- Supporting information – A Comparative Study of Walking-induced Dust Resuspension using a Consistent Test Mechanism Yilin Tian1*, Kyung Sul2, Jing Qian2, Sumona Mondal3 and Andrea R. Ferro2 1 Institute for a Sustainable Environment, Clarkson University, 8 Clarkson Ave, Potsdam, NY, 13699 2 Department of Civil and Environmental Engineering, Clarkson University, 8 Clarkson Ave, Potsdam, NY, 13699 3 Department of Mathematics, Clarkson University, 8 Clarkson Ave, Potsdam, NY, 13699 Table of Content Literature review: residential surface dust loading ..................................................................... 2 Photograph of the resuspension mechanism ............................................................................... 5 Carpet type descriptions .............................................................................................................. 5 Volume and number size distributions of house dust ................................................................. 6 Photograph of the seeding system............................................................................................... 6 Summary of average resuspension fractions............................................................................... 7 Summary of ANOVA results ...................................................................................................... 7 Results of Wilcoxon rank sum test ............................................................................................. 8 Emission rate plot ..................................................................................................................... 12 Literature review: residential surface dust loading Table S1 shows a summary of surface dust loadings measured in U.S residence and offices. According to table S1, measured residential surface dust loadings varied over several orders of magnitude: from the minimum value of 0.05 g/m2 reported by Adgate et al. (1995) to the maximum value of 465 g/m2 provided by Robert et al. (2004). When reporting surface dust loadings, different size fractions were used in previous studies due to various research purposes, sampling methods and dust process methods. In the current work, the collected house dust was sieved to 44 micron to remove debris, hair and fibers which could tangle or block the dust inlet of the seeding system. There is a need to determine the fractions of less than 44 micron particles for dust sieved to 150 micron and 500 micron, respectively. Table S2 listed the average size fractions of house or office dust reported in previous studies. According to table S2, we assumed that less than 44 micron particles account for 30% and 20% of the bulk dust sieved to 150 micron and 500 micron, respectively. To estimate the proper surface dust loading for less than 44 micron particles, these fractions were applied to the median/geometric mean/average surface dust loadings reported in Table S1. The estimated median/geometric mean/average surface dust loadings ranged from 0.08 to 4.8 g/m2 (average: 1.4 g/m2; median: 0.8 g/m2). The specific Grimm 1.108 aerosol spectrometer used in the study had a reduced performance when the airborne particle concentrations were close to its detection limit of 1 particle/L. Hence, instead of the median value of 0.8 g/m2, the average value of 1.4 g/m2 (rounded to 2 g/m2) was selected as the low level of surface dust loading to reduce the error generated by the instrument. The high level of surface dust loading of 8 g/m2 was selected based on the amount of house dust obtained from the 18 vacuum bags collected. 