Quantification of Heavy Metals in Indoor Dust for Health Risk Assessment in Macao

Abstract

The presence of heavy metals plays a significant role in indoor air quality, which poses a serious public health problem since most of the population spends over 90% of their time in indoor environments. This work investigates heavy metals in indoor dust across different occupational settings in Macao. Field sampling was conducted in five representative locations, which included restaurants, student dormitories, auto repair shops, offices, and parking security rooms, with a total of 11 samples collected in this study. Dust in the form of particulate matter was collected from air conditioning filters to quantify 14 heavy metal contents. The PMF model was applied for source apportionments of the heavy metals, while a health exposure model was used to assess health risks and evaluate the non-carcinogenic and carcinogenic risks in the five representative workplaces. The PMF model identified six major pollution sources: traffic emissions (23.800%), building materials (21.600%), cooking activities (18.500%), chemicals (15.200%), electronic devices (12.300%), and outdoor seaport activities (8.600%). The health risk assessment showed that the overall non-carcinogenic risk (HI = 6.160 × 10⁻⁶ for inhalation, 1.720 × 10⁻³ for oral ingestion, and 2.270 × 10⁻⁵ for dermal contact) and total HI (1.749 × 10⁻³) and carcinogenic risk (6.570 × 10⁻⁹) were below the safety threshold, showing minimal health risk problems. Nevertheless, nickel and chromium were identified as the main contributors to potential long-term risks.

1. Introduction

Indoor air quality (IAQ) has emerged as a critical public health concern, particularly in densely populated urban areas like Macao, where residents spend approximately 90% of their time in enclosed environments. Indoor pollutants, including particulate matter (PM₁₀ and PM_2.5), volatile organic compounds (VOCs), and heavy metals, are linked to respiratory diseases, cardiovascular conditions, and cognitive impairments [1]. The World Health Organization (WHO) estimates that indoor air pollution contributes to 4.3 million premature deaths annually, with economic losses exceeding $40 billion in the U.S. alone due to reduced productivity and healthcare costs. In Macao, rapid urbanization and high population density exacerbate IAQ challenges, necessitating targeted intervention to mitigate health risks.

While the Macao Meteorological and Geophysical Bureau (SMG) monitors the outdoor air pollutants (such as PM_2.5 and NO₂), comprehensive IAQ assessments—especially of indoor dust—remain limited. Indoor dust acts as a reservoir for hazardous substances, including heavy metals (such as Pb and Cd) and organic pollutants (such as PAHs), which accumulate via ventilation systems, outdoor infiltration, and human activities [2]. These pollutants pose long-term health risks through inhalation, dermal contact, and ingestion, particularly in occupational settings like auto repair shops and restaurants, where combustion sources and poor ventilation elevate exposure [3].

This study investigates indoor dust composition across five locations in Macao (student dormitories, offices, restaurants, auto repair shops, and parking security rooms) to address one of the gaps in air pollution risk assessment. IAQ in educational facilities is crucial due to the extended time students spend in those environments, affecting their health, academic performance, and attendance. Underground garages, covered car parks, and other similar structures are another high-risk environment, with the main sources of pollution being motor vehicles and emissions from the heating-ventilation-air-conditioning (HVAC) systems of the joint building [4]. The key parameters of IAQ included building location, layout and construction materials, ventilation and air cleaning systems, finishing materials, occupant demographics, occupancy, and activities [5].

Using air conditioning filters as sampling media, heavy metal contents in indoor dust were analyzed and served as inputs to a receptor model for identifying primary sources (such as vehicular emissions and cooking fumes). Integrating the U.S. EPA exposure model, a health risk assessment was then conducted on the studied indoor environment via multi-pathway exposure to heavy metals [6,7].

The novelty of this work lies in being the first preliminary study of indoor heavy metal assessment in Macao. It may serve as a benchmark for future studies in this region. By bridging local data and well-established risk assessment framework, the research contributes to Macao’s public health objectives and aligns with UN Sustainable Development Goal (SDG) 3: Good Health and Well-being [8,9,10].

