A pilot study on particle emission from printer paper shredders

ABSTRACT

This project investigated airborne particle releases derived from paper shredding. The characteristics of these emissions have not been studied previously. The objective of this project was to characterize released particles in terms of particle size distribution, particle mass and number concentration, particle morphology and chemical composition, as well as the structure of paper fibers. A significant amount of paper particles were found in the close vicinity of an unenclosed paper shredder during shredding and the manual manipulation of shredded paper. The particle release from shredding using two types of printer paper, two types of shredders, and various shredding frequencies was measured inside the shredder basket. The particles released from paper shredding were found to be in the nanometer to micrometer size ranges, and the particles contained calcium and other metal elements. Manual stirring of shredded paper released the highest particle concentration as measured at the opening of the container. Shredding with 30-s intervals increased the duration of airborne particle release as compared to intermittent shredding. Multiple quantitative analyses were employed to evaluate released particles at various particle size ranges. Two real-time instruments which measured thoracic (<10 μm) and respirable (<4 μm) particle sizes showed comparable concentrations measured inside and outside of the shredder basket.

Copyright © 2017 American Association for Aerosol Research

EDITOR :

• Tiina Reponen

Introduction

Past research in the paper industry has evaluated exposures in paper mills, which includes various toxic vapors, particles, and volatile organic compounds (VOC) released from pulp and paper manufacturing. These emissions are reported to cause health issues for workers and the surrounding community (Toren et al. 1996; Szadkowska-Stanczyk and Szymczak 2001; Soskolne and Sieswerda 2010; Andersson et al. 2010). Health problems such as occupational asthma and allergic reactions to paper dust have been reported by the American Postal Workers Union (APWU). The U.S. Postal Service (USPS) studied the issue in 1998 (American Postal Workers 2009) and discovered that sorting machines could send potentially carcinogenic VOCs (i.e., ink) and other irritants such as dust mites into the air (American Postal Workers 2009). A population-based cross-sectional study concluded that exposure to paper dust and to fumes from photocopiers and printers is related to an increased risk of sick building syndrome and that carbonless copy paper exposure increases the risk of eye symptoms (Jaakkola and Jaakkola 2007; Jaakkola et al. 2007). In particular, paper dust exposure is related to an increased risk of respiratory symptoms (Toren et al. 1994) and adult-onset asthma (Jaakkola and Jaakkola 2007; Jaakkola et al. 2007). The Occupational Safety and Health Administration (OSHA) states that paper fiber (i.e., cellulose) exposure associated with tissue paper, cellulose insulation, and reed dust could cause respiratory symptoms; however, the studied exposures didn't include printer paper, and paper particles were not further characterized for their particle size or elemental composition (Occupational and Health 2005). Any such dust/particle exposure is presently considered to be a nuisance hazard (Occupational and Health 2005). In addition, paper dust emissions have not been investigated when shredding paper for disposal. Shredding may produce fine and ultrafine particles which can deposit in lung alveoli and penetrate into the interstitium, possibly evoking greater inflammation, cardiopulmonary health effects, and overall toxicity as compared to larger respirable particles (Wolff et al. 1988: Ferin et al. 1990, 1991; Zhang et al. 2000). Paper shredding tasks generate fine particles which could cause exposure to office workers and the general population using shredders at home. Shredding particle exposure typically occurs in the indoor environment of residences or administrative offices where air circulation is limited when compared to a production facility. For example, the printing room and classroom have same required air exhaust rate, which is a third of the exhaust rate required for an auto repair room (7.5 L/s-m²) (ASHRAE 2003). People who have prolonged exposure to paper shredding dust could develop respiratory tract irritation. However, such paper aerosols have not been characterized as to their elemental composition, particle size distribution, morphology, or particle number concentration.

