摘要:

Determination of alkylamine permeation through protective gloves using aliphatic amine pads{ Evanly Vo and Stephen P. Berardinelli National Institute for Occupational Safety and Health, Division of Safety Research, 1095 Willowdale Road, Morgantown, WV, 26505, USA Received 9th August 1999, Accepted 8th October 1999 A quantitative study of alkylamine permeation through a glove material using Permea-Tec aliphatic amine pads, used for the detection of chemical breakthrough of protective clothing, was performed for triethylamine following a microwave-extraction process and gas chromatographic analysis.Triethylamine exhibited w99% adsorption on the pads at a spiking level of 729 ng (1.0 ml). Triethylamine showed recoveries from 63 to 90% (RSD °5%) over the range 0.2±1.0 ml (146±729 ng) applied to pads.The ASTM F739 standard and direct permeability testing procedures were used to determine breakthrough times for Æve protective glove materials using triethylamine as a challenge chemical. Breakthrough times for six protective gloves were determined ranging from 40 s to w4 h. The quantitative concentration of triethylamine on the pads following permeation through the gloves was also determined, ranging from 101 to 103 ng cm22 (382±386 ng per pad).Introduction Respiratory and dermal exposures to chemical agents are known to occur in the workplace.1±4 A major occupational health focus has been respiratory monitoring and control, yet damage caused by other routes of exposure is also a major cause of occupational ill-health.5,6 Other routes of exposure are mainly through the skin; therefore, workers are required to wear chemically resistant gloves and protective clothing to prevent skin exposure to toxic chemicals in the workplace.Aliphatic amine pads have been developed to enhance workers' ability to protect themselves from skin exposure to alkylamines. Fresh pads are attached to the hands of a worker before gloving.Permeation of alkylamines through the glove will result in adsorption on the pads, which can be quantiÆed through subsequent analysis. Although the glove-selection process is typically based on manufacturers' chemical permeation data, many factors, such as temperature, pressure and product variation among suppliers, bring into question the reliability of this process.Passive dermal monitoring could be used to evaluate glove performance during `actual use' conditions and could bridge the gap between laboratory data and `real world' performance. We report here the validation of an assay using pads attached to the palm, the cuff or the Ængers of gloves to determine time to breakthrough and the concentration of triethylamine at timed intervals.Experimental Chemicals Triethylamine was used as a spiked, standard chemical for testing pads. It was also used as a neat challenge chemical for glove permeation testing. Cyclohexane was used as the extraction solvent. Both chemicals (ACS reagent grade) were purchased from Aldrich Chemical (Milwaukee, WI, USA). Selected gloves and pads Six different types of gloves were selected for this study: nonsterile vinyl gloves purchased from Baxter Healthcare (Valencia, CA, USA); polymer-coated latex, co-polymer, and powdered nitrile gloves purchased from VWR ScientiÆc Products (West Chester, PA, USA); and disposable vinyl and Sol-Vex gloves purchased from Ansell-Edmont (Coshocton, OH, USA).Permea-Tec aliphatic amine pads were purchased from Colormetric Laboratories (Des Plaines, IL, USA).Apparatus An Ames (Waltham, MA, USA), 214-10 micrometer with a pressure foot of 1 cm was used to determine the thickness of each glove. A Miran (Miniature Infrared Analyzer) closed-loop conÆguration which consists of a metal bellows pump (Model MB-41), a 2.5 cm chemical permeation cell and a Miran was used to determine time to breakthrough for the glove material.A CEM (Matthews, NC, USA) MES-100 microwave extraction (ME) system was used to extract triethylamine in pads. A Perkin-Elmer (Norwalk, CT, USA) gas chromatographic system, which consisted of a PE Nelson Model 1022 Personal Integrator and an AutoSystem gas chromatograph, was used to analyze chemicals. Evaluation of sorption properties of pads Spiking pads. New pads were removed from sealed packages and the adhesive areas removed with scissors.Pads were tested to determine if there was