Laser Printers: A Source of Nanoparticles in Indoor Office Environments

Transcription

1 Laser Printers: A Source of Nanoparticles in Indoor Office Environments Lidia Morawska 1, Congrong He 1, Graham Johnson 1, Rohan Jayaratne 1, Tunga Salthammer 2, Hao Wang 1, Erik Uhde 2, Thor Bostrom 1, Robert Modini 1, Godwin Ayoko 1, Peter McGarry 1,3, Michael Wensing 2 1 International Laboratory for Air Quality and Health, Institute of Health and Biomedical Innovation Queensland University of Technology. PO Box Brisbane. Queensland Australia, Tel: l.morawska@qut.edu.au 2 Fraunhofer Wilhelm-Klauditz-Institute (WKI), Material Analysis & Indoor Chemistry, Bienroder Weg 54 E, 38108, Braunschweig, Germany 3 Workplace Health and Safety Queensland. ABSTRACT It is now well known that the operation of some laser printers results in the emission of high concentrations of ultrafine particles. A significant fraction of these particles lie in the nanometre size range and the health effects of these particles are still largely unknown. In this respect, understanding the process of particle formation is important considering the widespread use of laser printers and human exposure to these particles. In this study, using state-of-art instrumental methods, we addressed three basic questions regarding the emissions: (1) what is the composition of the particles? (2) what is their formation mechanism? and (3) what factors determine the rate of emission from a printer? The results indicate that the emission rates of ultrafine particles from printers are directly related to the fuser temperature. We also show that the particles are volatile and are of secondary nature, being formed in the air from volatile organic compounds originating from both the paper and hot toner. It is evident that some of the toner is initially deposited on the fuser roller, after which the organic compounds evaporate and then form particles, through one of two main reaction pathways: homogenous nucleation or secondary particle formation involving ozone. INTRODUCTION The emission of ultrafine particles (<0.1 μm) during the operation of laser printers is now well-established (He et al, 2007; Kagi et al, 2007, Schripp et al, 2008). This is of significance to indoor air quality as the toxicological effects of ultrafine particles have recently been acknowledged by the World Health Organisation (WHO, 2005). Laser printers are also known to emit larger particles, volatile organic compounds (VOC) and ozone (Tuomi et al, 2000; Lee and Hsu, 2007; Lee et al, 2001). However, until now, we have had very little knowledge of the composition of the particles and their formation mechanisms and relationship to other emitted pollutants (Destaillats et al, 2008). Despite the apparent 1

2 simplicity of testing the emissions from these relatively small devices, the persistent challenge is in interpreting the results, which stem from several potential contributors to the emissions such as paper, toner powder and fuser roller material, as well as a number of possible physicochemical pathways for particle formation. This paper describes the results of an extensive set of measurements aimed towards identifying the source and nature of the emitted particles and the factors that determine the rate of emission of ultrafine particles from laser printers. METHODS 2.1 Printers and Media Study 1 comprised a detailed investigation into the emission properties and was conducted using a relatively high particle number emitting printer that we shall denote Printer A. (see He et al, 2007 for a list of printers classified according to their particle number emission rates). It was operated with 0%, 5%, or 50% toner coverage within a flow tunnel and in a closed box chamber. For details of the experimental systems, see Morawska et al (2009). The tunnel study provided information on characteristics of particles, ozone and VOC approximately 30s after emission. The chamber study provided information on particle ageing and their volatility and hygroscopic properties that required a longer time than that available in the tunnel. In addition, emissions from the idle belts of the fuser rollers, paper and toner powder were investigated in a controlled-temperature furnace to provide insight into particle formation from each of these materials independently from the other components of the printer. The study involved over 100 emission tests. Study 2 included over 30 laser printers belonging to five popular brands. Each printer was placed in the box chamber and tested for particle number and VOC emissions. 2.2 Instrumentation, sampling and analytical methods Total particle number concentration in the submicrometer size range was measured by two TSI Condensation Particle Counters (CPCs), Model 3022A and Model 3781, with a sampling time of 1 s and a size range of µm. Particle size distribution in the submicrometer range was measured by a Scanning Mobility Particle Sizer (SMPS) incorporating either a TSI Model 3080 or 3071 Electrostatic Classifier (EC) with a CPC Model 3025A, with a sampling time of 120 s. The particle size range of the two SMPS systems were nm and nm respectively. Particle size distribution in the supermicrometer range was measured by a TSI Model 3320 and a TSI Model 3312A Aerodynamic Particle Sizer (APS), with a sampling time of 10 s and a size range of µm. 2

