AtmosphericPollutionResearch1(2010)94Ͳ101
Atmospheric Pollution Research
www.atmospolres.com
Spatialandtemporalvariationofoutdoorandindoorexposureof volatileorganiccompoundsinGreaterCairo SilkeMatysik1,AbouBakrRamadan2,UweSchlink3 1
DepartmentofClinicalChemistryandLaboratoryMedicine,UniversityHospitalRegensburg,FranzͲJosefͲStraussͲAllee11,DͲ93053Regensburg,Germany NationalCentreforNuclearSafety,AtomicEnergyAuthority,Cairo,Egypt 3 HelmholtzCentreforEnvironmentalResearchͲUFZ,Permoserstraße15,04318Leipzig,Germany 2
ABSTRACT
Monthlymeasurementsofvolatileorganiccompounds(VOCs)werecarriedoutinafieldstudyinGreater Cairo in the period of 2005–2007. Ten apartments were chosen as sites for passive sampling of VOCs in different parts of Greater Cairo, Egypt taking into consideration the traffic volume profile across the city andthesuburbs.Theconcentrationsof29VOCs,belongingtothegroupsofalkanes,cycloalkanes,aromatic hydrocarbons, halogenated hydrocarbons and terpenes were measured indoor and outdoor at the same time. A strong correlation was found between sites with high traffic volume and the concentration of aromatic compounds. The measured VOC burden in the outdoor environment of the whole city is dominatedbycompoundsrelatedtovehicleemissions,e.g.theBTEXcompounds(benzene,toluene,ethyl benzeneandxylene).ThespatialvariationoftheindoorBTEXexposureissimilartotheoutdoorexposure. Thisimpliesastronginfiltrationofoutdoorambientairpollutionintotheindoorenvironment.Thehighest 3 indoor BTEX concentration was measured in the city centre (172 μg/m ). Significant seasonal cycles with highvaluesinwinterandlowervaluesinsummermonthscouldbemodelledforindoorconcentrationsof BTEX and terpenes. This is in accordance with findings in Middle Europe (Germany) with completely differentclimateandbuildingcharacteristics.
©Author(s)2010.ThisworkisdistributedundertheCreativeCommonsAttribution3.0License.
Keywords: VOC BTEX Indoorairpollution Seasonalvariation GreaterCairo
ArticleHistory: Received:23December2009 Revised:18March2010 Accepted:21March2010
CorrespondingAuthors:
SilkeMatysik,UweSchlink Tel:+49Ͳ941Ͳ9446281 EͲmail:
[email protected]Ͳregensburg.de
doi:10.5094/APR.2010.012
1.Introduction Megacities with several million inhabitants and an everͲ growing urbanization are typical for serious air pollution resulting from high traffic volume, combustion processes and industrial areas nearby. Besides SO2, NOX, and particulate matter, volatile organiccompounds(VOCs)areanimportantclassofpollutantsdue totheirstrongassociationwithadversehealtheffects(Woodruffet al., 1998). Especially the BTEX aromatics (benzene, toluene, ethylbenzene,andxylene)shouldbeconsideredindetailbecause oftheirharmfulpotentialandtheirstrongcorrelationwithmotor vehicleexhaustemissions.Additionally,VOCsaresuspectedtobe involvedincausingallergicdisorders(atopiceczemaandasthma), especially in combination with long–term exposures and/or early childhood(Herbarthetal.,2002;Rumchevetal.,2004;Herbarthet al.,2006). Thehugevehicleemissionsasthemainpartofairpollutionin Cairohavebeenknownforseveralyearsandafewstudiestowards the concentrations of VOCs and particulate matter have been published (Doskey et al., 1999; Abu–Allaban et al., 2002). As published recently, the ambient BTEX concentrations in Cairo are amongthehighestworldwide(Khoder,2007). Asmostpeoplespendthemajorityoftimeindoorstheindoor exposureplaysanimportantroleintheoverallpersonalexposure and therefore in any health risk assessment. In industrialised countriesatmoderateclimate(e.g.Germany)pollutantsinindoor air are accumulated due to central heating and low air exchange
rates,especiallyinwintermonths.Asaconsequence,inregionsof moderate climate indoor concentrations of VOCs can exceed the outdoor burden (Begerow et al., 1995; Rehwagen et al., 1999; Mitchell et al., 2007). A typical seasonal cycle of the indoor total VOCs was found in German apartments (Rehwagen et al., 2003; Schlinketal.,2004).Thesefindingsarepresumablynotnecessarily validfortheNorthernEgyptwhichischaracterisedbysubtropical climate and a different lifestyle. However, there is a lack of informationabouttheindoorVOCexposurescenarioinCairoand the influence of the high outdoor BTEX concentrations. There is only one study published on indoor exposure in Cairo which evaluates the concentrations of aromatic hydrocarbons in office buildings (Khoder, 2006). More detailed data about the outdoor andindoorVOCsituationacrosstheGreaterCairoareaareneeded to perform risk assessments in the future. In contrast, there are extensive studies in Europe to evaluate systematically the relationshipbetweenindoorairpollutionandhumanexposureto pollutants,e.g.theAIRMEXproject(Kotzias,2005). In the present study the concentrations of 29 VOCs (BTEX included)weremeasuredat10differentsitesinCairotoachievea spatial assessment of the VOC burden across the Greater Cairo. The measurements were done outside and inside apartments for severaltimesoveraone–yearperiodtoinvestigatetheseasonality oftheVOCconcentrations.Toourbestknowledge,anassessment oftheVOCexposureinapartmentsinCairowasdoneforthefirst time.