In this study, we used an advanced exposure assessment methodology, which includes the use of direct-reading instruments to measure airborne particle number concentration and size distribution, gravimetric sampling, and the Tsai diffusion sampler (Tsai et al. 2009) to collect particles for elemental and morphological analysis using electron microscopy. Measurements were (1) taken in the close vicinity of the shredder during both shredding and manipulating shredded paper, and (2) taken inside the shredders in order to characterize the airborne particle release at the source. Since it is important to understand the source strength, measurements at both outside and inside the shredder were measured. The particle release level for nanometer to micrometer-sized particles associated with various types of papers and shredders was analyzed.

Methods

Equipment

Aerosol particle assessments were performed using the following instruments and techniques: (1) real-time instruments (RTI) including the nanoscan scanning mobility particle sizer (SMPS) (Model 3910, TSI, Shoreview, MN, USA) which measures particle sizes of 10–420 nm and the optical particle sizer (OPS) (Model 3330, TSI) which measures particle sizes of 0.3–10 µm; using both together to analyze airborne particle concentration from 10 nm to 10 µm with a 1-min response time; (2) a Tsai diffusion sampler (TDS) with attached transmission electron microscopy (TEM) grid on the filter to collect airborne particles at a flow rate of 0.3 L/min; TEM-copper grids (400 mesh with SiO₂ film coating, SPI, West Chester, PA, USA) and 25-mm-diameter polycarbonate membrane filters (0.2-μm pore size, Millipore, Billerica, MA, USA) were used together to deposit particles for analysis; (3) and a gravimetric sampling method (NIOSH 0500, closed face cassette, 2 L/min flow rate for total dust) to collect particles for mass concentration analysis.

Paper shredding and sampling

Two types of printer papers: (1) coated paper (Staples color laser paper, gloss finish, heavy duty, item # 633215) and (2) uncoated paper (Staples white printer paper) were shredded using a cross cut shredder (GearHead model PS80000MXW, shredding speed 196 cm/min, basket volume of 17 L) and a straight cut shredder (Techko, SH2106PA, shredding speed 366 cm/min, basket volume of 14 L) to collect particle release data and aerosol samples for analysis.

Paper particles released from the straight cut shredder were measured inside and outside the shredder. The straight cut shredder had an unenclosed basket (opening at top), with measurements outside the shredder taken approximately 15 cm vertically above the shredder center line as shown in Figure 1a. The cross cut shredder had an enclosed basket and was studied for particle release inside the shredder only. For measurements taken inside the shredder, the feeding slot was the only opening for air to flow into the shredder as shown in Figure 1b. The direct reading instruments and sampling pumps withdrew air from the shredder through tubing (Tygon) and were placed in the shredder basket at approximately 10 cm vertically below the shredder as shown in Figure 1b. The sampling tubing was water washed and vacuum dried after each use. Particle loss inside the tubing can be ignored based on results of a previous study (Tsai 2015). The air flow rate for measurements taken during shredding was driven by the air suction from two instruments (SMPS 1 L/min, OPS 1 L/min) and one sampling pump (TDS 0.3 L/min) with a total air flow rate of 2.3 L/min. The airflow rate, airflow direction, dimension of paper feeding slot, and shredding frequency were the same for all experiments.

Figure 1. Illustration of experimental set up. (a) Measurements outside shredder basket and (b) measurements inside shredder basket.

For measurements outside the shredder basket, the straight cut shredder, basket, sampler, and tubing were placed in a glovebox (Series 100, Terra Universal, Fullerton, CA, USA) equipped with ultrafiltered clean air with RTIs placed outside the glovebox. The shredder surfaces and blades were wiped clean with isopropyl alcohol and deionized water and all surfaces in the glovebox were cleaned before each experiment. Uncoated printer paper was shredded and particle releases from paper shredding (20 min) and manual manipulation (i.e., stirring [10 min] and transferring [5 min]) of shredded paper pieces were measured. Shredding was run with 40 sheets of paper and one shredding frequency (single sheet every 30 s). Each experiment was monitored for up to 100 min starting from background concentration measurement through activities, breaks, and final venting. Control data including the background concentration inside the glovebox before the experiment and the particle concentration while running the shredder without any paper were measured for comparison. The experiment was repeated three times with measurements taken each time for analysis. Airborne paper particles were collected on TEM grids using the TDS. The airflow rate was 2.3 L/min total (1 L/min for each RTI and 0.3 L/min for TDS). The ultrafiltered air circulation was turned off during the experiment.

