Skip to main content

Identification and control of the volatile organic compounds activity in confined environments (Mosques)


Mosques within the professional environments such as universities, banks and shopping malls possess different features that from the public mosques. These mosques are of relatively smaller size, less airy, populated for a shorter duration of the day and the visitors with more formal dressing. All these factors contribute to an increased activity of Volatile Organic Compounds (VOCs) generated by the visitors. Improper drying of the wet feet after ablution and/or prolonged wearing of socks, while being in professional environments, further worsens the situation. A prolonged activity of these VOCs in confined environments is not only unpleasant but also poses certain health issues. This study aims at identification of the need to control this activity within an acceptable limit by deploying a low-cost smart computational configuration of sensor array. The configuration parameters depend on a number of factors such as mosque space, ventilation, air conditioning and the quantity and quality of VOCs generated on average. Pertinent VOC data monitoring, computing and averaging across a network is done in real-time using mobile sensing stations. A proof of concept-based technical feasibility on the adoption of a control strategy is conducted by establishing an acceptable threshold level for the VOCs activity in varying conditions. As the sensitivity of the sensor and the lifespan & proliferation of the VOCs are affected by the humidity, temperature and the air circulation, these factors are closely monitored. Mosque being a sacred place, all these experiments are constrained to be conducted in a controlled manner without any disturbance whatsoever to the environment, visitors and the religious activities being performed.


There is quite a large number of organic compounds in the air we breathe in. However, among the three most abundant yet less reactive organic compounds are CO2 (Carbon dioxide, ~ 370 ppm), CH4 (Methane, ~ 1.8 ppm) and CO (0.15 ppm). VOCs are certain compounds -from a broad group of carbon-based organic chemical- found in gaseous state at room temperature and usually participating in atmospheric photochemical reactions. These are comprised of one hydrogen atom and a minimum of one and maximum of 15 carbon atoms. Some of the compounds (e.g., formaldehyde) are though VOC, however treated differently because of their unique characteristics.

There is a large number and type of VOCs. Based upon boiling point, they are grouped into highly volatile (0 °C to 50–100 °C), volatile (50–100 °C to 240–260 °C) and semi-volatile compounds (240–260 °C to 380–400 °C) [1]. VOC are also grouped on the basis of their chemical structure such as aliphatic, aromatic, chlorinated, aldehydes, ketones, alcohols, etc.

Due to an ever-increasing trend to stay indoor (whether at home, office or within a vehicle), especially in the urban life, associated challenges are also rising [2, 3]. These is a significant rise on the health concern and the comfort level of the occupants of a building due to various indoor pollutants along with the temperature and humidity variations [4]. The VOCs that are considered as the major source of diseases include propanol acetaldehyde, n-propyl acetate, methyl methacrylate, 1-dipropoxypropane and styrene [5]. Almost all the activities performed by an individual results the emission of various kinds of organic compounds [6, 7]. These include carbonyls, alcohols, alkanes, alkenes, esters, aromatics, ethers, amides, etc. Extended exposure may result in symptoms such as throat irritation, eye-redness, itchy nose, intermittent headache, nausea, fatigue, dizziness, etc. An increased sensitivity may be experienced as well to a single or a mixture of two or more, or the sum of all chemicals (commonly referred to as a total volatile organic compound or TVOC). The term MVOC (Microbial VOCs) is also used for specific type of VOCs that are produced by microbial organism e.g., those produced by the soiled clothing or pungent odor socks.

Curran et al. [8] claim that there is no significant work done on the VOCs present in the human odor as compared to studies made on the human sweat. Study of human sweat itself had not been sufficient to study the nature of VOCs present. Further to this, study on the VOCs from foot odor has received even more limited attention than from hands, forearm and axillae [9]. This is despite the fact that bacteria population responsible for the VOC activity is quite dense on the foot region, considered as the favorite biting site of the anthropophilic mosquitoes causing malaria [9]. Socks and various types of canvas shoes as well as leather shoes may retain these VOCs for a long time. The growth of such organic compounds and their retention in socks and shoes increases with moist and favorable temperature especially in closed places with limited air circulation. In a study by [10], six volatile compounds that include butyric acid, dimethyl disulfide, dimethyl trisulfide, 2-heptanone, 2-nonanone and 2-octanone, in decreasing order of their presence, were identified as the major source of malodor in soiled clothing (such as used socks).

Sometimes VOCs exhibit sink effect due to the presence of wooden furniture, carpets, curtains or clothing. This effect is due to earlier absorption of the VOCs while the emission rate is relatively higher and later a slow release of these VOCs. Usually, higher the boiling point of a specific VOC, higher the sink effect. This implies that the sensor data obtained is less dynamic in nature. In lab-measurements, sink effect is usually minimized using special arrangements. on the other hand, deep learning-based models, such as recurrent neural network employing dynamic training [11], can also be used to compensate such effect.

