO₂, with the addition of H₂O₂ to ensure complete oxidation to SO₄²⁻ [30]) and analysed using ion chromatography (IC). The analyses were conducted using an ICS1000 instrument from Thermo-Fisher (Sunnyvale, CA). Ammonia content was determined using an ION PAC CS12A 4 mm column, a pre-column ION PAC CG12A 4 mm, and a CSRS-ULTRA 4 mm suppressor. Elution was performed with methane sulfonic acid in isocratic mode, utilising a concentration of 20 mM and a flow rate of 1.2 mL/min. Inorganic species were identified using an AS11 4 mm column, a pre-column AG11 4 mm, and an ASRS-ULTRA 4 mm suppressor. Elution was achieved with potassium hydroxide, generated by ECG40 EGC II KOH from Thermo-Fisher in an isocratic mode with a 10 mM concentration. The analysis of organic acids was carried out utilising an AS11-HC 4 mm column and a pre-column AG11-HC 4 under an isocratic mode with a KOH concentration of 4 mM and a flow rate of 1.50 mL/min. Each analysis was followed by regeneration with KOH at 50 mM.

Each sample was also analysed three times, yielding a standard deviation of less than 3%. The precision for each type of passive sampler, based on triplicate measurements, was determined to be below 5%.

The pollutant’s air concentration was calculated by averaging the three measurements obtained from each sampling site. The average values in the different sampling locations were then compared.

Additionally, the results were compared with the existing literature to assess the consistency and reliability of the measurements.

The identification of materials via observation was performed with a Celestron Handheld Digital Microscope Pro 5MP 20 and 200×. The plastic items were classified in broad groups using FTIR: Fourier Transform Infrared spectra were recorded on micro-samples collected under the objective of an optical microscope in diamond anvil cell in transmission mode by a Thermo Scientific Nicolet iS10 instrument, equipped with a Continuum microscope in the 4000–700 cm⁻¹ range, with a resolution of 4 cm⁻¹, collecting 64/128 scans. Absorption peaks have been assigned by comparing them to internal references and the IRUG database (www.irug.org).

External reflection FTIR (ER-FTIR) investigations were performed by an Alpha II spectrometer (Bruker Optics, Ettlingen, Germany) equipped with an external reflection module. The ER-FTIR spectra were acquired in the near- and mid-infrared region from 8000 to 400 cm⁻¹, at a resolution of 4 cm⁻¹ by summing 128 scans. The ER-FTIR investigations were performed non-invasively and contactless (working distance of about 2 cm) with a circular investigated area of 6 mm diameter on the top of samples and without any sample preparation. The acquired data collection and qualitative analysis were processed and qualitatively analysed using OPUS 8.5 spectroscopy software. The reflectance spectra are presented without any spectral correction.

4. Results and Discussion

4.1. Classifying Plastic Materials Using FTIR Spectra

A set of 35 plastic objects were analysed using FTIR in a diamond anvil cell (Table 3). The objects have been chosen according to the following criteria:

• Typology;
• Representativeness;
• State of conservation.

For instance, a homogeneous group of toilette objects was represented by a single comb or a mirror frame.

ER-FTIR measurements were carried out to generate a series of references intended for use in future studies. Those spectra were compared to the traditional ones recorded in transmission mode.

The classification resulted in the following grouping:

Table 3. Typology of polymeric materials in the analysed set of plastic objects.

Plastic Material Type		Number of Items
Cellulose nitrate	CN	9
Cellulose acetate	CA	8
Casein	Protein	5
Polystyrene	PS	4
Acrylics	PMMA	3
Phenolics	PF	2
Urea–Formaldehyde	UF	2
Polyethylene	PE	1
Polyethylene terephthalate	PET	1

Table 3. Typology of polymeric materials in the analysed set of plastic objects.

Plastic Material Type		Number of Items
Cellulose nitrate	CN	9
Cellulose acetate	CA	8
Casein	Protein	5
Polystyrene	PS	4
Acrylics	PMMA	3
Phenolics	PF	2
Urea–Formaldehyde	UF	2
Polyethylene	PE	1
Polyethylene terephthalate	PET	1

At first glance, it was possible to note that the various groups did not align the current production trends of plastic objects, where PE, PP polypropylene, and PET account for approximately 45% of the manufactured materials [8]; notably, some significant plastic material, for example, in the building industry, such as PVC, were entirely absent; conversely, some historical plastic materials such as cellulose derivates, casein, or urea–formaldehyde were prominent represented [18]. For one object, a finishing coating has been macroscopically individuated and then analysed, resulting in a PET bulk coated with cellulose nitrate.