For measurements inside the shredder basket, openings or gaps on the basket were sealed. Four sets of experimental shredding scenarios (two paper types, two shredders) were conducted for paper particle release assessment. The SMPS and OPS measured airborne particle concentrations inside the shredder basket for 35 min during each experiment; concentration levels were measured before, during, and postshredding. The two types of paper were manually cut from the standard 8.5” × 11” size to a dimension of 6.5” × 11”, for smoother feeding into the 8.5” wide slot. A total of 40 sheets were used in each batch of shredding and were shredded using four different shredding frequencies. This included (1) five sheets at a time fed continuously (<2 min shredding), (2) five sheets at a time fed every 30 s (∼4 – 5-min shredding), (3) one sheet at a time fed continuously (∼5 min shredding), and (4) one sheet at a time fed every 30 s (∼20-min shredding). The gravimetric particle samples were collected for 30 min from the start of each shredding run. The shredder was placed on a mobile workstation, with connected instruments located adjacent to the shredder in order to minimize sample line length. At the end of each experiment, the shredder was moved into a high efficiency particulate air (HEPA) filtered hood enclosure (Model: Nanohood, Labconco, Kansas City, MO, USA) to remove the shredded paper from the basket by placing it in a plastic bag and sealing for disposal. The shredder basket was cleaned using wet paper towels to wipe the inner surface after each experiment. Researchers were protected with gloves, laboratory coats, and N100 disposable respirators for the shredded paper removal and cleaning.

Characterization and analysis

Paper particles collected on copper grids and polycarbonate filters were analyzed using TEM, scanning electron microscopy (SEM), and energy dispersive spectroscopy (EDS) to characterize particle morphology, size, and elemental composition. TEM grids collected airborne particles on a silicon-coated film on the grid space, and silicon was excluded in some EDS analyses. TEM analysis was performed using a Philips TEM (Model CM12, FEI, Hillsboro, OR, USA) at 100 kV equipped with a Kevex EDS detector with an IXRF digital imaging system and digital camera, and a JEOL TEM (Model 2100F, JOEL, Peabody, MA, USA) at 200 kV equipped with a digital Gatan Ultrascan camera. TEM grids were also analyzed using SEM (Model 6500F, JEOL) at 15 kV equipped with a super octane silicon drift detector EDS system. Polycarbonate filters were analyzed using SEM; a portion of the filter was cut out, placed onto a carbon-taped stub, then gold coated. The sample was then analyzed by a Tescan SEM (Tescan, Kohoutovice, Česká Republika) equipped with a Gresham light element detector and an IXRF EDS, and a FEI NOVA SEM (FEI) equipped with an Oxford INCA Energy 250 system electron dispersive X-ray analysis (EDX).

RTI data were analyzed for particle concentration and diameter in two stages, i.e., data calculation from raw data sheets, and an advanced statistical analysis to look for data correlation and significance using the statistical package for the social sciences (SPSS) program. The airborne particle concentrations were analyzed and correlated with shredder use, shredding scenario (paper and shredder types), shredding frequency (continuous or periodic shredding), and paper type.