3 The thermal decomposition and hygroscopic behaviour of particles in the dominant size range was measured by a Volatilisation and Humidification Tandem Differential Mobility Analyser (VH-TDMA) (Johnson et al, 2004). This instrument comprises two TSI Model 3010 CPCs and three TSI Model 3071A ECs). The VH-TDMA selects particles of a very well defined size from within the instrument range ( μm), and subjects them to thermal treatment at successively higher temperatures, while measuring the volume and water activity (via hygroscopic growth) of the thermally modified particles. Ozone and Total Volatile Organic Compounds (TVOC) concentrations were monitored using an UV-106 Ozone Analyzer (Ozone Solutions, Inc. IA, USA), with a sampling time of 10 s and a PPB Rae Plus Photoionisation Detector (PID), respectively. Samples were collected using Tenax tubes and analysed for volatile and semivolatile organic compounds (VOC and SVOC). Particles were collected on Transmission Electron Microscope (TEM) grids placed normal to the flow in the tunnel and analysed for individual particle chemical composition and morphology. The average fuser heater surface temperature during printing was monitored using a 1s resolution thermocouple placed between the fuser pressure roller and idle belt. Temperature readings were stored digitally in real time. 2.3 Study 1a: Furnace Experiments Paper and toner powder samples were individually placed in a small ceramic dish and heated in the furnace. In another experiment, the fuser roller was removed from the printer and placed within the furnace. The temperature of the samples was measured by a thermocouple. Particle number and size distribution were measured directly from the furnace. Air passing through the furnace, at a flow rate of 1.4 L min -1, was introduced into the box chamber. Determination of emissions other than particles and the VH-TDMA tests were carried out from the box chamber. 2.4 Study 1b: Flow Tunnel Experiments Printer A was placed within a tunnel of diameter 0.5 m and made to print 150 monochrome pages of different toner coverage. The downstream end of the tunnel was tapered, so that the diameter of the tunnel cross section reduced to m over a distance of 0.85 m. The air velocity at the exit of the tunnel was 0.7 m s-1. The distance from the printer to the sampling point was 2.0 m. 2.5 Study 1c: Box Chamber Experiments Printer A, was placed within the box chamber. The chamber was flushed with clean air before the printer was allowed to print 150 pages. Each print run lasted about 7 min. Particle number and size distribution and TVOC concentrations were measured in real time. A series 3

4 of volatilization and humidification experiments were carried out for up to 2 h after printing. Each of the different experiments with Printer A was repeated at least three times. 2.6 Study 2: Effect of Fuser Temperature This study was performed, as in Study 1c, with each of the 30 printers placed within the chamber. The fuser temperatures were measured with a thermocouple. RESULTS 3.1 Study 1a: Furnace experiments The furnace experiments showed that the emissions depended on the temperature of the material. No emissions of particle number and TVOC occurred until some threshold temperature was attained. For each material tested, the threshold temperature was similar for both particles and TVOC; however it differed between the materials tested (Figure 1). A further increase in temperature above this threshold value resulted in a rapid increase in particle number emission while, if the furnace temperature was held steady for more than 30 to 60 s, the emissions began to decrease, before increasing again, in response to each successive temperature increase. The SMPS showed consistent differences in particle size distributions for different materials, with distinct unimodal, bimodal and trimodal distributions. Thus, the size location of the modes might be considered as source signatures of the heated materials. The experiments also showed that there was no increase in ozone in the furnace above the indoor background level which varied from 7-10 ppbv (parts per billion volume). The volatility and humidification tests showed that the particles were both volatile and water insoluble, consistent with organic species composition. Within the limits of the particle counters used for these tests there was no evidence of a non-volatile core residue suggestive of seeded nucleation. 4

5 Threshold Emission Temperature [ o C] (a) Particle Number 75 Paper Used Roller New Roller Toner Threshold Emission Temperature [ o C] (b) TVOC 75 Paper Used Roller New Roller Toner Figure 1: Mean threshold temperatures of the materials heated in the furnace, for (a) particle number and (b) TVOC emission. Error bars represent one standard deviation. 3.2 Study 1b: Flow tunnel experiments The tunnel experiments provided us with an insight into the origin of the emissions associated with the operation of a printer. Figure 2 shows the particle number and TVOC concentrations measured during a typical print run of 150 pages using 5% toner coverage. Maximum particle number concentrations exceeded 8 x 10 4 particles cm -3. The TVOC concentration showed a slow increase during printing, reaching a maximum after the completion of printing. Since toner was a very likely source of precursors for particle formation, it was expected that an increase in toner coverage would result in an increase in particle concentrations. However, experiments with 0% and 50% toner coverage showed that there was no apparent relationship between toner coverage and level of submicrometer particle or TVOC concentration. The absence of an increase was surprising and therefore, the tests for different coverage were repeated several times and the results confirmed. To investigate whether paper contributes to the particle and TVOC concentrations, independently from the toner, several tests were conducted on printer A for each type of paper, using a new fuser roller, under 0% coverage. The tests showed that the type of paper was linked to particle concentrations (although, in most cases, to a lesser degree than in the 5