any pad±media interference. Known volumes of triethylamine (0.2±1.0 ml) were added directly to the surface of the pads using a syringe (0.1±1.0 ml). The pads were then inserted into 300 ml of extraction solvent in 3 ml Savillex vials and the vials were covered with the vial caps.The ME process and GC analysis. The ME-GC procedure was run according to the method of Vo et al.7 as follows. The closed vials were immersed in 25 ml of water in extraction vessels, then, the vessels were placed in the MES-100 system and extracted for 15 min at 100 �C, 70 lb in22, and 60% power. {This article was prepared by US Government employees as part of their ofÆcial duties and legally may not be copyrighted in the United States of America. J.Environ. Monit., 1999, 1, 545±548 545 This journal is # The Royal Society of Chemistry 1999The extracted solutions were allowed to cool to room temperature for 20 min before opening the vial caps for GC analysis in order to obtain maximum recovery of these chemicals without evaporation loss.The optimum GC conditions were as follows: column, 3.05 m63.1, 8 mm id, 3% SP-1500 on 80±120-mesh Carbopack B (Supelco, Bellefonte, PA, USA); helium Øow rate, 25 ml min21; oven temperature, 225 �C; column temperature, 215 �C; temperature of the Øame ionization detector (FID), 235 �C; data collection time, 0±21 min; main plot time, 13± 21 min; and y-maximum, 50 mV. Volumes of 5 ml of extracted samples of spiked triethylamine in extracted solutions were injected into the GC column using a syringe (0.1±5.0 ml), up to four times for each sample.The areas of the spiked triethylamine peaks in the resulting gas chromatograms were used for spiked triethylamine determinations. EfÆciency of adsorption during triethylamine spiking. In order to assess whether complete adsorption of triethylamine is determined during spiking on pads, two Savillex vials were used. In the Ærst vial, 1 ml of triethylamine was added to a pad and the vial was covered with the vial cap.The vial was left at room temperature for 30 min before the pad was inserted in 300 ml of extraction solvent in the second vial. Then, 300 ml of the extraction solvent were added directly to the Ærst vial to extract excess triethylamine which did not adsorb on the spiked pad.These vials were then used for the extraction process and GC analysis. Triethylamine recovery. Standard triethylamine determinations were performed using the same spiked triethylamine procedure, but without using pads. Known volumes of triethylamine (0.2±1.0 ml) were added directly to 300 ml of extraction solvent in the vials.These vials were then used for the extraction process, and 5 ml samples of the extracted solutions were subjected to GC analysis. The areas of the standard peaks in the resulting gas chromatograms were used for standard triethylamine determinations. Triethylamine recovery was calculated as the percentage of spiked triethylamine peak area divided by standard triethylamine peak area. Glove thickness test The thickness test was performed at three positions on each glove: the palm, the entire middle Ænger and the cuff (3 cm from the open end), as these represent the areas of highest contact and glove abrasion, which enhances chemical penetration.Five thickness measurements for each position on the glove were recorded and the mean thickness and variation were calculated. All measurements were made to the nearest °0.01 mm.Glove breakthrough time determination A modiÆed ASTM F739 method was used to determine breakthrough time.8 The IR conditions were as follows: slit, 1.0 mm; wavelength, 9.3 mm; pathlength, 20.25 m; and minimum detectable concentratippm. The 2.5 cm permeation cell is divided into a `challenge side' which contains the chemicals and a `collection side' which contains the sweep gas.Sections from the palm and the cuff of gloves containing a pad attached to the inner surface of the section were used as a membrane between two halves of the permeation cell with the outer surface toward the challenge side of the permeation cell. The experiment was conducted at room temperature (22°1 �C).A 15 ml volume of challenge chemical was injected into the challenge side of the cell and a timer was immediately started. Permeation of triethylamine through the glove was detected by the change in color of the pads or the