Results and discussion

Paper particle concentrations in close vicinity of the shredder

The area concentrations at approximately 15 cm above the straight cut shredder were monitored for the entire paper shredding and manipulation periods. The same operation was run for three repetitions. Activities including paper shredding, stirring, and transferring shredded paper were monitored with breaks in between each activity. Total particle number concentrations measured by the OPS and SMPS are shown in Figure 2. Background concentrations before activities in the filtered glovebox were close to zero for both the OPS and SMPS measurements. Paper shredding (20 min) and stirring shredded paper (10 min) generated significant amounts of particles, with stirring shredded paper releasing more particles than shredding paper. Stirring shredded paper caused airborne particles to reach the highest concentration and as a process it contributed over 4000 particles/cm³ to the total concentration of 5000 particles/cm³ at 0.3–10 μm size range (Figure 2a), and it exceeded 2.5 × 10⁵ particles/cm³ at the 10 – 420-nm-size range (Figure 2b). The peak concentration during stirring shredded paper was found at the beginning of stirring by OPS measurement and at the end of stirring by SMPS measurement. Particle concentrations were reduced at the first break (4 min) between paper shredding and stirring. However, the concentrations at the second break (20 min) between stirring and transferring shredded paper showed a reduction on 10 – 420-nm particle size but an increase on 0.3 – 10-μm particle size, and this difference was found to be the same from all three repetitions. For transferring shredded paper (4-min activity), shredded paper pieces were gently dumped into a trash bag from the basket, and an elevated concentration was found for particles in the 10 – 420-nm particle size range measured by the SMPS, but a reduced concentration was found by OPS measurement. Concentrations were reduced at the last break (20 min) after transferring shredded paper into a trash bag. The filter ventilation of the glove box was turned on for the last 6 min of monitoring and the concentrations were quickly reduced to close to zero particle count. Manually stirring shredded paper by hand was found to generate the highest particle concentration.

Figure 2. Total particle number concentration at area nearby straight-cut shredder during paper shredding and manipulating shredded paper pieces. (a) OPS (0.3–10 μm) concentration and (b) SMPS (10–420 nm) concentration.

The straight cut shredder was run without feeding paper, and the particle concentrations were measured for 25 min of each run for all three repetitions. The average particle concentration and size distribution from the shredder running without paper is shown in Figures 3a and b (diamond [red] line) for OPS and SMPS measurements, respectively. The average concentration and size distribution of other activities were also shown in the same figure for comparison. The shredder run without paper generated few particles at the size of 0.3–10 μm (Figure 3a) but it generated a significant amount of particles at a size less than 200 nm, with mode size around 50 nm and the peak concentration above 2.5 × 10⁵ particles/cm³ (line with circles [red line] of Figure 3b). These particles were shown with a single-mode distribution, which was different to particles released from other activities involving paper shredding and manipulation, which displayed multiple mode distributions (Figure 3b). According to SMPS measurement, paper shredding released small particles less than 100 nm. Stirring and transferring shredded paper released particles both larger and smaller than 100 nm. Activities involving paper were shown to significantly generate particles larger than 300 nm and the concentrations of micrometer-sized particles (>1 μm) were found to be orders of magnitude higher than particles in the range of 300 nm to 1 μm. Overall, the size distributions were similar for all activities involving paper for both nanometer- and micrometer-size ranges. A significant amount of paper particles, both small and large, were found in the area above the straight cut shredder during shredding and manual manipulation of shredded paper.

Figure 3. Paper particle size distribution and concentration at area nearby straight-cut shredder (a) 0.3–10 μm by OPS and (b) 10–420 nm by SMPS. Normalized particle number concentration is the concentration of dN/dlog(Dp). Each line presents an average of particle number concentrations during the time period (4–20 min) of each activity named in figure legend.

Paper particle concentrations at the shredding source

Paper particles released from paper shredding activities were studied further by measuring the concentration inside the basket (source location) to differentiate possible factors (e.g., different shredding frequencies and paper types) affecting the released particle concentration. Number concentrations were measured.

The average total particle number concentrations and standard deviations are shown in two size ranges, 10 – 420 nm by Nanoscan SMPS and 0.3–10 µm by OPS in Table 1. The number concentrations measured in the 0.3–10-µm-size range showed consistent trends for all four shredding scenarios. The cross-cut regular paper had the highest average number concentration of 1492 particles/cm³ (including background particles), and straight-cut coated paper had the lowest average number concentration of 92 particles/cm³. However, the particle number concentrations in the 10–420-nm-size range showed a different trend of concentration levels for the four shredding scenarios, which were not correlated with the OPS (0.3–10 μm) measurement. The highest number concentration of particles in the 10–420-nm-size range was 109,111 particles/cm³ measured using the straight-cut shredder with regular paper, and the lowest concentration was 35,875 particles/cm³ measured using the cross-cut shredder with regular paper. Statistical significance (t-test with significance at 0.05 level) was analyzed on average concentrations in Table 1 to compare the difference between paper types (SI Table S1) and shredding frequencies (SI Table S2). t-test results and p values are presented in Tables S1 and S2. Regular paper released statistically significant higher particle concentrations than coated paper at a size ranging 0.3–10 μm by OPS measurement (p < 0.05) regardless of shredding frequencies (Table S1). However, particle concentrations at sizes ranging 10–420 nm by SMPS measurement were significantly higher on shredding regular paper only for three out of eight measurements (Table S1). Two experiments showed significantly higher concentration from shredding coated paper but the other three experiments do not have significant differences between regular and coated paper shredding. There was no consistent relationship between measurements of the two size range RTIs.