2. Materials and Methods

2.1. Study Area and Sampling Locations

The region of Macao is located in the southern part of China, and the land capacity is extremely limited due to geographical constraints and the rapid growth of the population. The United Nations World Prospects Report had listed Macao as the number one most densely populated region in the world. The longitude and latitude of the Macau Peninsula range from 113°31′4″ to 113°37′5″ east and 22°04′0″ to 22°13′03″ north. Five representative sampling locations were selected across Macao (Figure 1) to assess IAQ in diverse indoor environments:

• Student Dormitory: A residential room on the 10th floor of a university campus, medium occupancy density with continuous residential activity.
• Office: Third floor of a university campus, standard working environment with regular working hours.
• Restaurant: Commercial food service establishment with cooking emissions.
• Auto Repair Workshop: Industrial setting with vehicle maintenance activities.
• Parking Security Office: Enclosed space with limited ventilation.

As a preliminary study of indoor dust in Macao, it focuses on setting of the most concern. Despite the limited sample size, this study informs disparities in indoor exposure while setting the stage for further, more comprehensive assessment.

2.2. Indoor Dust Collection and Analysis

This study collects dust from air conditioning (AC) filters. Specifically, the AC units used washable nylon filters that may be reused for many years. It is essential to ensure these filters were not cleaned for at least two months prior to the sampling. Figure 2 shows a dust-loaded AC filter. A stiff brush is used to gently sweep the dust into a collection dish. An acid digestion method was used to treat the dust samples for elemental analysis.

The specific digestion-analysis protocol was optimized from the U.S. EPA reference method [11]. First, 0.500 g of homogenized sample was weighed into a polytetrafluoroethylene digestion tank, and 10.000 mL of concentrated nitric acid (HNO₃, 69.000%) and 2.000 mL of hydrogen peroxide (H₂O₂, 30.000%) were added in sequence. The digestion occurred under a two-stage temperature plateau (pre-digestion at 120 °C for 1 h and main digestion at 180 °C for 4 h). The digestion solution was then filtered through a 0.45 μm nylon filter membrane and fixed to 50.000 mL, in which the contents of heavy metals such as Pb, Cd, and As were determined using an inductively coupled plasma mass spectrometer (ICP-MS) [12,13].

Systematic method validations include: (1) the recovery rate as evaluated by a 0.100–10.000 mg/kg gradient spike experiment (in compliance with NIOSH 7300 standards [14], controlled at 85.000–115.000%); (2) the repeatability as characterized by the intra-day relative standard deviation of 6 sets of parallel samples (RSD < 5.000%); (3) the reproducibility as cross-validated by three certified laboratories (referenced to the ISO/IEC 17025 [15] framework, inter-laboratory RSD < 8.000%); (4) the limit of detection (LOD) quantified as 3 times the standard deviation of the blank signal (0.0010–0.0100 mg/kg) and the limit of quantification (LOQ) quantified as 10 × LOD; and (5) the matrix effect corrected by the standard addition method (according to EPA 6020B, the slope deviation of the calibration curve between the spiked sample and the pure solvent was <10.000%) [2].

2.3. Receptor Modeling Source Apportionment

The selection of the U.S. EPA Positive Matrix Factorization (PMF) model for source apportionment in this study was based on its well-established capability to handle complex environmental mixtures while adhering to physically realistic constraints [16]. PMF analyzes the concentration matrix of measured pollutants to find two non-negative matrices, G and F, minimizing the difference between the product GF and the original data matrix X. The objective function Q is defined as the weighted sum of squared differences between X_ij (pollutant j’s concentration in sample i) and the corresponding value from GF, divided by the uncertainty U_ij:

$$X_{ij}=\sum _{k=1}^p\left(G_{ik}\times F_{kj}\right)+E_{ij}$$

(1)

$$Q=\sum _{i=1}^n\sum _{j=1}^m\left({\displaystyle \frac{E_{ij}}{U_{ij}}}\right)$$

(2)

The weights are based on the measurement uncertainties, and thus the model gives more weight to data points with smaller uncertainties. PMF’s non-negativity requirement for both G and F matrices aligns with the inherent characteristics of indoor dust pollutants, where negative source contributions are physically implausible. Uncertainty weighting was implemented following EPA guidelines, with measurement uncertainties calculated differentially for concentrations above and below MDLs, an approach validated in similar studies [17,18]:

$$U_{ij}={\displaystyle \frac{5}{6}}\times {\mathrm{M}\mathrm{D}\mathrm{L}}_j$$

(3)

when X_ij < MDL, and

$$U_{ij}=\sqrt{{({\delta }_j\times X_{ij})}^2+{(0.5\times {\mathrm{M}\mathrm{D}\mathrm{L}}_j)}^2}$$

(4)

when X_ij ≥ MDL. In Equations (3) and (4), δ_j and MDL_j are pollutant-specific measurement precision and detection limit.