6 case of toner), as well as TVOC concentrations. They also indicated that increases in particle and TVOC concentration were linked to the occurrence of two separate processes during printing, as indicated by the increase in concentrations at different stages of printing. 8.E+04 Particle Number Conc TVOCs 1.2E+03 Particle Number Conc [p/cm 3 ] 6.E+04 4.E+04 2.E E E E+02 TVOCs [ppb] 0.E Elapsed Time [s] 4.0E+02 Figure 2: Submicrometer particle number and TVOC concentrations in the flow tunnel from Printer A during and after printing of 150 pages. Toner coverage = 5%. Printing commenced at time=0 and ended at 434s (broken line). The initial particle count median diameter (CMD) of submicrometer particles in the tunnel during a print run was about 63 nm. The CMD gradually decreased to about 28 nm during printing. The variation of CMD did not vary significantly with toner coverage. Supermicrometer particle CMD, on the other hand, remained at approximately 1 μm throughout the entire printing process, irrespective of toner coverage. There was an increase in supermicrometer particle concentration during printing, indicating that the particles were generated through a mechanical process, such as re-suspension of particles due to the movement of the rollers and the paper. In another experiment, Printer A was allowed to operate in the tunnel with the fuser unit removed. With no temperature increases within the printer, no particle emission was detected. 3.3 Study 1c: Box chamber experiments In general, the experiments conducted in the chamber with Printer A supported the findings of the tunnel experiments. Of particular significance was the small increase in ozone concentration (no more than about 50 ppbv). This low rate of production was not detectible under the dynamic conditions in the tunnel because of the high dilution rate of the emissions. Two other relationships seen more clearly in these experiments than in the tunnel were a positive correlation between submicrometer particle and ozone concentrations and a contrasting negative correlation between submicrometer particles and TVOC concentration. To investigate the role of ozone in this process, an experiment with 5% toner coverage was 6

7 repeated with pure nitrogen instead of air in the chamber. Particles were still generated, but at much lower concentrations (9.8x10 4 cm -3 ) than in the presence of air (4.3x10 5 cm -3 ). Therefore, it can be concluded that ozone plays a role in particle formation, but that there are also other formation pathways, not requiring ozone. As in the tunnel experiments, the initially higher particle CMD in the chamber decreased as printing continued, however, unlike in the tunnel, after several minutes the CMD began to increase again. This was expected, considering that the particles would accumulate in the chamber, leading to an increase in diameter due to coagulation. In a subsequent chamber experiment, Printer A was made to operate without paper and toner. This was only possible for single operations as the printer mechanism reported error messages. Operations at consecutive intervals of about 20s, showed a significant increase in particle concentration in the chamber, albeit about one order of magnitude lower than for normal printing. The VH-TDMA results showed that the particles were volatile. Figure 3 shows a typical run at 0% toner coverage. The boiling point or decomposition temperature of this substance is likely to be greater than about 130 C. These particles showed no growth when exposed to high humidity and were therefore deemed to be non-hygroscopic. 1.2 Unvolatilised Particle Volume Fraction (V/V0) Temperature ( o C) Figure 3: Volatilisation temperature curve of the particles generated by Printer A, measured by the VH-TDMA. 3.4 Study 2: Effect of Fuser Temperature Experiments with the 30 printers indicated that there was a statistically significant (p < 0.05) positive relationship between the fuser roller temperature and particle emission rate. Figure 4, shows the particle number emission rate per minute as a function of the excess fuser temperature, that is the (fuser temperature room temperature). Printers with relatively low fuser temperatures showed low emissions, while the printers with the higher fuser 7

8 temperatures showed the highest emission rates. The trendline shown is exponential and has an R 2 value of Particle Number Emission Rate x 1E+15 (per min) Excess Temperature (deg C) Figure 4: Particle number emission rate as a function of average fuser temperature for 30 printers investigated in the box chamber. The trendline shown is exponential (R2=0.43). DISCUSSION The results provide strong evidence that the fuser temperature is the key parameter driving the process of particle formation. It has been shown that: (1) the printer paper and toner both emit particles when heated to above a threshold temperature value; (2) no particle emission is observed when the printer is operated with the fuser removed; and (3) there is a significant positive relationship between the particle number emission rate and the excess fuser temperature (Fig 4). We hypothesise that, following the evaporation of VOC and SVOC, as explained above, two processes occur, leading to particle formation: (1) homogenous nucleation of SVOC species to form particles; and (2) secondary particle formation through a reaction between VOC species and ozone to produce a further SVOC species. During printing, SVOC released from the toner as it is being fused onto the paper will evaporate and then condense as new particles. Condensation may occur through homogenous nucleation, as well as onto the cooler leading portion of the idle belt. Directly deposited toner residues also remain on the idle belt when it separates from the paper. These deposits then volatilise when brought back to the fuser heater strip during subsequent printing. The return of these deposits to the heater, through idle belt rotation, explains the rapid appearance of VOC in the air when printing begins which occurs even for blank pages where no new toner is being applied to the paper. Given the observed correlation between ozone and particle concentration during printing, it is pertinent to consider a second reaction involving ozone. 8