In some objects, the decay phenomena resulted in the formation of surface exudate or some whitish crystals, which could be mechanically separated; FTIR spectra revealed the presence of some additives such as tri-phenyl-phosphate (in correspondence with a CA object) or phthalates originated from an acrylic substrate.

Several of the collected spectra demonstrated excellent quality, with a satisfactory signal-to-noise ratio (for example, those collected on PS objects), enabling precise identification and accurate assignment of the absorption peaks.

For some objects, only a broad categorisation was possible: the spectra assigned to casein objects, in fact, displayed the absorption pattern of a generic protein.

As previously mentioned, CN and CA are classified as malignant plastics [31]; cellulose nitrate degrades to produce acidic and oxidising nitrogen oxide gases, which can seriously damage nearby or contacting objects; this could also combine with water, converting in HNO₃, entering the cracks and attacking the objects internally. This deterioration was accelerated by elevated temperatures, high relative humidity, and acidic environments. Cellulose nitrate was commonly plasticised with camphor (e.g., celluloid). Over time, camphor sublimates from the plastic, leading to shrinkage and increased brittleness. The shrinkage creates tension within the brittle plastic often resulting in severe cracking or crizzling. This issue tends to be so less severe in cellulose nitrate plasticised with less volatile phthalates (DIOP, DEP, and DEHP).

CN artefacts should be stored at low temperatures (T = 2–5 °C) to slow hydrolysis reactions. Low humidity levels are also important and should be kept in the range of RH = 20–30%; lighting conditions should be limited to a maximum of max 50 lux, UV 75 W/lumen, eliminating the UV component from the sources as much as possible by using low UV component lamps or UV filters on the light sources.

The display case interior should be ventilated, to remove acidic and oxidising substances, avoiding forced ventilation as this would induce further loss of plasticisers and accelerate degradation; monitoring tools such as AD strips, Dancheck acid detection test strips, or self-prepared Cresol red strips can be employed to detect and manage acidic conditions effectively.

CA primarily degrades through acid hydrolysis, causing deacetylation, which cleaves pendant acetate groups from the cellulose polymer backbone, causing depolymerisation. A by-product of this degradation is the emission of acetic acid gas, which creates acidic surfaces on the plastic and acidic atmospheres within enclosed spaces. Depolymerisation reduces mechanical strength and fractures, leading to deformations and warpages. Acetic acid is a volatile gas that diffuses through the display or storage space and causes the corrosion of metals, or acidic-catalysed degradation of other materials and textiles when multi-materials objects are involved. Additives, especially plasticisers, migrate and may be lost, or are hydrolysed or oxidised to acidic compounds. This process leads to warpage, embrittlement, and fracture and the development of acidic and sticky surfaces, sometimes with surface deposits of plasticiser of flame retardant (like tri-phenyl-phosphate TPP) or acidic degradation products.

Similar preventive conservation parameters and measures should be adopted also for CA objects.

4.2. Indoor Air Quality

The average concentration of gaseous pollutants obtained during the monitoring campaign is presented in Figure 2. The NO concentration was obtained as the difference between NO_x and NO₂.

The concentration of indoor pollutants depends on several factors: the outdoor pollutants levels, the rate of exchange between indoor and outdoor space, the efficiency of indoor surfaces in eliminating pollutants, and the rate at which they are either generated or eliminated through deposition and chemical reactions within indoor spaces, as described by Weschler et al. [32]. The surface removal rate is quantified as the deposition velocity of a pollutant, relative to the surface-to-volume ratio of the room (S/V); an increase in the surface-to-volume ratio results in a reduction in pollutant concentration within the confined environment.

In the context of monitoring air quality in confined spaces, calculating the indoor-to-outdoor (I/O) ratio of the concentration of potentially harmful species is highly beneficial. The I/O ratio proves valuable in determining whether external pollutants are infiltrating the building or, conversely, if the materials themselves are emitting pollutants. When the ratio exceeds 1, the confined space is a source of pollution.