PMF derives source profiles (F) and contributions (G) through iterative calculations to minimize the objective function Q. The number of factors (p) is determined by analyzing different factor numbers and using criteria like the Q value plot and residual distribution. Once the optimal factor number is selected, the final G and F matrices are obtained, reporting source contributions and profiles, respectively. In this analysis, the number of factors selected for analysis was 3–8. After 20 iterations of each selected number of factors, the calculation results were boosted 100 times to verify their stability. The most tested shadow factor number is p = 6.

Although the sample size was modest, the model’s applicability in this study was confirmed through several key considerations: limiting the analysis to six factors or fewer as recommended for smaller datasets, the presence of strong marker elements enabling clear source differentiation, and bootstrap analysis demonstrating factor stability. The resolved factors were further validated against local emission inventories, regional source signatures, and technical reports on characteristic emission ratios. Moreover, model performance metrics, including specific ratio values, the percentage of scaled residuals within an acceptable range, and factor correlations in bootstrap analysis, all met established acceptability criteria [19].

2.4. Exposure Assessment Model

There are three main ways for heavy metals in dust to enter the human body: direct ingestion through the hand-mouth route, inhalation through the respiratory system, and skin contact. This study focused on eight heavy metals (Be, Cr, Ni, Cu, Zn, As, Cd, Pd), where the International Agency for Research on Cancer (IARC) classifies As, Ni, Cr, Cd, and Be as human carcinogens potentially causing urinary bladder, lung, skin, prostate, and nasal cancers [1]. The most dangerous route for Cd, Cr, Ni, and Be exposure is through inhalation due to high bioavailability, while oral ingestion of As and Cd is the primary concern and they accumulate within the human body, in general, with medium bioavailability. Dermal contact of Ni and Cr may lead to skin damage though they have a relatively low bioavailability [20].

This study assesses the non-carcinogenic and carcinogenic risks of the heavy metals in dust. The corresponding long-term average daily dose (ADD) via inhalation (Inh), oral ingestion (Ing), and dermal contact (Der) is calculated according to Cai et al. (2017) and Wang et al. (2025) [4,20]:

$$ADDInh={\displaystyle \frac{(c\times IR\times EF\times ED)}{(PEF\times BW\times AT)}}$$

(5)

$$ADDIng={\displaystyle \frac{Cd\times IRd\times CF\times FI\times EF\times ED}{BW\times AT}}$$

(6)

$$ADDDer={\displaystyle \frac{Cd\times CF\times SAd\times AF\times ABSd\times EF\times ED}{BW\times AT}}$$

(7)

where IR represents the breathing rate of adults (m³/day); EF represents the annual exposure frequency of adults via the respiratory route (days/year); ED represents the exposure years of adults (years); BW represents the average body weight of adults (kg); AT represents the average exposure time of pollutants (days); c and PEF is the heavy mental content in dust (mg/kg) and particle emission factor (m³/kg), respectively, Cd represents the concentration of contaminant in indoor dust (mg/kg); IRd represents the indoor dust ingestion rate (mg/day); CF represents conversion factor; SAd represents skin surface area in contact with indoor dust (cm²/event); and AF represents skin adherence factor for dust (mg/cm²). The values of relevant exposure parameters are shown in

Table 1. Parameters of average daily exposure of heavy metals in dust.