analyzerdetected response. Direct permeability testing procedure was also performed on the Ænger position of gloves. A Ænger of the gloves was turned inside out and a pad was attached to the Ænger.The glove Ænger was then fastened to a glass cylinder (2615 cm) by using duct tape at the open end of the glove Ænger as the attachment area. The glass cylinder served as a Æll tube. A 10 ml volume of challenge chemical was injected into the glass cylinder and a timer was immediately started. Permeation of triethylamine through the glove Ænger was detected by the change in color of the pads.Breakthrough time was recorded and the pads were immediately removed from the glove and inserted into 300 ml of extraction solvent in the vials for the extraction process. Quantitative determination of triethylamine The closed vials of triethylamine were extracted as described above. Volumes of 5 ml of extracted samples were subjected to GC analysis.The areas of the peaks in the resulting gas chromatograms were used for challenge triethylamine quantiÆcation. The quantitative concentration of challenge triethylamine was determined against a known concentration of spiked triethylamine based on its linear equation and the degree of triethylamine recovery. Results Resolution Good resolution of the extraction solvent and triethylamine was achieved by setting the optimum GC conditions as described.None of the blanks (unexposed pads in the extraction solvent) produced chromatograms containing peaks corresponding to triethylamine. The GC retention times for triethylamine and extraction solvent obtained under these conditions are given in Table 1. EfÆciency of adsorption during triethylamine spiking Excess triethylamine which did not adsorb on the spiked pad in the Ærst vial was low, being v1% (the area peak was v1% of the area peak for the triethylamine found in vial No. 2). Triethylamine exhibited w99% adsorption on the pads at spiking level of 729 ng. Calibrations of pads Triethylamine spiked on the pads (0.2±1.0 ml) could be detected on the chromatograms. The relationship between signals (peak area) on the gas chromatograms and volumes over the range 0.2±1.0 ml (n~5) of triethylamine applied to the pads was analyzed using Microsoft Excel software.A linear correlation for triethylamine was obtained over the range 0.2±1.0 ml [r2~0.9989, pv0.001, with the linear equation y~197x232 (x~volume; y~peak area)].No such linear correlation for triethylamine was obtained when °0.1 ml of triethylamine was added to pads. Triethylamine recovery For experiments performed with repeated extraction process and GC measurements, recoveries from 63 to 90% (RSD°5%) were obtained over the range 0.2±1.0 ml (146±729 ng) of triethylamine applied to the pads (Table 1). No signiÆcant improvement in the recovery was observed when ¢1.2 ml of triethylamine was added to pads.Glove thickness test Table 2 summarizes the results obtained for the mean thickness of Æve thickness measurements for each type of glove. 546 J. Environ. Monit., 1999, 1, 545±548Glove breakthrough time determination The data demonstrated breakthrough with any of the gloves tested with the ASTM F739 standard and direct permeability testing procedures, except the Sol-Vex glove (Table 2).Permeation of triethylamine through the gloves was detected by the change in color of the pads (triethylamine causes the pads to change from yellow to blue), but not by the analyzerdetected response. Quantitative determination of triethylamine The concentration of challenge triethylamine on the pads following permeation through the gloves was determined, ranging from 101 to 103 ng cm22 or 382±386 ng per pad [pad: 2.0 cm61.9 cm; 382±386 ng per pad (0.52±0.53 ml per pad) of triethylamine based on adsorption of about 0.35±0.36 ml per pad from its linear equation and its recovery of 68% at 0.4 ml added].Precision The recovery precision was good, with RSD °5% for triethylamine (over the range 0.2±1.0 ml of triethylamine applied to pads and 5 ml of extracted solutions for GC analysis).Sensitivity The gas chromatograms of triethylamine were obtained with 0.2 ml of triethylamine spiked onto the pads, and 5.0 ml of extracted solutions of triethylamine applied for GC analysis. These volumes represent 146 ng of triethylamine per pad and 2.43 ng of triethylamine applied for GC