VOC emission is usually controlled at source level using proactive approach. This means eliminating or reducing the sources of VOCs instead of treating them. This is done by proper ventilation and selecting low-emitting materials/products in wall paints, floorings, carpets, furniture and other office equipment as well as limited use of the aerosol products, bleach, detergents, disinfectants, cleaning solutions, air-fresheners and adhesives.

Rest of the article is structured as follows. In the next section, a brief review of various measuring methods of VOC activities is discussed. In the subsequent section, a detailed description of the methodology and experimental setup is given justifying its use in the environment like the mosques at professional places. This follows with the presentation of the key results and discussion. Finally, conclusions of the study are presented.

Measurement of VOCs

Response of the human nose has been widely used for odor identification in many studies but lack of quantification makes it subjective. This calls for the use of variety of other methods like mass spectrometry, sensitive sensors, chemical and atomic emission detectors. Sensors-based measurements are relatively fast, reliable for the continuous monitoring and more cost-effective. Six major types of sensors that are used as VOC sensors include: Photo-Ionization Detectors (PID) electrochemical sensors (ampere-metric or potentio-metric), metal oxide sensors (MOx), optical sensors, micro-gas chromatograph (µGC) and electronic noses and sensor arrays [12].

Metal Oxide (MOx) sensors make use of tin-dioxide surface (or some other metal oxide such as ZnO), a heater and a sensing chip in order to measure conductivity variations. When the sensing surface comes in contact with a reducing agent, such as a reducing gas, oxygen molecules that are absorbed to the metal oxide surface and causing the electrons to be trapped, are removed [13]..This allows the trapped electrons to flow hence causing an increased conductivity across the metal surface [12]. Although the principle of MOx sensor measurement is quite straight-forward, however identification and individual quantification of the contaminant concentration are quite complicated. This is not only due to the limitations of the construction of the sensor but also due to the impact of the variation in temperature and relative humidity [14, 15].. The measurement process gets further complicated due to the cross-sensitivities other than the target contaminants [16]. Sensitivity is further influenced due to the nano-structure of the MOx surface, limiting its use to relatively higher concentrations of the contaminants [17]. These concentration of the VOCs does not simply increase over time due to certain removal processes, known as sinks, being taking place at the same time. Atmospheric chemical oxidation is one of such sink process mainly responsible for the reduction of VOCs concentration [18]. Poor ventilation rates result in elevated VOCs’ activity enabling low-cost sensors to measure them proportionally. This helped to resolve the issue faced by majority of the low-cost VOC sensors, that is the limited range of detection (usually 1 part-per-million (ppm)). Under normal conditions, parts-per-billion level measurements require sophisticated equipment with a very high associated cost. However, with the use of array of sensors, reduced ventilation and proportionally confined chamber helps to resolve these issues.

For a complete breakdown of each individual compound along with quantification, a detailed chemical analysis is required in a laboratory. This requires collection of the specimen under controlled environments. This study is limited to continuous monitoring of the TVOC level with the usage of low-cost sensors without any need of the individual quantification.

Experimental setup

This research focuses on the entrance areas of the prayer halls of three mosques in the faculty of engineering. One of the mosques is purposefully designed to be a prayer area for relatively larger number of people, while the other two are not custom built, rather designated for this purpose and are relatively smaller in size.

Configuration of mosque area

The layouts of the mosque areas are given in Fig. 1. Table 1 describes the parameters of theses mosques, such as area, height, flooring, carpeted area, area specified for the shoes, etc.

Fig. 1

Layout of the three mosques namely M1, M2 and M3

Table 1 Configuration Parameters

The sampling of the VOC measurements in the mosques was performed for two hours from 1220–1330 h and 1510–1620 h. These hours represent the peak hours for the mosques due to the gathering for prayers of Duhar and Asar. The sampling data were collected for 30 working days, for each mosque with a carpeted floor. All the samples were collected with continuous temperature (T) and relative humidity (RH) monitoring. All of the mosques rely on passive ventilation (i.e., windows and doors) and active ventilation (i.e., HVAC) systems during prayer time. The doors were kept open during the monitoring period, and the HVAC system was in normal working conditions throughout the experiments. During these measurements, arrival patterns of the visitors and certain characteristics about these visitors are also recorded. These include lead time, wearing of socks (yes/no) and fresh ablution (yes/no).