Parameter *	Unit	Value
c	mg/kg	Heavy metal content in dust from sampling sites
IR	mg/day	12.800
EF	days/year	200.000
ED	years	30.000
BW	Kg	55.900
AT	days	ED × 365 (for non-carcinogenic), 70 × 365 (for carcinogenic)
PEF	mg/kg	1.360 × 109
CF	kg/mg	1.000 × 10−6
FI		1.000
SA	cm2/event	400.000
AF	mg/cm2	0.500
ABS		0.001

* IR, EF, BW, AT, PEF, CF, FI, SA, AF, and ABS in the table are all referenced to [4], while ED and c are from the data in the study.

, assuming the population involved in this study is all working adults (aged 18–65 years).

The respiration rate (12.800 m³/day) derives from Chinese adult studies, exceeding the US EPA default (10.000 m³/day) to reflect moderate-activity respiratory patterns in Asian populations; the annual exposure frequency (200.000 days/year) is sourced from Macao Labour Affairs Bureau data (accounting for holidays) and complies with ISO 17734 [21] occupational standards; exposure duration (30 years) considers the working-life span (25–55 years), consistent with WHO chronic exposure guidelines; body weight (55.900 kg) represents a weighted average (male: 62.100 kg; female: 49.700 kg) from cited 2023 demographic statistics; exposure time differentiates non-carcinogenic risk (ED × 365 days) from carcinogenic risk (70-year lifetime, per WHO East Asia data) for cumulative effect assessment; while the particulate emission factor directly references US EPA datasets.

2.5. Health Risks Calculations

For non-carcinogenic risk, the hazard quotient is calculated as:

$$HQ=ADD/RFD$$

(8)

In Equation (8), RFD represents the reference dose of the pollutant in mg/kg-d, which is the maximum number of pollutants that can be taken in per unit body weight per unit time without causing adverse reactions in the human body. If HQ < 1, the risk is considered small or negligible; if HQ > 1, it is considered that there is a non-carcinogenic risk.

For carcinogenic risk, the product of long-term average dose and carcinogenic slope factor is used to measure it, thus:

$$Risk=ADD\times SF$$

(9)

In Equation (9), SF is the carcinogenic slope factor, which indicates the maximum probability of a certain dose of a pollutant causing a carcinogenic effect in humans, [mg/(kg-d)]⁻¹. It is generally believed that when the carcinogenic risk is lower than 10⁻⁶ ~10⁻⁴ (one cancer patient is added for every 10,000 to 1 million people at risk), the pollutant is considered to have no carcinogenic risk [22,23,24,25].

The RFD and SF of each heavy metal element are shown in

Table 2. Reference doses of metals and slope factors of heavy metal exposure.

Item	RfD Inh * mg/kg-Day	RfD Ing * mg/kg-Day	RfD Der * mg/kg-Day	SF * [mg/kg-Day]−1
Zn	3.000 × 10−1	3.000 × 10−1	3.000 × 102
Pb	3.520 × 10−3
Cu	4.020 × 10−2	4.000 × 10−2	4.000 × 101
Be	2.000 × 10−2	2.000 × 10−3	2.000	1.100
As	1.230 × 10−4	3.000 × 10−4	1.000 × 10−2	4.300 × 10−3
Ni	2.060 × 10−2	2.000 × 10−2	2.000	0.840
Cr	2.860 × 10−5	3.000 × 10−3	3.000 × 10−1	42.000
Cd	1.000 × 10−3	5.000 × 10−4	5.000 × 10−1	6.300

* The RFDs of Zn, Pb, Cu, Be, As, Ni, Cr, and Cd are all from the reference (Cai et al. [4]); As, Ni, Cr, and Cd are carcinogenic metals.

.

3. Results and Discussion

3.1. Analysis of Heavy Metal Pollution Characteristics

A total of 11 indoor dust samples were collected from the five distinct indoor environments in Macao, including three office samples, two restaurant samples, two security room samples, two auto repair shop samples, and two dormitory samples. The analysis determined their contents of 26 metal elements: Be, B, Na, Mg, Al, Si, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Mo, Ag, Cd, Sb, Ba, Tl, and Pb.