Table 1. Average total particle concentrations for each shredding scenario and frequency.

	CC+R		CC+C		SC+R		SC+C
Shredding Scenario	Avg.	SD	Avg.	SD	Avg.	SD	Avg.	SD
10–420 nm, nanoscan SMPS	35,900^*	n/a	41,600^*	n/a	109,000^*	n/a	62,700^*	n/a
5 sheet continuous (<2 min)^#	32,200	44,475	63,200	82,100	48,443	112,061	47,800	103,048
5 sheet 30 s (4–5 min)	39,000	47,305	63,000	74,087	27,200	46,198	20,800	32,224
1 sheet continuous (5 min)	23,500	29,863	11,300	12,319	95,800	263,130	47,000	129,209
1 sheet 30 s (20 min)	48,800	29,850	29,000	19,875	265,000	224,992	135,000	109,608
0.3–10 µm, OPS	1,492^*	n/a	193^*	n/a	279^*	n/a	92^*	n/a
5 sheet continuous (<2 min)	883	1,315	189	421	188	306	75	139
5 sheet 30 s (4–5 min)	894	1,590	196	296	220	315	27	57
1 sheet continuous (5 min)	1,480	2,105	125	175	373	680	66	210
1 sheet 30 s (20 min)	2,710	1,606	263	117	334	247	202	276

CC: Cross-cut, SC: Straight-cut, R: Regular paper, C: Coated paper, Avg.: average number concentration, SD: standard deviation, and n/a: not available.

* Overall average of particle number concentration (particles/cm³) from four shredding frequencies shown in the following rows. Data of each shredding scenario are the average of a 30-min particle concentration starting the beginning of shredding paper until after shredding.

^# Time period of shredding paper is noted in the parenthesis.

The number concentration of each shredding frequency in Table 1 showed that shredding one sheet every 30 s generated the highest number of particles on three (CC+R, SC+R, SC+C) out of four shredding scenarios for the 10–420-nm size range, and on three (CC+R, CC+S, SC+C) out of four for the 0.3–10-µm size range. The concentrations of different shredding frequencies were analyzed with a t-test for significance on shredding regular paper using the straight-cut shredder. Results (Table S2) showed that the high concentration (265,000 #/cm³) of one sheet 30-s shredding was not significantly different to some other shredding frequencies due to the high variation of concentrations. The correlation among concentration profiles was further analyzed.

Further analysis of particle number concentrations are presented in two types of graphs. First is the total number concentration change over the entire measurement period, and second is the particle size and concentration distribution. Total number concentration profiles through the overall measuring period of 35 min for each shredding frequency are shown in Figure 4 for cross-cut shredding and online supplementary information (SI) Figure S1 for straight-cut shredding. The differences in total number concentrations among various shredding frequencies are also shown in these figures. Variations in the number of papers shredded at a time as well as the total shredding time resulted in different concentration profiles: a shorter shredding period, such as continuous shredding and five sheets per shred, had a higher concentration accumulating within the first 5 min of shredding and decreased quickly after all 40 sheets had been shredded.

Figure 4. Total particle number concentration of cross-cut shredding measured by OPS (left y-axis) and SMPS (right y-axis), (a) regular paper and (b) coated paper. The concentration at each time/min presents an average of 13 and 16 size-fractional concentrations for SMPS and OPS, respectively.