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
2.Methods 2.1.Fieldstudies Urbanizationandindustrializationhaveincreasedveryrapidly in Greater Cairo. The Greater Cairo area has about 15million inhabitants. For the Greater Cairo area, the statistics of the Egyptian government reported 1735724 vehicles in April 2003. The number of vehicles has increased by 10% per annum. About 64% of the industries in Egypt are concentrated in Greater Cairo area.ThesefactsintensifytheproblemofcontaminationofCairo’s environment with various impurities and environmental hazards (Robaa,2006). Climatologically,Cairofollowsasubtropicalclimate.According to the characteristic meteorological conditions of Egypt, the average daily mean, maximum, and minimum temperatures over Cairorangefrom13.8,19.7and8.8°C,respectively,inJanuary,to 27.9, 34.9 and 22.3°C, respectively, in July. Wind speed ranges from 2.8 kt during October to 5.6 kt during May. The prevailing windsareN,NWandW,withpercentageofoccurrence31.8,12.9, 12.8, respectively. These wind directions could cause rapid transportation of pollutants and other urban wastes from the adjacent industrial area of Shubra El–Kheima to the urban Cairo area. In summer, strong vertical temperature gradients lead to enhanced vertical convection in this area, carrying air pollutants, such as organic compounds and aerosols, aloft and enhancing turbidity(Zakeyetal.,2004). Weselected10apartmentsassamplingsitesindifferentparts ofCairoasshowninFigure1:ElHaram(Pyramides)fromthewest, El Mokatam hill and Nasr city from the east and Helwan and El Maadi from the South, Shubra, Heliopolice and El Obour from north and north east and El Kasr El Ani and El Attaba from the centre. These sites are representing different microenvironments (streets, near industries, urban areas and residential areas). The 10ambient air monitoring sites are classified according to the predominant land use surrounding the site (see Table 1). Seven sites are in residential areas influenced by industries nearby or traffic. Three sites were considered as pure residential sites (El Mokatam,Nasr cityandEl Obour).Thesampleswerecollectedin the period from September, 2005 to March, 2007. Generally, samplingwascarriedoutonceamonthinSeptember,December, January, February, April, May, August, September, November and December for the sites 1–9. In all cases, data were collected for indoor and outdoor microenvironments. Because of technical restrictions sampling in site No.10 (El Obour) was performed only threetimes.Intotal,185samplesweretaken. 2.2.Sampling ForpassivesamplingofVOCspassivediffusionmonitors(type OVM 3500, 3M Company, Canada) were used. During a 4–week periodthemonitorswereexposedtotheindoorandoutdoorairof rooms,whichwererepresentativeforthesamplingsite.Generally, thesamplerswerehunginthemiddleoftheroomataheightof2– 2.5m. TheresultsbyBegerowetal.(1999)clearlydemonstratedthe usefulness of passive sampling for the determination of VOC concentrations in indoor and outdoor air at environmental conditions. For air monitoring down to the ng/m³ range usually longsamplingperiods,uptoseveralweeksareapplied.According to Shields and Weschler (1987) even after an 8–week sampling period at environmental conditions, no overload of the passive samplersoccurred. 2.3.Samplepreparationandgaschromatographicanalysis Afterexposure,theVOCsweredesorbedfromtheadsorption layers (charcoal pads) by means of carbon disulfide (promochem
95
GmbH, Wesel, Gemany) which contained the internal standards cyclododecaneandbenzene–d6(bothanalyticalgradequalityfrom Fluka, Deisenhofen, Germany). After the extraction period of 30minthesolutionwasdecantedintoGCvialsforanalysis.
Figure1.MapofGreaterCairowithsamplingsites.
The VOC analysis was performed on a Perkin Elmer GC–MS system equipped with an RTX–1 column (Restek; 60mx0.32mm I.D., 1.0μm film thickness). The oven temperature was held at 43°C for 5 min, and then programmed to 200°C at a rate of 2.5°C/min. A sample volume of 1μl was injected in the splitless mode.Integratedareasofselectedfragmentionsfromeachofthe 29VOCswereobtainedwiththesoftwareTurbomass,Version4.4 (PerkinElmer). After calibration of all compounds of interest, the adsorbed amounts of VOC were calculated according to the 3M Technical Bulletin1028.Adetaileddescriptionofcalibration,calculation,and the analytical parameters is given in Rehwagen et al. (2003). Recovery coefficients were determined by direct injection of a known amount of the standard into a 3M sampler. The recovery was between 98 and 102%. Detection limits were found to range from 0.01 to 0.05 μg/m3 for a sampling interval of 4 weeks. The relativestandarddeviationwastypicallybelow10%. 2.4.Statisticalmethods For statistical processing, GenStat 10.1 was applied (Payne, 2000). The analysis was based on the median and its 95% confidenceintervalofthemeasuredVOCconcentrations. To calculate the rank values qu ,l of upper and lower confidencelimitsforthemedianofmonthlysubgroupsweused the normal approximation qu ,l | E >q @r u10.05 / 2 var>q @ with
u10.05 / 2 1.96 representing the 95th percentile of the standard normal distribution. This approximation is nearly valid as the samplesizesare N t 9 (Lienert,1978). As percentiles follow a binomial distribution (Lienert, 1978), forthemedian(whichisthe50thpercentile,i.e.,q=1/2)we
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
96
Table1.VOCsumbasedon29VOCandBTEXconcentrationsfromoutdoorandindoormeasurementsacrossGreaterCairo (September2005–March2007).Descriptionofthemicroenvironment:R=residential,C=commercial,T=traffic,I=industrial
Attaba(7)
R,C,T
Sampling Frequency 10
ElKasrElAni(1)
Site(SiteNo.)