The results showed that the metal contents varied significantly across different dust samples, with the ranges shown as follows: Be (0.004–0.023 μg/g), B (4.266–10.309 μg/g), Na (5493.669–24815.662 μg/g), Mg (1181.895–4693.406 μg/g), Al (13.352–103.909 μg/g), Si (10.894–67.222 μg/g), K (669.589–4257.889 μg/g), Ca (3725.682–16024.947 μg/g), Ti (1.377–8.304 μg/g), V (0.129–0.584 μg/g), Cr (0.591–3.353 μg/g), Mn (9.341–448.040 μg/g), Fe (14.412–447.950 μg/g), Co (0.092–222.406 μg/g), Ni (3.423–992.715 μg/g), Cu (2.473–30.545 μg/g), Zn (41.400–1659.581 μg/g), As (0.747–2.999 μg/g), Se (0.431–2.309 μg/g), Mo (0.344–1.102 μg/g), Ag (0.002–0.018 μg/g), Cd (0.003–0.208 μg/g), Sb (0.180–1.821 μg/g), Ba (0.136–3.243 μg/g), Tl (0.004–0.064 μg/g), and Pb (0.015–1.003 μg/g).

Certain metals were significantly enriched in specific types of environments. For instance, Mn, Fe, Co, and Ni are relatively high in auto repair shops; Ca, Mg, and Al are relatively high in restaurants; Se, and Pb are relatively high in dormitory rooms; and Na, Sb, and Zn are relatively high in offices. The maximum content of each element and where it occurs is shown in

Table 3. The maximum content of each metal element and the environment in which it occurs.

Element	Maximum Value (μg/g)	Indoor Environment Where It Occurs
Be	0.020	Dining Room 1
B	10.310	Security Hall 1
Na	24815.660	Office 2
Mg	4693.410	Dining Room 2
Al	103.910	Dining Room 2
Si	67.220	Auto Repair Shop 1
K	4257.890	Office 2
Ca	16024.950	Dining Room 2
Ti	8.300	Dining Room 2
V	0.580	Dining Room 2
Cr	3.350	Office 1
Mn	448.040	Auto Repair Shop 2
Fe	447.950	Auto Repair Shop 2
Co	222.410	Auto Repair Shop 1
Ni	992.710	Auto Repair Shop 2
Cu	30.540	Auto Repair Shop 1
Zn	1659.580	Office 1
As	3.000	Dormitory Room 2
Se	2.310	Dormitory Room 2
Mo	1.100	Auto Repair Shop 1
Ag	0.020	Office 3
Cd	0.210	Office 1
Sb	1.820	Office 2
Ba	3.240	Dining Room 1
Tl	0.060	Dormitory Room 2
Pb	1.000	Dormitory Room 2

.

The Na content of indoor dust in the office setting (21,500.520 ± 2809.870 μg/g) is significantly higher than that in other settings. The office area is mainly used by schoolteachers, with high hygiene requirements, and is cleaned every week. The Na may result from frequent human activities and frequent use of detergents. K is also enriched in office dust (up to 4257.890 ug/g), which may be related to common plant fertilizers and/or certain office supplies. In addition, the highest Sb content appears in Office 2 sample (1.820 ug/g), possibly coming from flame-retardant materials in some office equipment.

The Zn content in Office 1 indoor dust was exceptionally high at 1659.580 μg/g, 17 times higher than the overall average. Investigations revealed that the specific office area contains many laboratory equipment and electronic devices, including printers and computers. These could be the underlying cause of the Zn pollution.

The Ca (14,596.310 ± 2014.250 μg/g) and Mg (4565.270 ± 138.140 μg/g) contents of indoor dust were significantly higher in the catering area serving Portuguese dishes, mainly seafood. It was found that seafood products contain a large amount of calcium and magnesium [26]. Compared with similar studies on mainland China catering areas, the calcium content in Macau’s catering areas is on average about 15.000% higher. This may reflect the seafood-based dietary characteristics of Macau.

The auto repair shop exhibited a distinct heavy metal pollution profile, with Mn (356.600 ± 129.250 μg/g), Fe (up to 447.950 μg/g), Co (335.180 ± 159.410 μg/g), and Ni (740.090 ± 357.030 μg/g) contents in indoor dust significantly higher than those in other settings. Especially in Auto Repair Shop 2, the Ni content reached as high as 992.710 μg/g, Co content reached 222.410 μg/g, and Mn content reached 448.040 μg/g.