Since there is no change in the rate of ventilation (or type of ventilation i.e., switching from active to passive or vice versa) during the entire period of monitoring, the air exchange rate is assumed to be constant over time and uniform across the space. By adding a proportionate confined space for the sensing station, air contaminant concentration is increased significantly within this space that is full of various types of shoes and socks. This enables to sense the air contamination within sub-ppm concentration range using an array of low-cost sensors as motivated by [19]. This configuration of sensors used in the sensing testbed is particularly useful due to geospatial nature of the sensors [20]. There is negligible outdoor air exchange as the entire building is centrally air-conditioned, and the doors of M1 and M2 are always open. Door of M3 is usually closed, however due to very efficient air conditioning system, any air exchange rates are neglected.

Sensing testbed is placed close to the entrance area but at a sufficient distance from the door. This area is reserved for the shoes/socks. Due to the use of low-end VOC sensors, detection level is relatively low. To increase the level of VOC activity, air exchange rate is limited by covering this area. This serves an approximation for the measurement of the VOC activity inside the area of interest. Since the time of stay of each individual is short, we are not concerned in any possible health risks associated with the exposure to VOCs or various contaminants. Covering is designed as a rack with transparent plastic to reduce the opening time to a minimum for stacking the shoes/socks. Back end of the racks is sealed and a set of six fans force the flow inside the housing on the sensor surface area via small steel tubing network. This ensures maximum possible exposure of the contaminated air to the surface of the VOC sensors. Temperature, pressure and CO sensors are installed within the housing toward a single outlet for the air outflow.

Visitors are requested to place their shoes in the rack and immediately cover them. Most of the time, a helper is placed to ensure that no element of the sensor configuration is disturbed during the loading or unloading of the shoes. It is required to remove the coverings within the last fifteen minutes and to perform cleaning of the entire room as usual. During this process, the cleaner vacuums the carpets, and makes use of air-fresheners.

Configuration of sensors

Three independent sensing stations are custom built to monitor the TVOC activity along with temperature and relative humidity measurements. These stations use a raspberry pi3 B + and a set of sensors TGS2602, TGS2600, TGS 2601 and DHT22.

Conductivity of the VOC sensors (TGS2602) varies due to varying conditions and the associated voltage-values vary accordingly. These values are normalized, recorded locally and transmitted directly.

According to the data provided by the sensor manufacturer, resistance is calculated using Eq. 1.

$$R_{s} = \frac{{V_{C} \times R_{L} }}{{V_{o} }} - R_{L}$$

where \(R_{s}\) is the resistance of the sensor that varies according to the air quality, \(V_{C}\) is the circuit voltage, \(R_{L}\) is the load resistance, and \(V_{o}\) is the output voltage. Equation 2 is used to obtain the normalize resistance value, \(R_{N}\).

$$R_{N} = \frac{{R_{s} }}{{R_{0} }}$$

where \(R_{0}\) is the resistance of sensor in clean configuration. Sensor data are transmitted in real-time to the central server while data related to the visitor characteristics are recorded manually and later merged to the central database. This process was kept manual due to privacy concerns of the visitors. Two similar forms are used to keep track of visitor arrival timing, stay and the desired characteristics on daily basis for the Duhar and Asar prayer schedules. Sample form is shown in figure. the Arrival and departure times are recorded through infrared sensors as well to ensure data integrity. Sensing stations once installed and started recording the data are never stopped until the last day. This is done to ensure the compliance with the conditioning period of each sensor that varies due to sensor construction.

Metal Oxide (MOx) sensors, based upon the conductivity variation in accordance with the exposure to different gases or VOCs level, are used. These conducti-metric MOx sensors are low cost, simple and robust in their operation. Sensor calibration is performed in an external setup by a commercial research lab. It is recommended to maintain the heater voltage at 5.0 V as varying this voltage results in changing the sensor characteristics toward different contaminants.

Three sensor values are recorded at the same time intervals at various locations in close proximity for a single mosque. Mean absolute value is calculated across two sensors using Eq. 3, as follows:

$$\mu_{*\# } = \frac{1}{m}\mathop \sum \limits_{i = 1}^{m} \left| {\frac{{x_{SV*} - x_{SV\# } }}{{x_{SV*} }}} \right|$$

where * indicates the reference sensor, and # represents the other sensor. For example, evaluating the mean absolute value of the sensor 1& 2 with sensor 1 as reference across all sensor values (m) is given as,

$$\mu_{12} = \frac{1}{m}\mathop \sum \limits_{i = 1}^{m} \left| {\frac{{x_{SV1} - x_{SV2} }}{{x_{SV1} }}} \right|$$

A lower value indicates a more stable environment and uniformly loaded rack. Furthermore, relatively less difference in the \(\mu_{*\# }\) and \(\mu_{\# *}\) indicates consistent data monitoring in case of balanced loading conditions.