Description
3
Outdoor median 343.8
VOCsum(μg/m ) Outdoor Indoor average median 381.1 364.9
3
Indoor average 381.3
Outdoor median 204.0
BTEX(μg/m ) Outdoor Indoor average median 211.4 171.8
Indoor average 178.9
R,C,T
10
226.4
214.1
207.0
184.7
130.9
118.0
105.4
95.4
ElMaadi(5)
R,T
10
175.7
196.6
223.6
245.7
98.7
108.0
65.0
73.5
Helwan(6)
R,I
10
152.5
233.7
180.2
336.9
58.1
54.8
53.0
54.0
ElHaram(2)
R,T
10
144.7
276.2
219.9
272.1
68.9
65.9
62.0
79.1
Heliopolice(3)
R,T
10
92.4
97.5
167.1
296.9
44.4
48.6
50.1
58.6
Shubra(8)
R,I
10
73.9
92.5
240.2
264.5
36.6
43.0
34.1
41.0
ElMukatam(9)
R
10
80.1
91.8
124.6
159.4
29.1
29.6
29.1
36.4
Nasrcity(4)
R
10
42.0
49.3
86.3
103.5
16.9
23.7
30.5
31.9
ElObour(10)
R
3
36.7
61.2
56.2
91.2
15.8
20.6
16.8
17.6
can specify the expectation value E >q @ Nq N / 2 and the variance var>q @ Nq1 q N / 4 . Asthedatadidnotfollowanormaldistributionfunction,we used the Mann–Whitney U–test and the H–test of Kruskal–Wallis to detect significant differences between groups of data. These testsreferonlytothestatisticsofthedataanddonotincludethe measurementerrors. ToassesstheamplitudeoftheobservedannualcycleofVOCs and to enable a conversion between concentrations observed in differentmonths,wefittedaseasonalmodeltothelogarithmsof the monthly medians of concentrations. The predictors are harmonicfunctionsofthetime(minmonths)andthecoefficients havebeenestimatedbymeansofregressionanalysis.
3.ResultsandDiscussion 3.1.SpatialvariationofoutdoorandindoorVOCs The outdoor and indoor VOC sum concentrations (median values)resultingfrom9 independentmeasurementsforeachsite aregiveninTable1.ThisVOCsumcontainsthealkanesC6–C16,the cycloalkanes methylcyclopentane, cyclohexane, and methylͲ cyclohexane, the aromatic hydrocarbons benzene, toluene, ethylbenzene, o–, m–, p–xylene, styrene, 4–, 3–, 2–ethyltoluene, naphthalene, and methylbenzoate, the chlorinated hydrocarbons chlorobenzene,trichloroethylene,andtetrachloroethyleneandthe terpenesɲ–pinene,ɴ–pinene,ȴ3–carene,andlimonene. It is obvious from Table 1 that the outdoor concentration of those VOCs which we measured is highest in the city centre (e.g. site No.7), lower in locations a few kilometres away from the centre(siteNo.3),andlowestindistantareas(siteNo.10).Thisisin accordancewiththefindingsbyKhoder(2007)whoalsofoundvery highvaluesintheinnercityanddecreasingvaluesinruralorpurely residentialareas.SimilarresultswereobtainedforVOCmonitoring in and around Athens, which is also characterized by MediterͲ raneanclimate(Rappenglucketal.,1998).Thisimpliesthefactthat thehighairpollutionisspreadingoutfromthecitycentre to the bordering area by transport. A similar hypothesis was also suggestedbyIlgenetal.(2001). Because of their health relevance and important role in the photochemical formation of ozone (Atkinson, 2000) the BTEX profilesweretakenintodeeperconsiderations.Especiallybenzene is considered as one of the most important compounds for this
studyduetoitscarcinogenicity(WHO,1993).Themajorhealthrisk associatedwithexposuretobenzeneisleukaemia. Table 1 shows also a clear local dependence of high BTEX values. Sites in the city centre are more burdened than sites in pureresidentialregions.ThesitesNo.1andNo.7arecharacterised byaBTEXexposurewhichis2.5to20timeshigherthanreported for Europe and North America, both indoor and outdoor (Rehwagenetal.,2003;Payne–Sturgesetal.,2004;Jiaetal.,2008; Iovino et al., 2009). From the indoor values in Table 1, it can be concludedthatthespatialvariationoftheindoorBTEXexposureis similar to the outdoor exposure. The indoor concentrations of BTEX in the city centre (e. g. Attaba and El Kasr El Ani) were 3–5 timeshigherthanthosemeasuredatdistantsites.Asanexample, an average indoor benzene concentration of 17.2 μg/m3 was observedforsiteNo.7incontrasttositeNo.10with3.0μg/m3.The fact,thattheindoorairespeciallyintheinnercityareaisburdened byBTEXtosuchahighextentisanalarmingaspectsincemostof the office buildings, banks, and shopping centres with a high numberofemployersarelocatedinthecitycentre. To assess the possible sources of the BTEX exposure, benzene/tolueneratioscanbecalculated.Anaveragevalueof0.35 was calculated for the benzene/toluene ratio for all the outdoor measurements.Thisisingoodagreementwithrecentlypublished measurements in Cairo by Khoder (2006) who also found a benzene/toluene ratio of 