In our study, the primary focus was on evaluating emissions from plastic objects stored in metal cabinets and from items stored within the storage room, as well as on studying the interaction between indoor and outdoor sources of pollutants, the latter represented by OSR (see Table 1). Therefore, we did not measure the concentration of pollutants in the actual outdoor environment.

Following this approach, the metal cabinet was labelled as ‘IN’ and the storage room as ‘OUT’, to differentiate the emission of gaseous species from various types of plastic objects, inside the metal cabinet. The IN/OUT ratios have been calculated and reported in

Table 4. Indoor (metal cabinets)/outdoor (storage room) average concentration ratio for SO₂, NH₃, O₃, NO₂, NO_x, NO, HONO, HNO₃, HCOOH, and CH₃COOH.

	SO2	NOX	NO2	NO	HNO2	HNO3	O3	CH3COOH	HCOOH	NH3
CA/SR	0.59	1.0	0.056	3.4	2.2	0.19	0.62	1.3	0.18	0.44
CN/SR	0.64	0.73	0.90	0.36	2.1	1.8	0.63	0.27	0.22	1.0
UF/SR	0.93	0.9	0.74	1.2	1.5	0.42	0.55	0.29	0.38	1.7

. The OSR sampling site was selected to serve as an indicator of the overall indoor air quality within the museum. Moreover, this sampling point helped identify the potential presence of additional pollutant sources within the environment.

The metal cabinets and the storage rooms displayed quite different surface-to-volume ratios (MCA and MCN s/v = 7.8 m⁻¹; MUF s/v = 12 m⁻¹; SR s/v = 0.73 m⁻¹), resulting in possible variations for removing pollutants.

Moreover, the SR and metal cabinets had different air exchange rates, especially considering that the latter are kept partially closed.

SO₂, O₃, and NO_x concentrations measured in the metal cabinets were always lower with respect to the other sampling sites; this result was expected, given their reactivity and the absence of internal sources.

The average concentration of SO₂ across all the monitored sites was 1.7 µg/m³, which was relatively low. The highest concentration was recorded in OSR (2.2 µg/m³), while the lowest was at MCA (1.2 µg/m³). It is worthwhile to note that all measured concentrations were below the recommended levels in a museum environment (5 µg/m³) [33].

The average concentration of O₃ across all the measured sites was 3.1 µg/m³. The highest concentration was observed in OSR (5.5 µg/m³), while the lowest concentration was at MUF (2.0 µg/m³). Moreover, in this case, all measured spots remained below the recommended levels for the museum environment, which is 10 µg/m³ [34].

For total nitrogen oxide (NO_x), the average concentration was 16 µg/m³, with OSR showing the highest value (21 µg/m³) and MCN the lowest one (12 µg/m³). The average, maximum, and minimum concentrations of NO_x, NO₂, and NO are reported in

Table 5. Average, maximum, and minimum concentrations of NO_x, NO₂, and NO at the measurement points.

Pollutant	Average Concentration (µg/m3)	Maximum (µg/m3)	Minimum (µg/m3)
NOx	16	21 (OSR)	12 (MCN)
NO2	8.7	13 (OSR)	0.85 (MCA)
NO	7.4	17 (MCA)	1.8 (MCN)

.

OSR was the site with the highest concentration of outdoor-generated pollutants. This was because OSR is partially connected to one of the museum’s exhibition rooms, which is close to the secondary entrance, fronting a heavy vehicular traffic road; as a consequence, pollutants enter through the ventilation system and directly through the entrance.

I/O NO was > 1 both in MCA (3.4) and UF (1.2) (see Table 4). This very high value recorded in MCA was associated with a very low one for NO₂ (I/O NO₂ = 0.056), showing a change in the distribution of NO/NO₂. For example, NO showed an average concentration of 7.4 µg/m³ but exhibited a peak concentration in MCA equal to 17 µg/m³, while the minimum value was recorded in MCN (1.8 µg/m³). NO₂, with an average concentration of 8.7 µg/m³, exhibited a peak concentration in OSR (13 µg/m³), while the minimum value was observed in MCA (0.85 µg/m³).