2 Table.S1 Surface dust loadings in U.S. homes Reference Adgate et al. 1995 Method Size fraction 216 houses sampled < 500 µm 0.05-7.0 (GM*: 0.42) Floor wipe samples Carpet vacuum samples Wang et al. 1995 Surface dust loading (g/m2) 0.3-99 73 houses sampled (GM: 7.8) < 500 µm 1-136 (Median: 16, SD: 16) < 150 µm 0.32-14.4 (median: 1.30) < 150 µm 0.7-21.1 (median:2.9) < 150 µm Average: 2.6 Carpet vacuum samples Robert et al. 1999 9 mid-class houses and 2 small offices HVS3: carpet Robert et al. 2004 10 old carpets HVS3: surface dust Takaro et al. 2004 214 low-income houses sampled (before environmental intervention) HVS3: child’s room or primary play area *: geometric mean Table.S2 Particle size fractions of house dust Reference Que Hee et al. 1985 Method Size fraction Sampling tube + sampling pump 392-833 µm: 11 % 149-392 µm: 13.3 % 44- 149 µm: 58 % <44 µm: 18 % (fraction for <53 µm: 24% (<150 µm) 20% (<500 µm) Lewis et al. 1999 25 middle-class homes sampled 500 µm-2mm: 13.3% Vacuum samples (~80% carpeted) 150-500 µm: 25.8% 53-150 µm: 33.1% 53-106 µm: 22.7% <53 µm: 27.8% (fraction for <53 µm: 33% (<150 µm) 25% (<500 µm) 3 Molhave et al. 2000 7 Danish office buildings Vacuum samples >125 µm: 40% 50-125 µm: 41% <50 µm: 18.8% (fraction for <53 µm: 31% (<150 µm) Reference: Adgate, J., Weisel, C., Wang, Y., Rhoads, G.G. and Lioy, P.J. (1995) Lead in house dust: Relationships between exposure metrics. Environ. Res., 70, 134-147. Lewis, R.G., Fortune, C.R., Willis, R.D., Camann, D.E. and Antley, J.T. (1999) Distribution of pesticides and polycyclic aromatic hydrocarbons in house dust as a function of particle size, Environ. Health Persp., 107, 721-726. Molhave, L., Schneider, T., Kjaergaard, S.K., Larsen, L., Norn, S. and Jorgensen, O. (2000) House dust in seven Danish offices, Atmos. Environ., 34, 4767-4779. Que Hee, S.S., Peace, B., Clark, C.S., Boyle, J.R., Bornschein, R.L. and Hammond, P.B. (1985) Evolution of efficient methods to sample lead sources, such as house dust and hand dust, in the homes of children, Environ. Res., 38, 77-95. Roberts, J.W., Clifford, W.S., Glass, G. and Hummer, P.C. (1999) Reducing dust, lead, dust mites, bacteria, and fungi in carpets by vacuuming, Arch. Environ. Con. Tox., 36, 477-484. Roberts, J.W., Glass, G. and Mickelson, L. (2004) A pilot study of the measurement and control of deep dust, surface dust, and lead in 10 old carpets using the 3-spot test while vacuuming, Arch. Environ. Con. Tox., 48, 16-23. Seifert, B., Becker, K., Helm, D., Krause, C., Schulz, C. and Seiwert, M. (2000) The German Environmental Survey 1990/1992 (GerES II): reference concentrations of selected environmental pollutants in blood, urine, hair, house dust, drinking water and indoor air, J. Expo. Anal. Env. Epid., 10, 552-565. Takaro, T.K., Krieger, J.W. and Song, L. (2004) Effect of environmental interventions to reduce exposure to asthma triggers in homes of low-income children in Seattle, J. Expo. Anal. Env. Epid., 14, S133-S143. Wang, E., Rhoads, G.G., Wainman,T. and Lioy, P.J. (1995) Effects of environmental and carpet ariables on vacuum sampler collection efficiency. Appl. Occup. Environ. Hyg., 10, 111-119. 