0.35 in ambient air. It is known from studies performed in the United States using a chemical mass balance receptor model, vehicle exhaust results in a benzene/toluene ratio of 0.5 (Scheff and Wadden, 1993). Thereforeitcanbeassumedthatmotorvehicleemissionisoneof thesourcesforthehighBTEXexposureinCairo. Thebenzene/toluenedistributioninsidebuildingsisnearlythe same as outside. An average value of 0.30 was calculated for the benzene/tolueneratioforallindoormeasurements.Thisimpliesa strong infiltration from the outdoor ambient BTEX pollution into the indoor microenvironment with only small effects from additionalindoorsourcesoftoluene.TheindoorexposureofBTEX is therefore dominated by the outdoor exposure. To verify the occurrence of a transfer from outdoor air into the indoor microͲ environment matched pairs of concentrations of benzene that were measured simultaneously, were analyzed. The positive correlation between outdoor and indoor benzene concentrations (r=0.92) is shown in Figure 2. A reasonable approximation is a linearrelationshipthatwasfittedbylinearregression.Asaresult, the monthly indoor concentration of benzene can be predicted fromtheoutdoorconcentrationaccordingto:
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
3
3
Cbenzene,indoors(μg/m )=(1.59±0.40)+(0.72±0.03).Cbenzene,indoors(μg/m )
(1)
benzene indoors / µg/m³
There are only three indoor air benzene concentrations that exceedthenormallevelandthismightbeattributedtohouseholds with smokers. In general, a differentiation between smoker and non–smokermicroenvironmentswasnotdonebecauseofmissing statementsoftheresidents. 35 30
97
Thecompoundswitha p–value<0.05mightfollowaseasonal function over 12 months. To demonstrate this behaviour and to enableaconversionbetweenconcentrationsmeasuredindifferent months, a seasonal model to the logarithms of the monthly mediansofconcentrationscouldbefittedforallcompoundswitha Kruskal–Wallis p–value<0.05. The maximum/ minimum ratios of the monthly concentrations (median values of all sites) and the summer/winterratios(seeTable2)ascribetheamplitudeofthese annualcycles.Itshouldbenotedthatthemodelbasedonlyonthe medianvaluesastheyarerobustwithoutliers.Thepredictorsare harmonic functions of time. For benzene, Equation (2) describes this harmonic function for the outdoor values. R2 expresses the fitnessofthemodel. 3
logCbenzene,outdoors(median)(μg/m )=2.00+0.40.cos(2ʋm/12) 2 +0.13.sin(2ʋm/12)(R =0.82)
25 20 15 10 5 0 0
5
10
15 20 25 30 benzene outdoors / µg/m³
35
40
45
Figure 2. Association between benzene exposure indoors and outdoors in apartmentsinGreaterCairo(n=84).
In general, the indoor VOC exposure is affected both by infiltration from the outdoor air and personal activities of the residents (Dodson et al., 2007). The latter includes everything related to the indoors such as furniture, carpets, paints and varnishes(HerbarthandMatysik,2010)andtothelifestylesuchas the use of cleaning detergents and smoking. This explains the discrepancies between the 29VOC sum concentrations measured indoorsandoutdoorsaspresentedinTable1,inparticularforthe site No.8. However, the measurement of only 29 selected VOCs cannotgiveageneralpictureoftheVOCexposuresincealdehydes, organicacids,otherpolarcompoundsandlow–boilingcompounds cannotbecollectedandanalysedreliablybyusingthe3Mpassive samplingdevices(ShieldsandWeschler,1987). 3.2.TemporalvariationofoutdoorandindoorVOCs Seasonal variations of the outdoor concentrations of VOCs were observed in Europe, Asia and North America (Cheng et al., 1997; Morikawa et al., 1998; Na and Kim, 2001). The indoor VOC concentrations undergo in continental and temperate climates seasonally–relateddeviations(Rehwagenetal.,2003;Schlinketal., 2004)whichcouldbeexpressedaspronouncedcycles. Multiple VOC measurements were performed in Cairo in the course of two years to examine the seasonal variations in indoor andoutdoorVOCexposure(seeSection2.1). At first, the seasonal effect was studied on the outdoor concentrationsbymeansoftheKruskal–Wallistest.Notethatthe test was conducted using the all data set that included the uncertaintiesofmonthlyobservations.Avalueof p<0.05indicates statistically significant differences in the VOC concentrations between monthly measurements. The annual median, the confidence limits, and the Kruskal–Wallis p–values are given in Table2.