Results and discussion

Time intervals are set to 141 ranging from 12:20 to 13:30. Results are plotted only for M1 with Duhar prayer for the sake of brevity. Figure 2a–c illustrates the TVOC concentration levels for M1-SV1, M1-SV2 and M1-SV3 during the three randomly selected monitoring sessions, respectively. Figure 2 shows that TVOC concentration starts increasing to initial peak within first 40 time intervals (20 min) i.e., zone 1 in. Figure 2a–c. Mosques are fully air-conditioned, however passive ventilation is assumed as the sensing station is isolated as a separate chamber. However, temperature variation is found to be negligible (variation of 1.5 °C on average), while the humidity variation is found to be 14%. Under these conditions, TVOC concentration levels reflect the effects of VOC activity caused by shoe and socks stored within the rack in addition to the response time of the sensor. The subsequent drop is associated with the rapid drop in the number of visitors right after the jamaat (zone 2 in Fig. 2a–c). Relatively steep drop is attributed to the quick and repeated mixing of air from the active ventilation area to the sensing station during the process of picking shoes from the racks by the visitors. Variation in the M1-SV2 and M1-SV3 is found to be quite higher due to the same reason. Being closer to the door, number of shoe-placings and removals, observed to much higher, contributed to higher variations.

Fig. 2

TVOC measurements for TGS2602 for three randomly selected days at Duhar session

Figure 3 shows the mean absolute sensor values (\(\mu_{12} , \mu_{21} ,\mu_{23} , \mu_{32} , \mu_{13} , \mu_{31}\)) for the M1. values in between the same sensor groups show the %age difference as the reference sensor is changed. For example, as we change the reference sensor from sensor 1 to sensor 2, 1.4% variation is observed in the value of mean absolute sensor value. The lower difference values of 1.4%, 0.1% and 0.9% among various sensor groups reflect that the sensor values are consistent in balanced rack conditions for the monitored session.

Fig. 3

Mean absolute sensor values for M1

Boxplots in Fig. 4 cover the 20-days data for the M1-SV1, M1-SV2 and M1-SV3. Although the mean concentration of the TVOC is below 3 ppm, there are some spikes of higher concentration as illustrated by the outliers in Fig. 4. This is caused by the cleaning activity performed by the helper within last 15 min of the monitoring session at M1. Use of cleaning agents and aerosol sprays in the vicinity of the station results in an immediate rise just before the first monitoring session of the day is automatically closed at 13:30. Since the same helper performs this cleaning activity at the M2, these spikes are not observed as the cleaning activity was performed after the monitoring period. Cleaning activity at M3 was not monitored closely due to administrative issues. However, data reflects that no activity of cleaning was performed during the monitoring session.

Fig. 4

Boxplots for M1 depicting the variation across days of measurement


Prolonged activity of the VOCs in confined environments is not only unpleasant but also poses certain health issues. This study aimed at identification of the need to control VOCs activity inside mosques of the professional environments within an acceptable limit by employing a certain configuration for low-cost array of sensors. The configuration depends on a number of factors such as mosque space, ventilation, air conditioning and potential quantity of VOCs generated on average. The average is estimated by sampling the visitors with different attributes. A proportionate sized chamber is used to simulate the area under consideration. The study creates an awareness in the community for the health that could arise in public places within confined/closed areas under specific environmental conditions using a network of low-cost intelligent sensor arrays. A future prospect is to conduct a technical and economic feasibility on the adoption of a control strategy by establishing an acceptable threshold level for the VOCs activity. A comparative study of the data with the international/national standards will help to improve the living/working conditions. Lastly, a control strategy/measure can be proposed to mitigate the negative effects of the contaminants.


  1. 1.

    Yu C, Crump D (1998) A review of the emission of VOCs from polymeric materials used in buildings. Build Environ 33(6):357–374

    Article  Google Scholar 

  2. 2.

    Klepeis NE et al (2001) The National Human Activity Pattern Survey (NHAPS): A resource for assessing exposure to environmental pollutants. J Expo Anal Environ Epidemiol 11(3):231–252

    Article  Google Scholar 

  3. 3.

    Yang CT, Chan YW, Liu JC, Lou BS (2020) An implementation of cloud-based platform with R packages for spatiotemporal analysis of air pollution. J Supercomput 76(3):1416–1437

    Article  Google Scholar 

  4. 4.

    Al-Homoud MS, Abdou AA, Budaiwi IM (2009) Assessment of monitored energy use and thermal comfort conditions in mosques in hot-humid climates. Energy Build 41(6):607–614

    Article  Google Scholar 

  5. 5.