9 However, the importance of this process relative to direct volatilisation and nucleation of a SVOC released directly from toner deposits cannot be quantified at this stage. In addition to formation, the nature of the particles is also important for assessing their potential toxicological impact. For example: The VH-TDMA study showed that the particles contain a species which is volatile in air when heated briefly to 130 C. However, it is not water soluble, nor does it become so during thermal treatment. Results show that several species are present in increased concentrations in samples containing the emitted particles and gases. Of these, the potential candidate for a nucleating species must be water insoluble with a boiling point above 130 C. The following species meet these requirements: pentadecane, hexadecane, heptadecane, dimethylphthalate, ethylbenzene, m,p-xylene, o-xylene, and styrene. Microanalysis of printer aerosols collected onto transmission electron grids suggested that, though the emissions appear to be almost entirely composed of SVOC, there were inorganic components such as Ca that appeared to be from calcium carbonate in the paper. Fe, Ti, Si, and Mg were also found in varying trace quantities. What is the likely exposure of office workers to ultrafine particles arising from laser printer operation? The answer to such a question and, therefore, the results of the laboratory tests presented in this paper, are of importance to occupational hygienists for anticipation and recognition of the ultrafine particle hazard. A study is currently underway that aims to evaluate and characterise the emission and transport of particles from a range of laser printers being operated in different office environments and so as to inform guidance on the operation and positioning of laser printers relative to occupied workstations. Condensation particle counters, a light scattering instrument (DustTrak), and an optical particle counter are being used at set distances from the various printers and in proximity to the air intake for the offices so as to characterise the laser printer and outside air derived ultrafine, fine, and supermicrometer particles. The office ventilation systems are being evaluated in an endeavour to identify the impact of ventilation and filtration systems on particle spatial and temporal characteristics. 9

10 Acknowledgments E. Uhde and T. Salthammer thank the German Federal Ministry of Education and Research for travel grants to QUT, provided through the German Aerospace Center (DLR). The authors would also like to thank the Queensland Government Chief Procurement Office for financial support and supply of the printers for Study 2. References Destaillats, H., Maddalena, R., Singer, B., Hodgson, A. and McKone, T Indoor pollutants emitted by office equipment: A review of reported data and information needs. Atmos. Environ. Vol. 42, pp He, C., Morawska, L. and Taplin, L Particle Emission Characteristics of Office Printers. Environ. Sci. Technol. 41: Johnson, G., Ristovski, Z. and Morawska, L. (2004). Method for measuring the hygroscopic behaviour of lower volatility fractions in an internally mixed aerosol. J Aerosol Sci. Vol. 35, no 4, pp Kagi, N., Fujii, S., Horiba, Y., Namiki, N., Ohtani, Y., Emi, H., Tamura, H. and Kim, Y Indoor air quality for chemical and ultrafine particle contaminants from printers. Building and Environ. Vol. 42, pp Lee, C.W. and Hsu. D.J Measurements of fine and ultrafine particles formation in photocopy centres in Taiwan. Atmos. Environ. Vol 41, pp Lee, S. C., Lam, S. and Fai, H. K Characterization of VOCs, ozone, and PM10 emissions from office equipment in an environmental chamber. Building and Environ. Vol. 36, no 7, pp Morawska, L., He, C., Johnson, G., Jayaratne, R., Salthammer, T., Wang, H., Uhde, E., Bostrom, T., Modini, R., Ayoko, G., McGarry, P. and Wensing, M An investigation into the characteristics and formation mechanisms of particles originating from the operation of laser printers. Environ. Sci Technol. Vol. 43, pp Schripp, T., Wensing, M., Uhde, E., Salthammer, T., He, C. and Morawska, L Evaluation of Ultrafine Particle Emissions From Laser Printers Using Emission Test Chambers. Environ. Sci. Technol. Vol. 42, no. 12, pp Tuomi, T., Engström, B., Niemelä, R., Svinhufvud, J. and Reijula, K Emission of Ozone and Organic Volatiles from a Selection of Laser Printers and Photocopiers. Applied Occupational and Environmental Hygiene vol. 15, no 8, pp WHO Guidelines for Air Quality. World Health Organization. 10