(2)
where m denotes the month of the year (between 1 and 12, JanuaryandDecember,respectively. Thesummer/winterratiosindicatestatisticallyhigherconcenͲ trationsinwinterthaninsummerforallcompounds(p<0.05). Thesamecalculationsweredoneforallindoordata(median valuesofallsamplingsites)asexpressedinTable3.Annualcycles for the indoor concentrations were modelled according to the procedurepresentedabove. The p<0.05 criterion was fulfilled for the indoor concentrationsofC6–C8alkanes,cyclohexane,methylcyclopentane, methylcyclohexane, ȴ3–carene, ɴ–pinene, limonene, benzene, ethylbenzene, propylbenzene, toluene, m–, p–, o–xylene and styrene. The lowest summer/winter ratio was obtained for limonene indicating a seasonal cycle with pronounced amplitude. The seasonal cycles of the indoor concentrations of benzene, the BTEX sum and the terpene sum of ȴ3–carene, ɴ–pinene, and limonenearepresentedinFigure3. The predictors for the indoor concentrations are given in Equations(3),(4),and(5): log >median (c benzene )@ 1.85 0.48 cos(2Sm / 12) 0.25 sin(2Sm / 12)
>
log median (cterpenes )
@
2.38 1.47 cos(2Sm / 12) 0.95 sin(2Sm / 12)
log >median (cBTEX )@ 3.82 0.62 cos(2Sm / 12) 0.13 sin(2Sm / 12)
(3) (4) (5)
Themaincharacteristicsofthesefunctions,i.e.highervalues in winter and lower values in summer are similar to the results obtainedinEurope(Rehwagenetal.,2003).Thelowerindoorand outdoor concentrations of BTEX in summer months can be explained by a faster degradation process in the lower troposphere. Benzene and alkyl–substituted benzenes react with OH radicals and NO3 radicals, with the OH radical reaction dominatingasthetroposphericremovalprocess(Atkinson,2000). This process can only partly be responsible for the monthly deviations,inparticularforbenzene.Othersporadicandcomplex influencesmightalsoplayarole,e.g.industrialemissionsfromthe industrialareasnearbyandthemeteorologicalconditions.
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
98
Table2.StatisticalresultsofoutdoorVOCsmeasuredbetweenSeptember2005–March2007inGreaterCairo. n=10forninesamplingsitesandn=3foronesamplingsite lower confidence limit (μg/m³)
annual median (μg/m³)
upper confidence limit (μg/m³)
max/minof monthly median
Summer/ Winter ratio
Kruskal– Wallisp– value
R²of modeled harmonic function
Hexane
6.66
9.06
10.70
3.32
0.23
0.005
Heptane
3.9
5.27
7.17
3.70
0.20
0.018
Compound
Octane
2.23
2.66
3.71
3.57
0.17
0.01
Nonane
1.91
2.32
3.15
3.31
0.23
0.011
Decane
2.63
3.13
3.89
2.07
0.40
0.524
Undecane
2.14
2.51
2.85
2.15
0.36
0.477
Dodecane
3.15
3.55
4.38
1.94
0.92
0.082
Tridecane
1.33
1.62
1.99
2.71
0.76
0.194
Tetradecane*
2.17
2.73
3.63
3.45
1.31
0.087
Pentadecane*
0.92
1.1
1.3
2.07
0.44
0.33
Hexadecane*
1
1.25
1.46
1.97
0.99
0.097
Pentamethylheptane
0.02
0.02
0.03
2.55
0.33
0.085
Heptamethylnonane
0.08
0.09
0.1
2.02
0.89
0.005
Cyclohexane
1.05
1.41
1.93
5.09
0.15
0
Methylcyclopentane
2.51
3.32
4.57
3.59
0.21
0.001
Methylcyclohexane
2.11
2.59
3.32
5.47
0.15
0.001
0.79
ѐ3ͲCarene
0.13
0.15
0.25
7.33
0.12
0.003
ɲͲPinene
1.04
1.21
1.58
2.92
2.21
0.098
ɴͲPinene
0.16
0.2
0.23
3.34
0.17
0.021
0.74
Limonene
1.28
1.68
2.14
16.34
0.17
0
0.94
Benzene
6.16
7.81
10.49
2.87
0.28
0.001
0.82
Ethylbenzene
2.5
3.07
4.35
3.14
0.29
0.022
0.89
Propylbenzene
0.81
1
1.27
3.55
0.18
0.07
Isopropylbenzene
0.3
0.35
0.47
3.68
0.16
0.098
1,2,3ͲTrimethylbenzene
1.05
1.37
1.67
3.63
0.16
0.091
1,2,4ͲTrimethylbenzene
3.73
5.2
6.65
3.83
0.18
0.101
1,3,5ͲTrimethylbenzene
1.09
1.52
1.86
3.88
0.16
0.156
Toluene
16.32
22.84
30.39
3.26
0.23
0.04
2ͲEthyltoluene
0.83
1.13
1.4
3.76
0.19
0.065
3ͲEthyltoluene
1.08
1.23
1.89
3.70
0.23
0.061
4ͲEthyltoluene
2.21
2.74
4.03
3.73
0.23
0.075
m,pͲXylene
8.82
11.57
16.43
3.73
0.26
0.027
0.84
oͲXylene
2.83
3.88
5.23
4.01
0.24
0.024
0.83
0.82
Styrene
0.33
0.46
0.57
4.03
0.21
0
Naphthalene
0.21
0.25
0.38
4.35
0.11
0.052
pͲDichlorobenzene
0.07
0.09
0.11
2.55
1.24
0.289
Methylisobutylketone
0.06
0.07
0.08
2.04
0.48
0.474
*ValuesmeasuredinMaywereomitted