    Traxler S et al (2018) VOC breath profile in spontaneously breathing awake swine during Influenza A infection. Sci Rep 8(1):1–10

    MathSciNet  Article  Google Scholar 

  6. 6.

    Rasli NBI et al (2019) Compliance of indoor air contaminants within the main prayer halls of mosques in malacca with malaysia’s indoor air quality standard. J Constr Dev Ctries 24(2):105–121

    Google Scholar 

  7. 7.

    Rasli NBI et al (2019) Indoor Chemical Air Contaminants in Main Prayer Hall during Jumaat Prayer in Mosques with Different Mechanical Ventilation. Int J Integr Eng 11(1):84–91

    Article  Google Scholar 

  8. 8.

    Curran AM, Rabin SI, Prada PA, Furton KG (2005) Comparison of the volatile organic compounds present in human odor using SPME-GC/MS. J Chem Ecol 31(7):1607–1619

    Article  Google Scholar 

  9. 9.

    Dormont L, Bessière JM, Cohuet A (2013) Human Skin Volatiles: A Review. J Chem Ecol 39(5):569–578

    Article  Google Scholar 

  10. 10.

    Denawaka CJ, Fowlis IA, Dean JR (2016) Source, impact and removal of malodour from soiled clothing. J Chromatogr A 1438:216–225

    Article  Google Scholar 

  11. 11.

    E. Kristiani, C. F. Lee, C. T. Yang, C. Y. Huang, Y. T. Tsan, and W. C. Chan, “Air quality monitoring and analysis with dynamic training using deep learning,” J. Supercomput., no. 0123456789, 2020.

  12. 12.

    Spinelle L, Gerboles M, Kok G, Persijn S, Sauerwald T (2017) Review of portable and low-cost sensors for the ambient air monitoring of benzene and other volatile organic compounds. Sensors (Switzerland) 17(7):1520

    Article  Google Scholar 

  13. 13.

    Szulczyński B, Gębicki J (2017) Currently commercially available chemical sensors employed for detection of volatile organic compounds in outdoor and indoor air. Environ - MDPI 4(1):1–15

    Google Scholar 

  14. 14.

    Al-ajmi FF (2010) Thermal comfort in air-conditioned mosques in the dry desert climate. Build Environ 45(11):2407–2413

    Article  Google Scholar 

  15. 15.

    Al-Dabbous AN, Khan AR, Al-Rashidi MS, Awadi L (2013) Carbon dioxide and volatile organic compounds levels in mosque in hot arid climate. Indoor Built Environ 22(2):456–464

    Article  Google Scholar 

  16. 16.

    A. M. Collier-Oxandale, J. Thorson, H. Halliday, J. Milford, and M. Hannigan, “Understanding the ability of low-cost MOx sensors to quantify ambient VOCs,” Atmos. Meas. Tech. Discuss., pp. 1–29, 2018.

  17. 17.

    Shen Z, Zhang X, Ma X, Mi R, Chen Y, Ruan S (2018) The significant improvement for BTX (benzene, toluene and xylene) sensing performance based on Au-decorated hierarchical ZnO porous rose-like architectures. Sensors Actuators, B Chem 262:86–94

    Article  Google Scholar 

  18. 18.

    Anaf W, Leyva Pernia D, Schalm O (2018) Standardized indoor air quality assessments as a tool to prepare heritage guardians for changing preservation conditions due to climate change. Geosci. 8(8):276

    Article  Google Scholar 

  19. 19.

    Wolfrum EJ, Meglen RM, Peterson D, Sluiter J (2006) Metal oxide sensor arrays for the detection, differentiation, and quantification of volatile organic compounds at sub-parts-per-million concentration levels. Sensors Actuators, B Chem 115(1):322–329

    Article  Google Scholar 

  20. 20.

    Shen JH et al (2020) Spatial air index with neighbor information for processing k-nearest neighbor searches in IoT mobile computing. J Supercomput 76(8):6177–6194

    Article  Google Scholar 

Download references


This project was funded by the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah, under grant no (G:1553-135-1440). The authors, therefore acknowledge with thanks DSR for technical and financial support.

Author information



Corresponding author

Correspondence to Atif Shahzad.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Shahzad, A., Hameed, A.Z. & Basahel, A. Identification and control of the volatile organic compounds activity in confined environments (Mosques). J Supercomput 77, 8716–8727 (2021).

Download citation


  • Air contaminants
  • VOC sensors
  • Low-cost computing
  • Intelligent sensing
  • Air quality monitoring
  • Soiled clothing