The shredding frequency of a single sheet every 30 s took the longest period of time to complete shredding of all 40 sheets. The concentration profiles are discussed below. For cross-cut shredding regular paper presented in Figure 4a, the highest average concentration was for a single sheet fed every 30 s. A concentration of about 4 × 10³ particles/cm³ was found for larger (0.3–10 µm) particles and concentrations of 5–9 × 10⁴ particles/cm³ for smaller (10–420 nm) particles were found throughout the entire shredding period. However, fluctuating peak concentrations were observed in the other three shredding scenarios (i.e., cross-cut coated, straight-cut regular, and straight-cut coated as shown in red solid and dashed lines in Figures 4b and S1a and b). These high peak concentrations could be related to the dynamic airflow inside the shredder basket and shredder vibration during shredding operations. Regardless of the fluctuation, this shredding frequency—single sheet every 30 s generated particles continuously for the longest period of time among all other frequencies.

Further statistical analyses were performed to compare the correlation of concentration profile patterns among the four shredding frequencies and the four shredding scenarios for both concentrations in the10–420-nm and 0.3–10-µm-size ranges. These correlational analyses presented the variation or similarity among the shredding scenarios in terms of the particle release pattern. Tables listing all analysis results of total concentration including Pearson correlation and significance are available in Tables S3 and S4 (SMPS and OPS). For cross-cut regular paper shredding and concentrations in the 10–420 nm (SMPS) range, as seen in Figure 4a, the 5-sheet continuous and 5 sheet every 30 s shredding frequencies showed the strongest correlations, with Pearson correlation coefficients of 0.964 with significance at 0.01 level (Table S3). Shredding using single-sheet continuous, 5-sheet continuous, and 5 sheets every 30 s were strongly correlated (Pearson 0.942, significance 0.01) with each other as well (Table S3). The shredding frequency of the single sheet every 30-s condition showed a low correlation of 0.606–0.73 (0.01 significance) to other conditions. Shredding a single sheet every 30 s is not correlated with other shredding frequencies, which indicates that the longer shredding and particle release period was unique as compared to other shredding frequencies. A single sheet every 30 s correlated negatively (-̶0.077 to -̶0.11, regular) and poorly (0.487–0.817, coated) among all shredding frequencies. However, the single-sheet shredding every 30-s condition did correlate significantly with the other shredding frequencies among the cross-cut coated paper scenario (Table S3).

The statistical analysis results for the total concentrations of 0.3–10 µm (OPS) particles are shown in SI Table S4. The shredding frequency of a single sheet every 30 s was consistently found to have a low correlation with other frequencies throughout all four shredding scenarios (CC+R, CC+C, SC+R, SC+C) for OPS measurements. To summarize, shredding frequencies showed various concentration patterns, the correlation analysis showed some relationship and the significance analysis of mean values showed that the highest average concentration of single-sheet 30 s was not significantly higher than all others due to high variation of data. However, there was significant difference between the two types of paper.

Particle number and size analysis

The RTI data were further analyzed for individual particle size and the corresponding concentration. The size-differentiated concentration graphs and the size distributions are presented in Figure 5 for 10–420 nm (SMPS) measurements and SI Figure S2 for 0.3–10 µm (OPS) measurements. Concentrations measured during shredding and before shredding (baseline) are included in the graphs. The baseline concentration before each shredding run was recorded and all baseline concentration levels were very low; thus the concentration data during shredding were not adjusted by subtracting the baseline concentration.