The main emission source of lower alkanes has been vehicle exhaust and gasoline evaporation. The indoor exposure of these alkanes is therefore dominated by the outdoor exposure. In contrast, alkanes with longer chain length are mostly emitted indoorsbytheinteriormaterials,e.g.paintandvarnish(Herbarth andMatysik,2010).ThiscouldexplainwhyonlytheC6–C8alkanes showsignificantmonthlydifferences. It can be seen from the R2 values in Table 3 that a more pronouncedseasonalcyclecouldbemodelledfortheterpenes
thanforthearomatichydrocarbons(BTEX).Thereisnoexplanation why ɲ–pinene does not follow this function. In general, terpenes are strongly related to indoor sources (Nazaroff and Weschler, 2004). The high terpene values in winter months indicate that indooractivitiesoftheresidentsinbuildingsinCairoaresimilarto citiesinEurope.Thatwasnotexpectedbecauseofthesubtropical climate and the different kinds of building constructions and life styleinEgypt.
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
99
Table3.StatisticresultsofindoorVOCsmeasuredbetweenSeptember2005–March2007inGreaterCairo. n=10forninesamplingsitesandn=3foronesamplingsite
lower confidence limit (μg/m³)
annual median (μg/m³)
upper confidence limit (μg/m³)
max/minof monthly median
Summer/ Winter ratio
Kruskal– Wallisp– value
R²of modelled harmonic function
Hexane
6.46
7.63
10.01
3.32
0.23
0.002
Heptane
4.21
4.84
6.07
3.54
0.23
0.002
Octane
2.54
2.93
3.74
3.48
0.21
0.005
Compound
Nonane
2.65
3.78
4.86
4.58
0.18
0.072
Decane
4.71
5.74
8.98
3.05
0.22
0.617
Undecane
3.60
4.89
8.06
3.19
0.24
0.734
Dodecane
5.53
6.45
9.02
2.43
1.90
0.629
Tridecane
2.23
2.70
3.71
2.14
0.97
0.354
Tetradecane*
3.38
3.81
4.72
2.39
1.93
0.151
Pentadecane*
1.34
1.57
2.08
1.82
1.54
0.716
Hexadecane*
1.30
1.74
1.97
2.36
2.22
0.281
Pentamethylheptane
0.04
0.05
0.07
3.59
1.52
0.162
Heptamethylnonane
0.09
0.10
0.13
3.48
1.10
0.081
Cyclohexane
1.00
1.40
1.83
3.91
0.25
0.000
Methylcyclopentane
2.30
2.87
4.03
3.01
0.27
0.001
Methylcyclohexane
2.26
2.57
4.07
5.42
0.15
0.000
0.71
ѐ3ͲCarene
0.59
1.07
1.66
9.04
0.16
0.023
ɲͲPinene
3.62
4.93
6.39
5.46
0.72
0.162
ɴͲPinene
0.81
1.30
2.08
6.89
0.13
0.006
0.94
Limonene
5.30
9.36
17.94
31.72
0.04
0.000
0.96
Benzene
5.75
7.08
8.96
2.72
0.37
0.000
0.85
Ethylbenzene
2.68
3.18
3.85
2.32
0.37
0.014
0.89
Propylbenzene
0.93
1.10
1.27
2.63
0.28
0.044
Isopropylbenzene
0.36
0.41
0.49
2.55
0.36
0.064
1,2,3ͲTrimethylbenzene
1.32
1.60
2.18
3.16
0.24
0.127
1,2,4ͲTrimethylbenzene
4.62
5.60
7.49
3.22
0.22
0.060
1,3,5ͲTrimethylbenzene
1.41
1.69
2.18
3.34
0.21
0.177
Toluene
19.40
23.40
29.16
3.65
0.25
0.005
0.88
2ͲEthyltoluene
1.02
1.23
1.61
3.03
0.24
0.063
3ͲEthyltoluene
1.14
1.38
1.97
2.48
0.35
0.100
4ͲEthyltoluene
2.37
2.83
4.12
2.80
0.31
0.071
m,pͲXylene
9.68
11.68
13.69
2.23
0.38
0.015
0.88
oͲXylene
3.30
3.88
4.66
2.18
0.43
0.015
0.86
Styrene
0.56
0.68
0.78
3.15
0.37
0.001
Naphthalene
0.35
0.43
0.57
2.70
0.81
0.400
pͲDichlorobenzene
0.10
0.13
0.18
3.52
1.98
0.484
Methylisobutylketone
0.09
0.11
0.15
1.56
1.13
0.953
*ValuesmeasuredinMaywereomitted
4.Conclusion ItwasdemonstratedthattheGreaterCairoareaisburdened by BTEX, especially in the city centre. In places with high traffic volume the BTEX compounds originating from vehicle exhaust emission exceed 100μg/m3. Similar concentrations for the BTEX compounds were measured inside the apartments. This is an alarming fact, especially in the case of the carcinogenic benzene. The overall personal exposure of individuals is a time function of their activities. Due to the high exposure to VOCs with high amounts of aromatic hydrocarbons in the outdoor and indoor microenvironments,thepersonaluptakeofVOCsforinhabitantsin
Cairo is critical. The seasonal dependence of the measured concentrations has to be taken into consideration to assess the indoorVOCexposureforanyhealthstudies.Inordertodothis,the measured concentrations in a certain period (e.g., summer months) have to be adjusted for other periods using appropriate cyclefactors(Schlinketal.,2004). As Cairo is characterized by warm climate with high solar radiation,specialconsiderationshouldbefocussedtomonitorthe
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
100
Acknowledgment
Financial support by the International Office of the German Ministry of Education and Research (BMBF) is gratefully acknowledged. The authors thank Ms. Martina Rehwagen for helpfuldiscussions.