4 Photograph of the resuspension mechanism Figure S1. Resuspension mechanism inside plaxiglass chamber. Toe and heel plates controlled by electric actuators and pressure controlled by compressed air flow. Carpet type descriptions Table.S3 Carpet type descriptions Characteristic Carpet LD Construction Cut pile Face weight (kg/m2) 2.12 Pile height (m) 0.02 Density (kg/m3) 101 Face fiber BCF nylon 6.6 Backing SoftBac® BCF: bulked continuous filament 5 Carpet HD Cut pile 1.65 0.01 138 BCF nylon 6.6 SoftBac® Carpet Loop Level loop 1.36 0.009 151 BCF nylon 6.6 SoftBac® Volume and number size distributions of house dust Figure S2. Volume and number size distributions of house dust measured by Malven Mastersizer 2000. Photograph of the seeding system Figure S3. Photo of the seeding system 6 Summary of average resuspension fractions Table.S4 Average resuspension fraction (n=3 for each treatment) Resuspension fraction ra Flooring type RH (%) SDL1 (g m-2) Carpet HD 40 2 (2.54±0.87)×10-6 (1.30±0.14)×10-6 (5.50±0.72)×10-6 (4.59±0.88)×10-5 (1.56±0.38)×10-4 70 2 -6 (4.59±3.02)×10 -6 (2.64±1.57)×10 -5 (1.57±1.06)×10 -4 (1.12±0.40)×10 (3.72±0.33)×10-4 40 8 (1.12±0.88)×10-6 (7.08±2.92)×10-7 (4.93±3.59)×10-6 (3.85±1.97)×10-5 (1.16±0.35)×10-4 70 8 (1.18±0.28)×10-6 (6.73±1.46)×10-7 (4.29±1.73)×10-6 (6.18±2.80)×10-5 (2.60±1.17)×10-4 40 2 -6 (4.99±3.56)×10 -6 (3.22±2.76)×10 -5 (1.62±0.91)×10 -4 (1.47±0.82)×10 (4.64±3.01)×10-4 70 2 (5.30±3.17)×10-6 (3.39±1.75)×10-6 (2.27±0.77)×10-5 (2.43±1.17)×10-4 (6.63±3.45)×10-4 40 8 -6 (2.41±0.34)×10 -6 (1.53±0.33)×10 -5 (1.53±0.98)×10 -4 (1.27±0.62)×10 (3.73±1.16)×10-4 70 8 (1.62±0.50)×10-6 (1.10±0.10)×10-6 (9.45±4.02)×10-6 (1.20±0.40)×10-4 (3.68±1.33)×10-4 40 2 (3.35±0.66)×10-6 (2.28±0.78)×10-6 (9.41±2.57)×10-6 (5.25±2.56)×10-5 (1.60±0.62)×10-4 70 2 -6 (3.18±1.28)×10 -6 (2.03±1.06)×10 -6 (6.46±1.17)×10 -5 (3.78±1.14)×10 (1.12±0.45)×10-4 40 8 (1.11±0.07)×10-6 (7.42±2.00)×10-7 (4.69±0.69)×10-6 (3.07±0.83)×10-5 (8.93±3.25)×10-5 70 8 -6 (1.04±0.68)×10 -7 (8.02±3.51)×10 -6 (3.69±1.34)×10 -5 (2.40±0.98)×10 (6.94±1.67)×10-5 40 2 (4.44±0.58)×10-6 (1.91±0.50)×10-6 (6.95±2.77)×10-6 (3.23±0.68)×10-5 (6.26±1.51)×10-5 70 2 (2.61±0.86)×10-6 (1.63±0.56)×10-6 (4.12±2.84)×10-6 (2.24±1.05)×10-5 (3.03±0.44)×10-5 40 8 -6 (2.74±1.17)×10 -6 (2.24±0.69)×10 -5 (1.56±0.31)×10 -5 (8.29±1.55)×10 (1.74±0.41)×10-4 70 8 (1.02±0.44)×10-6 (7.77±2.34)×10-7 (4.68±2.59)×10-6 (2.48±1.44)×10-5 (5.93±3.15)×10-5 40 2 -6 (4.16±1.04)×10 -6 (1.50±0.43)×10 -6 (5.98±0.42)×10 -5 (2.28±0.70)×10 (4.27±0.99)×10-5 70 2 (3.76±1.68)×10-6 (2.41±0.37)×10-6 (7.37±0.91)×10-6 (2.67±1.20)×10-5 (6.13±4.20)×10-5 40 8 (2.20±0.67)×10-6 (1.39±0.20)×10-6 (1.11±0.39)×10-5 (6.29±1.41)×10-5 (1.27±0.33)×10-4 8 -7 -7 -6 -5 (3.25±3.93)×10-5 Carpet LD Carpet Loop Hardwood Vinyl 70 0.4-0.5 µm 0.5-1.0 µm (8.26±3.53)×10 1.0-3.0 µm (4.82±2.71)×10 3.0-5.0 µm (2.55±2.52)×10 5.0-10 µm (1.03±0.80)×10 Summary of ANOVA results Table.S5 Summary of ANOVA results. ANOVA p-value ≤ 0.1 Factor 0.4-0.5 Particle size (µm) 0.5-1.0 1.0-3.0 3.0-5.0 0.027 0.018 0.000 0.017 0.715 0.044 0.022 0.158 0.000 0.131 0.271 0.009 Flooring type RH Surface dust loading (SDL) Flooring type & RH Flooring type & SDL RH & SDL 7 0.000 0.010 0.017 0.009 0.041 0.002 0.000 0.048 0.150 0.000 0.080 0.002 5.0-10.0 0.000 0.116 0.635 0.001 0.030 0.040 Results of Wilcoxon rank sum test Variable: flooring type Table.S6 Two-sided Mann-Whitney rank sum test results for comparison of resuspension fractions among different flooring types (n=6, α=0.1) * LD Carpet vs. HD Carpet LD Carpet vs. Loop Carpet LD Carpet vs. Hardwood LD Carpet vs. Vinyl HD Carpet vs. Loop Carpet HD Carpet vs. Hardwood HD Carpet vs. Vinyl Loop Carpet vs. Hardwood Loop Carpet vs. Vinyl Hardwood vs. vinyl 0.4-0.5 µm 0.0783 0.5-1.0 µm 0.0400 1.0-3.0 µm 0.0051 3.0-5.0 µm 0.0017 5.0-10.0 µm 0.0051 0.1572 0.7508 0.7508 0.0990 0.5834 0.3408 0.0007 0.0120 0.0050 0.0001 0.0002 0.0001 0.0001 0.0000 0.0000 0.4885 0.2855 0.3408 0.7075 0.1260 0.4025 0.977 0.7075 0.6650 0.0226 0.0531 0.0141 0.0051 0.0014 0.0002 0.9770 0.4357 1.0000 0.5067 0.9770 0.6236 0.5834 0.4357 0.7075 0.7950 0.3408 0.2855 0.1260 0.0531 0.5444 *: H0: 1 = 2; H1: 1 ≠2, where is the population median. Table.S7 One-sided Mann-Whitney rank sum test results for comparison of resuspension fractions among different flooring types (n=6, α=0.1)*. Only the floorings associated with significantly different ra values were tested. LD Carpet > HD Carpet LD Carpet > Loop Carpet LD Carpet > Hardwood LD Carpet > Vinyl HD Carpet > Loop Carpet HD Carpet > Hardwood HD Carpet > Vinyl Loop Carpet > Hardwood Loop Carpet > Vinyl 0.4-0.5 µm 0.0391 0.5-1.0 µm 0.02 1.0-3.0 µm 0.0026 3.0-5.0 µm 0.0008 5.0-10.0 µm 0.0026 0.0786 NA NA 0.0499 NA NA 0.0004 0.006 0.0026 0.0000 0.0001 0.0000 0.0000 0.0000 0.0000 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 0.0113 0.0265 0.0071 NA NA 0.0026 0.0007 0.0001 0.0265 0.0630 *: H0: 1 = 2; H1: 1 >2, where is the population median. 8 Variable: relative humidity Table.S8 Two-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 40% and 70% relative humidity levels (n=6, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm Hardwood 0.0306 0.0306 0.0202 0.0453 0.0306 0.3785 1.0000 0.1282 0.2298 Vinyl 0.0927 0.6889 0.3785 1.0000 0.4712 0.5752 LD Carpet 0.5752 0.4712 0.2298 HD Carpet 0.0656 0.0082 0.6889 0.6889 0.2980 0.2980 0.2980 Loop Carpet *: H0: 1 = 2 ; H1: 1 ≠2, where is the population median. Table.S9 One-sided Mann-Whitney rank sum test results for comparison of resuspension fractions between 40% and 70% relative humidity levels (n=6, α=0.1) *. Only the floorings associated with significantly different ra values were tested. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm Hardwood 0.0153 0.0153 0.0101 0.0227 0.0153 0.1892 0.5000 0.1149 Vinyl 0.0641 0.0463 *: H0: 1 = 2; H1: 1 >2, where is the population median. Table.S10 One-sided Mann-Whitney rank sum test results for comparison of resuspension fractions between 40% and 70% relative humidity levels (n=6, α=0.1) *. Only the floorings associated with significantly different ra values were tested. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 0.2876 0.2356 0.1149 HD Carpet 0.0328 0.0041 *: H0: 1 = 2; H1: 1 <2, where is the population median. Variable: surface dust loading The main effect of surface dust loading on dust resuspension was evaluated for carpets. Because RH & SDL interaction was significant for hard floorings, the effect of SDL on hard floorings was evaluated on each RH level separately. 9 Carpets Table.S11 Two-sided Student t-test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings on carpets(n=6, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 0.1282 0.298 0.298 LD Carpet 0.0082 0.0202 0.1735 0.3785 HD Carpet 0.0082 0.0051 0.0927 Loop Carpet 0.0051 0.0131 0.0082 0.0927 0.0453 *: H0: 1 = 2; H1: 1 ≠2, where is the population median. Table.S12 One-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings on carpets(n=6, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 0.149 0.149 LD Carpet 0.0041 0.0101 0.0641 0.1892 HD Carpet 0.0041 0.0025 0.0463 0.0867 Loop Carpet 0.0025 0.0065 0.0041 0.0463 0.0227 *: H0: 1 = 2; H1: 1 >2, where is the population median. Hard floorings Under 40% RH level Table.S13 Two-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 40% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 0.1904 0.6625 Hardwood 0.0809 0.0809 0.0809 1.0000 Vinyl 0.0809 0.0809 0.0809 0.0809 *: H0: 1 = 2; H1: 1 ≠2, where is the population median Table.S14 One-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 40% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm NA NA NA NA Hardwood 0.0952 0.5000 NA NA NA Vinyl 0.0404 *: H0: 1 = 2; H1: 1 >2, where is the population median 10 Table.S15 One-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 70% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm NA 0.3313 Hardwood 0.0404 0.0404 0.0404 NA NA Vinyl 0.0404 0.0404 0.0404 *: H0: 1 = 2; H1: 1 <2, where is the population median Under 70% RH level Table.S16 Two-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 70% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 1.0000 1.0000 0.6625 Hardwood 0.0809 0.0809 0.1904 0.3827 Vinyl 0.0809 0.0809 0.0809 *: H0: 1 = 2; H1: 1 ≠2, where is the population median Table.S17 One-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 70% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm 0.5000 NA NA Hardwood 0.0404 0.0404 0.1904 Vinyl 0.0404 0.0404 0.0404 0.0952 *: H0: 1 = 2; H1: 1 >2, where is the population median Table.S18 One-sided Mann-Whitney rank sum test results for comparison of resuspension fraction between 2 g/m2 and 8 g/m2 surface dust loadings for hard floorings under 70% RH (n=3, α=0.1) *. Flooring type 0.4-0.5 µm 0.5-1.0 µm 1.0-3.0 µm 3.0-5.0 µm 5.0-10.0 µm NA NA NA 0.5000 0.3313 Hardwood NA NA NA NA NA Vinyl *: H0: 1 = 2; H1: 1 <2, where is the population median 11 Emission rate plot Figure S4. Size-resolved emission for different types of floorings with two levels of initial surface dust loading. The symbols represent the median values for all tested conditions (n=60, replicates=6). The upper and lower ends of the error bars represent the 75th and 25th percentiles, respectively. 12