The particle mode diameters and corresponding concentrations from the size distribution analysis are summarized in Table 2. For SMPS measurements, the particle mode size was always between 27 and 48 nm, with the majority equal to 36 nm for all shredding scenarios and frequencies (Table 2). Coated paper was shown to release smaller particles with a mode of 27 nm for some measurements. The size distributions were primarily in the 10–100 nm range (Figure 5a) but a few of the shredding frequencies extended from 10–200 nm (Figures 5b–d). For OPS measurements, the particle mode diameter varied by the shredding scenario or frequency with a wide range from 337 to 1994 nm, and multiple peaks are typically seen on all data. The highest peak SMPS concentration was for shredding a single sheet every 30 s for three scenarios, i.e., straight-cut on regular and coated paper and cross-cut on regular paper(but not cross-cut on coated paper Figure 5). This highest concentration was seen on all OPS measurements (Figure S2). The highest mode diameter concentration from SMPS measurements was 4.5 × 10⁵ particles/cm³ and 668 particles/cm³ from OPS measurements using straight-cut shredding on regular paper with a single sheet every 30 s (Table 2). The single sheet shredding every 30 s condition showed high peaks on most measurements, however, the statistical analysis didn't show a significant difference compared with other data.

Table 2. Mode particle diameter and particle concentration.

	CC+R			CC+C			SC+R			SC+C
SMPS 10–420 nm	D	Conc.	SD	D	Conc.	SD	D	Conc.	SD	D	Conc.	SD
5 sheet continuous	36.5	54,600	72,223	36.5	110,000	141,476	36.5	93,200	219,171	36.5	72,000	162,790
5 sheet 30 s	36.5	63,100	67,234	36.5	108,000	126,979	36.5	40,800	69,014	36.5	33,000	56,382
1 sheet continuous	36.5	33,100	34,457	27.4	17,800	17,790	27.4	158,000	463,605	27.4	96,000	258,842
1 sheet 30 s	36.5	68,400	74,516	27.4	45,800	30,131	48.7	452,000	435,485	36.5	212,000	187,236
OPS 0.3–10 µm^*	D	Conc.	SD	D	Conc.	SD	D	Conc.	SD	D	Conc.	SD
5 sheet continuous	1,994	1,120	1,770	337	306	468	337	435	585	337	274	342
5 sheet 30 s	1,994	940	1,852	337	356	395	419	459	756	1,007	51	112
1 sheet continuous	1,994	1,550	2,532	522	195	250	337	577	807	1,007	108	354
1 sheet 30 s	1,994	2,940	1,859	522	450	347	337	668	359	1,007	292	487

* Multiple peaks were shown on concentration, the particle size of the highest peak concentration was noted as the mode particle size.

D: diameter in nm, data are from particle sizer; Conc.: average number concentration at the mode diameter (particles/cm³); SD: standard deviation.

Figure 5. Paper particle size distribution and concentration at 10–420 nm measured by SMPS, (a) cross-cut regular paper, (b) cross-cut coated paper, (c) straight-cut regular paper, and (d) straight-cut coated paper. Normalized particle number concentration is the concentration of dN/dlog (Dp). Each line presents an average of particle number concentrations during the whole shredding experiment (30 mins).

Microscopic analysis

Paper particles collected outside the shredder were analyzed on TEM grids and images are shown in Figure 6. Particles were successfully collected on filmed grids using TDS, and the particle distribution (Figures 6a and b) showed various sizes of particles, with few larger than 10 μm (Figure 6a). Fibrous shaped particles were found agglomerated with irregular shaped particles as marked with arrows. The EDX analysis is shown in Figures 4a and d. Calcium (Ca) and Carbon (C) were the main elements found in paper particles.

Figure 6. TEM images of collected paper particles on grid: (a) and (b) overview of particles on two grid spaces; (c) and (d) various shapes of paper particles. Fibrous particles are marked with arrows (red).

Released paper particles collected on polycarbonate filters using the TDS were analyzed by SEM and particle diameters were found in a wide size range of nanometers to micrometers. Figure 7 presents images of typical airborne particles released from shredding regular and coated paper (Figures 7a–d), and coated paper pieces (Figures 7e and f). Various sizes of paper particles in irregular shapes mixed with fibrous and rounded agglomerates were seen on the filters (Figures 7a–d). The circular pores (dark pores) on the polycarbonate filter have an average diameter of 200 nm. Collected particles found deposited on filters between pores were less than 100 nm with many less than 50 nm in diameter, and are marked with arrows (Figures 7b and d). Large particle agglomerates greater than 5 µm were seen on the filter as well (Figures 7a and c).