The scientific literature on this topic is quite scarce. Notably, other studies highlighted a peculiar mechanism for several types of materials [35]: up to 15% of NO₂ was scavenged by the indoor surface and re-emitted as NO, whereas in our case, more than 96% of NO₂ (MCA) was scavenged.

A case study carried out at a museum in Kuala Lumpur [36] reported a decrease in NO₂ concentration, moving from the exhibition halls to the storage room: in the exhibition hall, NO₂ accounted for 52% of NO_x, while in the storage room, it was only 13%. Godoi et al. [37] published measurements taken inside a display case; although NO_x levels were not monitored, the concentration of NO₂ exhibited a decreasing trend as well. It decreases from 35 µg/m³ to 28 µg/m³, in the storage room, with a further dropping at 10 µg/m³ inside the display case.

It is well known that NO₂ is a precursor of HONO and HNO₃ through reaction (1) involving water absorbed on surfaces [38,39]:

2NO₂ (g) + H₂O/surface → HONO (aq) + H⁺ NO₃⁻ (aq)

(1)

Afterwards, HONO can evolve (2) or react further with NO₂ to form NO (3) while HNO₃ is adsorbed on the surface:

HONO (aq) ⇌ HONO (gas)

(2)

NO₂ (g) + HONO (aq) → H⁺ + NO₃⁻ + NO(g)

(3)

Reaction (3) increases the amount of nitric acid deposited on the surface. According to the literature, the amount of HNO₃ deposited on a surface can be evaluated by measuring HONO measurements [40]. Considering the reaction of NO₂ with the surface moisture film, the actual amount of deposited HNO₃ could also be much higher than expected.

Cellulose acetate-based materials are highly hygroscopic, leading to the formation of a water film, allowing the accumulation of pollutants at the interface, and facilitating the formation of new species [5].

The high water content and S/V ratio of the MCA cabinet (7.8 m⁻¹), even if generally correlated with lower pollutant concentrations, provide a larger surface area for reactions (3). This increased surface area could result in higher NO production, as cellulose acetate is known to readily absorb nitrogen dioxide [41].

The concentration of HONO in OSR was lower (1.5 µg/m³) compared to the other sampling location. This reduction was likely due to the photolysis reaction triggered by both sunlight and artificial light [42], as the OSR is connected to the outdoors and partially illuminated by artificial lighting.

The average concentration was 4.5 µg/m³, with a maximum value in MCA equal to 6.4 µg/m³, as expected from the previous considerations.

The average concentration of NH₃ was 12 µg/m³, with a minimum value of 4.9 µg/m³ in MCA and a maximum value of 19 µg/m³ in MUF, with a ratio UF/SR equal to 1.7. This value indicated an inner source of NH₃, as UF objects undergo a degradation process that releases NH₃ [17].

The lower concentration of ammonia and nitric acid in MCA could result from the reaction between these two gases forming ammonium nitrate in the aerosol phase, which then deposits on the surfaces [43] (4):

HNO₃ + NH₃ ⇌ NH₄NO₃

(4)

This hypothesis is further supported by the I/O ratios of nitric acid and ammonia in MCA, which were 0.19 and 0.44, respectively. In those cases, the low I/O ratios for both species suggested that the concentration of gaseous pollutants decreased because of the reaction (4).

It is well known that HNO₃ is adsorbed onto any surface with a lifetime of 30 minutes or less [44,45]. The highest HNO₃ concentration (5.9 µg/m³) was recorded in OSR, in correspondence with a low S/V ratio (0.73). Considering the MCN/SR ratios for HNO₃ (1.8) and HONO (2.1), these values confirmed the degradation of cellulose nitrate, leading to the formation of nitric and nitrous acid [14,46].

Regarding the organic acids, CH₃COOH, with an average concentration of 22 µg/m³, had the peak value in MCA (44 µg/m³). According to the literature [14], these elevated values were due to the degradation mechanism of objects made with cellulose acetate. There was also a comparable level of acid in the storage room, with an MCA/SR equal to 1.3. The release of acetic acid should have been due to the several wooden and plastic objects conserved in the SR. A high concentration of formic acid was detected in the SR (29 µg/m³), with the lowest value detected in the OSR. Considering these ratios, no inner sources of formic acid were assumed within the metal cabinets.