References AbuͲAllaban, M., Gertler, A.W., Lowenthal, D.H., 2002. A preliminary apportionment of the sources of ambient PM10, PM2.5, and VOCs in Cairo.AtmosphericEnvironment36,5549Ͳ5557. Atkinson,R.,2000.AtmosphericchemistryofVOCsandNOx.Atmospheric Environment34,2063Ͳ2101.
(a)
Begerow, J., Jermann, E., Keles, T., Dunemann, L., 1999. Performance of two different types of passive samplers for the GC/ECDͲFID determinationofenvironmentalVOClevelsinair.FreseniusJournalof AnalyticalChemistry363,399Ͳ403. Begerow, J., Jermann,E.,Keles,T., Ranft,U.,Dunemann,L.,1995.Passive sampling for volatile organic compounds (VOCs) in air at environmentally relevant concentration levels. Fresenius Journal of AnalyticalChemistry351,549Ͳ554. Cheng, L., Fu, L., Angle, R.P., Sandhu, H.S., 1997. Seasonal variations of volatile organic compounds in Edmonton, Alberta. Atmospheric Environment31,239Ͳ246. Dodson,R.E.,Houseman,E.A.,Levy,J.I.,Spengler,J.D.,Shine,J.P.,Bennett, D.H., 2007. Measured and modelled personal exposures to and risks from volatile organic compounds. Environmental Science and Technology41,8498Ͳ8505. Doskey,P.V.,Fukui,Y.,Sultan,M.,AlMaghraby,A.,Taher,A.,1999.Source profiles for nonmethane organic compounds in the atmosphere of Cairo,Egypt.JournalofTheAir&WasteManagementAssociation49, 814Ͳ822.
(b)
Herbarth, O., Matysik, S., 2010. Decreasing concentrations of volatile organiccompounds(VOC)emittedfollowinghomerenovations.Indoor Air20,141Ͳ146. Herbarth, O., Fritz, G.J., Rehwagen, M., Richter, M., Roder, S., Schlink, U., 2006.Associationbetweenindoorrenovationactivitiesandeczemain early childhood. International Journal of Hygiene and Environmental Health209,241Ͳ247. Herbarth, O., Diez, U., Borte, M., Lehmann, I., Fritz, G., Kroessner, T., Metzner, G., Rehwagen, M., Richter, M., Schulz, R., Wetzig, H., 2002. Effectofindoorchemicalexposureonthedevelopmentofallergiesin infancy.Epidemiology13,156Ͳ157. Ilgen, E., Karfich, N., Levsen, K., Angerer, J., Schneider, P., Heinrich, J., Wichmann, H.E., Dunemann, L., Begerow, J., 2001. Aromatic hydrocarbons in the atmospheric environment: Part I. Indoor versus outdoorsources,theinfluenceoftraffic.AtmosphericEnvironment35, 1235Ͳ1252.
(c) Figure 3. Annual pattern of (a) benzene (b) terpene sum and (c) BTEX concentrations based on indoor measurements (10 sampling sites in GreaterCairo).
ozone concentration and secondary emission products inside buildings.Theseproducts areformedbythereactions of volatile organic compounds with oxidants (e.g. ozone, nitrogen dioxide)andmaybeconsiderablymoreharmfulthanthereactants (Nazaroff and Weschler, 2004; Weschler, 2006; Wolkoff et al., 2006). Products of terpene–ozone reactions are of particular interest because terpene concentrations play also an important roleinCairo’sapartments.