The dormitory dust showed relatively high Pb (0.820 ± 0.270 μg/g), As (2.770 ± 0.300 μg/g), and Ti (up to 0.060 μg/g) contents. Pb content in Dormitory Room 2 reached 1.000 μg/g, while As content reached 2.990 μg/g. Residential buildings are commonly built of concrete and sea sand mixture in Macao, and sea sand may contain high level of natural heavy metals. Macao’s high temperature and humidity conditions accelerate the corrosion and aging of building materials, leading to an increase in heavy-metal release.

3.2. PMF Heavy Metal Source Analysis

In this study, the US EPA PMF 5.0 model was applied for the source apportionment of heavy metals in indoor dust. The initial base model run of this study was selected from factors 3 to 10. After numerous trial runs, the optimal solution contains 6 factors. Figure 3a,b show the source apportionment results [26,27,28,29].

Factor 1 contributes significantly to Zn and Sb, at 43.910% and 43.590%, respectively, on average. Previous studies by Thorpe et al. [30] and Hjortenkrans et al. [31] attribute Zn and Sb to tire wear and brake wear emissions in urban areas without significant industrial or combustion sources, such as in Macao. Thus, Factor 1 is identified as a traffic source which has contaminated the indoor air through infiltration.

Factor 2 marks contributions to Ag and B at 38.620% and 34.210%, respectively. Some air purifiers, air conditioning filters, or antibacterial coatings contain silver ions or nano silver [32], while boron compounds (such as boric acid and borax) are present in cleaning agents, flame retardants, or fiberglass products and can enter the air via spraying or volatilization [33]. Given the widespread use of cleaning devices/agents in the sampled public spaces, Factor 2 is deemed to originate from coatings and chemical products.

Factor 3 made substantial contributions to Al and Ti, at 62.110% and 47.430%, respectively. Al is extensively used in building structures (e.g., aluminum-alloy doors and windows), coatings, and fire-resistant materials, and long-term wear or aging can release Al-containing particulates [34]. Titanium dioxide (TiO₂), a common additive in coatings, plastics, and ceramics, can enter indoor air through the degradation of materials or during construction processes [35]. Hence, Factor 3 is attributed to building and furnishing materials.

Factor 4 marks a significant contribution to Tl at 49.150%. Surveys revealed the prevalence of old electronic products in the sampled indoor settings, with most containing over 10 such items, particularly in school offices with laboratory equipment. Old electronics like CRT monitors and semiconductor materials contain Tl, which can be released through improper use and disposal [36,37,38]. Factor 4 is identified as an electronic device source.

Factor 5 contributes notedly to Na at 33.060% overall. High-temperature cooking (such as frying and barbecuing) releases abundant Na-containing particulates, especially when sodium-rich seasonings like table salt are used [39]. Factor 5 is therefore recognized as a cooking source.

Factor 6 shows a significant contribution to Co and Ni, at 89.450% and 86.840%, respectively. Heavy fuel oils (HFO) and lubricates used in marine vessels contain high concentrations of Ni and Co [40,41]. The nearby Hong Kong SAR has one of the busiest ports in the world [20]. In some cases, Co and Ni emitted from HFO-powered ships can enter the room through the ventilation system or gaps in doors and windows. Factor 6 is thus related to shipping and port operations.

Figure 3a shows the source profile of six factors of indoor dust inferred from the PMF model. The color patterns represent the concentrations of each species within a given source, and the dots indicate the percentages of each species contributing to each factor. Figure 3b shows the source contribution rate of the selected heavy metal variables based on the PMF 5.0 model.

The contributions of the six major sources to heavy metal pollution of indoor dust in Macau are ranked as follows: traffic source (23.800%), building and furniture materials (21.600%), cooking source (18.500%), coating and chemical products (15.200%), electronic device source (12.300%), and outdoor shipping activity (8.600%). As shown in Figure 4, traffic emissions (Zn/Sb released by motor vehicle brake wear), building material aging (Al/Tl precipitated from sea sand concrete), and cooking activities (Na/Ca enriched by high-temperature processed seafood) contribute to a total of 63.900% of heavy metals, highlighting the compound impact of Macau’s high-density urban setting and unique food culture. The relative contribution of cross-border pollution (such as regional ship emissions input Co/Ni) is less than 10.000%, reflecting the limited impact of marine shipping activities on the indoor environment. It is worth noting that the Tl released by electronic equipment (accounting for 12.300% of heavy metals) exposes the weak links in the management of old electronic products, while coating and chemical products (contributing Ag/B, etc.) point to the lack of specifications for the use of chemicals.