An ER-FTIR campaign was carried out on several plastic objects to gather a series of spectra for future studies. Those spectra were compared to the traditional ones recorded in transmission mode, with a summary reported in Table 3.

The spectra collected on the surface of the object labelled IGB12207 reveal interesting vibrational features; the object is a hexagonal bag, whose body is made of cellulose acetate coated with an acrylic resin. The body of the bag displays regions where the plastic material appears intact; in fact, both µFTIR and ER-FTIR collected in different regions of the bag show the typical vibrational pattern of cellulose acetate (ER-FTIR Figure 3).

On the contrary, some parts of the bag exhibited a severe decay pattern with the surface covered by white crystals. The ER-FTIR spectra collected from those regions revealed a very complex vibrational pattern (Figure 4). Indeed, the reflectance bands localised at about 1720 cm⁻¹ and 1695 cm⁻¹ could be attributed to the C = O stretching [47]. In particular, the former was likely belonging to an ester group, while the latter was likely due to an acid, such as the one of a carboxylic or dicarboxylic acid, compatible with the presence of phthalic, formic, or benzoic carboxylic acids. At the same time, the characteristic vibrational bands observed around about 1280 cm⁻¹ and in the region between 1160 and 960 cm⁻¹ were consistent with the presence of a nitrate group, and, in particular, with cellulose nitrate. The presence of marker reflectance bands in the N-O stretching region suggests the potential formation of ammonium nitrate; however, its presence could not be unequivocally confirmed. This uncertainty arises from the very weak intensity of NH stretching vibrations in ER-FTIR measurements, particularly when ammonium nitrate is potentially mixed with other compounds.

The mixture of newly formed decay products identified by ER-FTIR suggested a complex interaction between the objects and their surrounding environment. This interaction likely involved contributions from the degradation of plastic materials additives (e.g., phthalates transformed into acids and migrated to the surface) and possible environmental degradation. The process appeared to involve the deposition of compounds that altered the surface composition of the polymer from cellulose acetate to cellulose nitrate.

5. Conclusions

This study examined the complexity and variability of indoor pollutant concentrations in confined spaces, focusing specifically on the influence of emissions from stored plastic objects. It analysed how pollutants coming from a storage room, treated as an external source (OUT), interact with the indoor environment (metal cabinet, IN).

Calculating the indoor-to-outdoor (I/O) ratios was crucial for identifying the sources of pollutants, even in the absence of a real outdoor reference.

All the measured values of SO₂ and O₃ were below the recommended levels for museum environments. The area connected to the outside (OSR) exhibited the highest concentration of NO_x and NO₂, as expected for pollutants originating outdoors. The concentrations of NOx in the storage room and inside the metal cabinets were lower but comparable. The same trend was observed for NO₂, except in the MCA cabinet, where the concentration was significantly lower. The interaction between the surface of plastic objects made of cellulose acetate and NO₂ changed the distribution of NO/NO₂. The average NO/NO₂ ratio at the different sampling points, excluding MCA, was 0.5, while in MCA, it was 20. Over 96% of NO₂ was removed and re-emitted as NO and HNO₃, resulting in an increase in nitric acid on the surfaces. Moreover, the presence of nitric acid could lead to the formation of ammonium nitrate.

ER-FTIR proved to be effective in determining the presence of nitro groups on the outer surface of some objects, likely due to the impact of NO_x on plastic items conserved in close proximity to their source. The results presented in this study highlighted the interaction between NO_x and CA objects. Based on these findings, it is possible to hypothesise a specific interaction via substitution on a cellulose derivative. Further research is required to confirm the hypothesis and expand it to other kinds of molecular structures.

The high concentration of HNO₃ and HONO in MCN confirms the degradation of cellulose nitrate; the presence of NH₃ in MUF indicates the degradation of urea–formaldehyde objects. This observation agrees with the established decay mechanisms of both CN and UF objects present in the scientific literature. Acetic acid concentrations were notably higher in MCA reflecting degradation of cellulose acetate objects. Formic acid levels were highest in the storage room, likely due to the presence of several wooden and plastic objects.

The results of this work have provided the museum conservator with a better understanding of the issue of the so-called malignant objects; this implies planning the design of the storage of the collection in order to store objects made from different polymer components separately.