Iovino, P., Polverino, R., Salvestrini, S., Capasso, S., 2009. Temporal and spatial distribution of BTEX pollutants in the atmosphere of metropolitan areas and neighbouring towns. Environmental MonitoringandAssessment150,437Ͳ444. Jia, C., Batterman, S., Godwin, C., 2008. VOCs in industrial, urban and suburban neighborhoods, Part 1: Indoor and outdoor concentrations, variation,andriskdrivers.AtmosphericEnvironment42,2083Ͳ2100. Khoder, M.I., 2007. Ambient levels of volatile organic compounds in the atmosphereofGreaterCairo.AtmosphericEnvironment41,554Ͳ566. Khoder, M.I., 2006. Formaldehyde and aromatic volatile hydrocarbons in the indoor air of Egyptian office buildings. Indoor and Built Environment15,379Ͳ387.
Matysiketal.–AtmosphericPollutionResearch1(2010)94Ͳ101
Kotzias,D.,2005.IndoorairandhumanexposureassessmentͲneedsand approaches.ExperimentalandToxicologicPathology571,5Ͳ7. Lienert, G.A., 1978. NonͲparametric techniques in biostatistics. Verlag AntonHain,MeisenheimamGlan(inGerman). Mitchell,C.S.,Zhang,J.J.,Sigsgaard,T.,Jantunen,M.,Lioy,P.J.,Samson,R., Karol, M.H., 2007. Current state of the science: Health effects and indoor environmental quality. Environmental Health Perspectives 115, 958Ͳ964. Morikawa,T.,Wakamatsu,S.,Tanaka,M., Uno,I.,Kamiura,T.,Maeda, T., 1998.C2ͲC5hydrocarbonconcentrationsincentralOsaka.Atmospheric Environment32,2007Ͳ2016. Na,K.,Kim,Y.P.,2001.Seasonalcharacteristicsofambientvolatileorganic compoundsinSeoul,Korea.AtmosphericEnvironment35,2603Ͳ2614. Nazaroff,W.W.,Weschler,C.J.,2004.Cleaningproductsandairfresheners: exposure to primary and secondary air pollutants. Atmospheric Environment38,2841Ͳ2865. Payne, R.W., 2000. The Guide to GenStat. Lawes Agricultural Trust, Rothamsted. PayneͲSturges,D.C.,Burke,T.A.,Breysse,P.,DienerͲWest,M.,Buckley,T.J., 2004. Personal exposure meets risk assessment: A comparison of measured and modeled exposures and risks in an urban community. EnvironmentalHealthPerspectives112,589Ͳ598. Rappengluck, B., Fabian, P., Kalabokas, P., Viras, L.G., Ziomas, I.C., 1998. QuasiͲcontinuous measurements of nonͲmethane hydrocarbons (NMHC) in the greater Athens area during MEDCAPHOTͲTRACE. AtmosphericEnvironment32,2103Ͳ2121. Rehwagen, M., Schlink, U., Herbarth, O., 2003. Seasonal cycle of VOCs in apartments.IndoorAir13,283Ͳ291. Rehwagen,M.,Schlink,U.,Herbarth,O.,Fritz,G.J.,1999.Pollutionprofiles at different kindergarten sites in Leipzig, Germany. Environmental Toxicology14,321Ͳ327.
101
Robaa, S.M., 2006. A study of solar radiation climate at Cairo urban area, Egypt and its environs. International Journal of Climatology 26, 1913Ͳ 1928. Rumchev, K., Spickett, J., Bulsara, M., Phillips, M., Stick, S., 2004. Association of domestic exposure to volatile organic compounds with asthmainyoungchildren.Thorax59,746Ͳ751. Scheff, P.A., Wadden, R.A., 1993. Receptor modeling of volatile organicͲ compounds.1. Emission inventory and validation. Environmental ScienceandTechnology27,617Ͳ625. Schlink,U.,Rehwagen,M.,Damm,M.,Richter,M.,Borte,M.,Herbarth,O., 2004. Seasonal cycle of indoorͲVOCs: comparison of apartments and cities.AtmosphericEnvironment38,1181Ͳ1190. Shields, H.C., Weschler, C.J., 1987. Analysis of ambient concentrations of organic vapors with a passive sampler. Abstracts of Papers of The AmericanChemicalSociety193,62ͲENVR. Weschler, C.J., 2006. Ozone's impact on public health: Contributions from indoorexposurestoozoneandproductsofozoneͲinitiatedchemistry. EnvironmentalHealthPerspectives114,1489Ͳ1496. WHO, 1993. Benzene (International Programme of Chemical Safety, Environmental Health Criteria No. 150). World Health Organization, Geneva. Wolkoff, P., Wilkins, C.K., Clausen, P.A., Nielsen, G.D., 2006. Organic compounds in office environments Ͳ sensory irritation, odor, measurementsandtheroleofreactivechemistry.IndoorAir16,7Ͳ19. Woodruff, T.J., Axelrad, D.A., Caldwell, J., MorelloͲFrosch, R., Rosenbaum, A., 1998. Public health implications of 1990 air toxics concentrations acrosstheUnitedStates.EnvironmentalHealthPerspectives106,245Ͳ 251. Zakey, A.S., Abdelwahab, M.M., Makar, P.A., 2004. Atmospheric turbidity overEgypt.AtmosphericEnvironment38,1579Ͳ1591.