Figure 7. SEM images: (a) and (b) airborne particles released from regular paper; (c) and (d) airborne particles released from coated paper; (e) side view of shredded coated paper piece; (f) top view of shredded coated paper piece. Representative particles are marked with arrows.

The shredded small paper pieces were analyzed using SEM for comparison to the released airborne particles. The images of coated paper pieces showed fibrous structures tangled with other fibers with a diameter of micrometer size as shown in Figure 7e. This image was taken in a side view showing the cut side of one shredded paper piece. In addition, a top view image of this coated paper piece was taken (Figure 7f) and the coating treatment was seen on the paper surface. This coated paper surface exhibited a round and layered structure with individual coatings in the submicron size. Micrometer-sized fibrous structures would be released as well but were not collected in this study. A filter sampling with a higher flow rate will be used together with the TDS for future study of larger particles. The EDX result for the coated paper piece is shown in Figures S3a and b. The spot scan showed elements including C, platinum (Pt), silica (Si), aluminum (Al), and Ca (Figure S3b). The images of regular paper pieces are available in Figures S3c and e, which show fibrous shapes, and EDX results are available in Figures S3d and f. The spot scan on fibers and flat structures of the paper (Figures S3c and e) showed elements including C, palladium (Pd), and Ca. Released airborne particles were found to have similar structures to these.

SEM-analyzed airborne paper particles were also analyzed for elemental composition and the results are shown in Figure 8. The EDX analysis targeted a single-paper particle (Figures 8a and c) and showed that the elements of C, Pt, Si, Al, magnesium (Mg), and Ca were found in coated paper particles as shown in Figures 8b and d. Elements including C, Si, Pt, and Ca were found in regular paper shredding particles (Figures 8e and f). In addition, TEM imaging and EDX analysis of the same airborne particles released from shredding regular paper are shown in Figures S4a and b. The elements detected by TEM associated with this released particle are C and Ca, which was confirmed to be associated with paper. The other elements detected on particles would be related to the shredder use. These studied particles were released from new printer paper, more may be released from aged or differently textured paper. Particles in nanometer to micrometer size were successfully collected using this TDS, and represent what would be inhaled by a person in that environment.

Figure 8. SEM images and EDX of airborne paper particles: (a) coated paper by cross cut; (b) EDX of (a), Pt, Si, Al, Mg, Ca, C; (c) coated paper by straight cut; (d) EDX of (c), Si, Al, Mg, Pt, Ca, C; (e) regular paper by cross cut; (f) EDX of (e), Si, Pt, Ca, C.

Conclusions

Several conclusions can be drawn from the data collected in this study. A significant amount of paper particles were released in the close vicinity of the shredder during shredding and manual manipulation of shredded paper using an unenclosed shredder and basket. Manual stirring of shredded paper generated the highest concentration of particles as compared to shredding paper and transferring shredded paper. Such activity should be avoided. Many particles, from nanometer to micrometer sizes, were measured in this study. Particle agglomerates were mixed with other round-like, irregular and fiber shaped particles. The regular uncoated paper released a higher concentration of large micrometer particles into the air for all shredding experiments. The amount of small particles (<420 nm) released couldn't be concluded to be significantly different between paper types. The particles released from coated paper were smaller than those released from uncoated paper. The difference among shredding frequencies was not significant. Various metal substances were found in released paper particles and C, Pt, Si, and Ca were consistently found in both regular and coated paper shredding. However, we can conclude that Ca and C were confirmed to be associated with paper materials, with other elements likely being released from the shredder. Those particles released from shredding unprinted printer paper would be of concern to indoor air quality. Future research is needed to investigate if printed paper includes additional substances that could be aerosolized during shredding.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgments

The authors thank Christopher J. Gilpin and Roy Geiss for technical support on TEM and EDX analysis, and Kaitline Soukup for assistance on laboratory activities.

Funding

This research study was supported by the National Institute for Occupational Safety and Health Pilot Research Project Training Program of the University of Cincinnati education and Research Center Grant #T42/OH008432-09.