The findings show that optimizing local traffic structure, upgrading anti-corrosion technology of building materials, and standardizing catering operation procedures are the core paths to reduce indoor heavy metal pollution [42,43,44,45,46]. At the same time, it is necessary to strengthen the full-cycle supervision of electronic waste to control emerging risk sources.

3.3. Heavy Metal Health Risks Assessment

The health risk assessment focused on non-carcinogenic and carcinogenic risks of eight heavy metals in indoor dust. In terms of non-carcinogenic risk, the health index (HI) for inhalation, oral ingestion, and dermal contact, calculated as the sum of HQ values, is about 6.160 × 10⁻⁶, 1.720 × 10⁻³, and 2.270 × 10⁻⁵, respectively (

Table 4. Daily non-carcinogenic exposure dose and non-carcinogenic risk of heavy metals in dust through inhalation.

Compound	ADD (mg/kg/Day)	HQ	HI
Be	1.070 × 10−12	5.350 × 10−10	6.160 × 10−6
Cr	1.260 × 10−10	4.400 × 10−6
Ni	1.190 × 10−8	5.780 × 10−7
Cu	1.080 × 10−9	2.670 × 10−8
Zn	1.800 × 10−8	5.990 × 10−8
As	1.320 × 10−10	1.070 × 10−6
Cd	7.760 × 10−12	7.760 × 10−9
Pb	3.460 × 10−11	9.840 × 10−9

,

Table 5. Daily non-carcinogenic exposure dose and non-carcinogenic risk of heavy metals in dust through oral ingestion.

Compound	ADD (mg/kg/Day)	HQ	HI
Be	1.470 × 10−9	7.370 × 10−7	1.720 × 10−3
Cr	1.840 × 10−7	6.130 × 10−5
Ni	1.740 × 10−5	8.690 × 10−4
Cu	1.570 × 10−6	3.920 × 10−5
Zn	2.620 × 10−5	8.740 × 10−5
As	1.920 × 10−7	6.400 × 10−4
Cd	1.130 × 10−8	2.270 × 10−5
Pb	5.060 × 10−8

and

Table 6. Daily non-carcinogenic exposure dose and non-carcinogenic risk of heavy metals in dust through dermal contact.

Compound	ADD (mg/kg/Day)	HQ	HI
Be	2.300 × 10−11	1.150 × 10−11	2.270 × 10−5
Cr	2.870 × 10−8	9.570 × 10−8
Ni	2.720 × 10−5	1.360 × 10−5
Cu	2.450 × 10−7	6.130 × 10−9
Zn	4.100 × 10−6	1.370 × 10−8
As	9.000 × 10−8	9.000 × 10−6
Cd	1.770 × 10−10	3.540 × 10−10
Pb	4.740 × 10−9

) and the total HI is 1.749 × 10⁻³, which is far below the safety limit of 1, indicating that there is no significant non-carcinogenic health risk from indoor dust exposure. The HQ values through inhalation, ranked from high to low, are: chromium (Cr, 4.400 × 10⁻⁶) > arsenic (As, 1.070 × 10⁻⁶) > nickel (Ni, 5.780 × 10⁻⁷) > zinc (Zn, 5.990 × 10⁻⁸) > copper (Cu, 2.670 × 10⁻⁸) > lead (Pb, 9.840 × 10⁻⁹) > cadmium (Cd, 7.760 × 10⁻⁹) > beryllium (Be, 5.350 × 10⁻¹⁰), through oral ingestion, Ni (8.690 × 10⁻⁴) > As (6.400 × 10⁻⁴) > Zn (8.740 × 10⁻⁵) Cr (6.130 × 10⁻⁵) > Cu (3.920 × 10⁻⁵) > Cd (2.270 × 10⁻⁵) > Be (7.370 × 10⁻⁷), and through dermal contact, Ni (1.360 × 10⁻⁵) > As (9.000 × 10⁻⁶) > Cr (9.570 × 10⁻⁸) > Zn (1.370 × 10⁻⁸) > Cu (6.130 × 10⁻⁹) > Cd (3.530 × 10⁻¹⁰) > Be (1.150 × 10⁻¹¹).

In terms of carcinogenic risk, as shown in

Table 7. Carcinogenic risk assessment of heavy metals in dust.

Compound	ADD (mg/kg/Day)	Risk	Total Risk
Be	4.580 × 10−13	5.040 × 10−13	6.570 × 10−9
Cr	5.400 × 10−11	2.270 × 10−9
Ni	5.100 × 10−9	4.290 × 10−9
As	5.640 × 10−11	2.260 × 10−13
Cd	3.330 × 10−12	2.100 × 10−11
Pb	1.480 × 10−11	1.260 × 10−13

, the total risk is about 6.570 × 10⁻⁹, which is lower than the safety threshold (1.000 × 10⁻⁶), indicating that the overall carcinogenic risk of heavy metals in indoor dust in the study area is low and will not cause health hazards to humans. However, nickel (Ni, 4.290 × 10⁻⁹) and chromium (Cr, 2.270 × 10⁻⁹) are the main contributing factors, accounting for more than 99.000% of the total carcinogenic risk. Although the assessment results show that the risk level has not been exceeded at present, the potential impact of nickel and chromium still needs attention. This work only calculated the risk through inhalation but did not calculate the risk through oral intake and other routes.

4. Conclusions

Combining environmental sampling, PMF source analysis, and US EPA exposure model, this study systematically evaluated the pollution characteristics, source contributions and health risks of heavy metals in indoor dust across multiple indoor settings in Macau. The study reveals the spatial heterogeneity of heavy metal pollution. Auto repair shops represent pollution hotspots for industrial heavy metals (mainly Mn, Fe, Co and Ni) due to motor vehicle maintenance activities (brake pad wear and welding operations. This type of pollution poses a serious health threat to practitioners, and long-term operation can lead to the accumulation of lung cancer risks. Catering places are driven by cooking activities. Mineral-rich seafood ingredients release a variety of characteristic elements (mainly calcium and magnesium) during high-temperature processing. The concentration is about 15.000% higher than similar studies in mainland China, confirming the unique pollution imprint of Macau’s food culture. In the office environment, high-density personnel activities and the use of detergents lead to enrichment of Na elements, flame retardant materials of electronic equipment release Sb, and Zn pollution in areas with dense experimental instruments highlights the challenges in electronic device management. The accumulation of As, Pb, and Tl in student dormitories is closely related to the aging of building materials [47,48,49,50,51].

Through PMF analysis, six major pollution sources in the indoor environment investigated were identified: traffic emissions (23.800%, contributing Zn/Sb), building material sedimentation (21.600%, mainly Al/Ti), cooking emissions (18.500%, enriched Na/Ca), chemical product emissions (15.200%, releasing Ag/B), electronic equipment emissions (12.300%, releasing Tl) and ship activities (8.600%, input Co/Ni). Among them, the contribution of local pollution sources reached 63.900%, highlighting the joint impact of high-density urban operations and special industrial activities; the contribution of cross-border pollution (such as Hong Kong ship emissions) was less than 10.000%, reflecting the indirectness of regional transmission on the indoor environment [52,53,54].

The total non-carcinogenic risk value (HI) is 1.749 × 10⁻³, which is far lower than the safety threshold HI < 1 of the US EPA. In terms of carcinogenic risk, the total risk is about 6.570 × 10⁻⁹, which is also lower than the safety threshold risk (1.000 × 10⁻⁶) of the US EPA. Therefore, it shows that the overall carcinogenic risk of heavy metals in indoor dust in the study area is low and will not cause health hazards to the human body [55,56,57]. The limitations of this study include a relatively small sample size and limited settings, which may not fully reflect the pollution characteristics of different indoor environments. Future studies will increase the sample size and the variety of sampling locations.