analysis.Discussion The efÆciency of adsorption of triethylamine on the pads was assessed. It was found that triethylamine was nearly completely adsorbed (99±100%) on the pads at a spiking level of 729 ng. The results obtained with the ME-GC procedure were consistent with a reproducible recovery of triethylamine from pads which had been exposed to this chemical. The recovery was obtained over the range 0.2±1.0 ml of triethylamine applied to pads and 5 ml of the extracted solutions subjected to GC analysis.The recovery was also dependent on the extraction process and the volume of triethylamine applied to the pad. No such linear correlation for triethylamine was obtained when°0.1 ml of triethylamine was added to pads. An excellent linear correlation (r2~0.9989 and pv0.001) was obtained for triethylamine over the range 0.2±1.0 ml added to pads and 5 ml of the extracted solutions used for GC analysis. The color change on the pad occurs on adsorption of about 0.35±0.36 ml per pad from its linear equation and its recovery of about 68%, and that this corresponds to volumes about 0.52±0.53 ml per pad (382±386 ng per pad) or 101±103 ng cm22, which is within this linear response range.These values are below the recommended exposure limit (REL) for triethylamine [ACGIH (American Conference of Government Industrial Hygienists) threshold limit value (TLV) of 3 ppm (12 mg m23) as a ceiling concentration, as its REL]. It was also shown that the GC system is sensitive enough to detect the presence of triethylamine in extraction solvent.Lownanogram amounts of triethylamine (2.43 ng) can be detected and they are clearly resolved under the GC conditions described. The results indicated that differences in material density and thickness yielded different breakthrough times. The results of the permeation experiments, as shown in Table 2, demonstrated that within a given material type of the gloves tested, thickness had a primary effect on the breakthrough time of the challenge chemical. Differences in glove materials yielded different breakthrough times (Table 2).For nitrile rubber materials, slow permeation by triethylamine was observed (¢70 min), while rapid permeation by triethylamine for natural latex rubber, polymerized alkenes or vinyl materials was observed.Table 2 also shows that there were signiÆcant differences in breakthrough times between the ASTM F739 standard and direct testing methods for these gloves where the thicknesses were almost similar. It is probable that the pressure which was generated by the pump within the closed-loop system caused triethylamine to permeate more quickly than actual breakthrough times. Permeation of triethylamine through the gloves was detected by the change in color of the pads before the IR analyzer detector responded.It is possible that triethylamine with a low vapor pressure gave the impression of longer than actual breakthrough times. Since bh thickness and density measurements represented an average of the swatch surface area, they did not indicate small regions of `thin' areas where permeation occurred more rapidly.The instrumentation used in this experiment is sensitive enough to detect concentrations in the pads caused by permeation in these thinner areas of the material. Hence, although we cannot accurately measure material variations, we can still observe the effects of these variations through their contribution to the permeation-rate variations.Quantitative data were obtained for triethylamine permeation through various glove materials. The results indicate that Table 1 Retention time and recovery of triethylamine (bp 88.8 �C) at different volumes applied to pads (mean recovery ° SD, n~3). A 300 ml volume of extraction solvent (retention time of cyclohexane~ 3.84 min) was used to extract triethylamine from the pads and 5.0 ml of the extracted solutions of triethylamine were analyzed by GC Volume of triethylamine applied to pad/ml Recovery (%) Retention time/ min 0.2 63°4.9 17.56 0.4 68°4.5 0.6 73°4.0 0.8 86°3.3 1.0 90°3.1 Table 2 Results of the glove permeation tests (n~3) for triethylamine Type of glove Material of glove Style or Model No.Thicknessa (palm and cuff)/mm Breakthrough timeb (palm and cuff) Thicknessa (Ænger)/ mm Breakthrough detection timec (Ænger) Mass of triethylamine found/ng per pad Non-sterile vinyl Vinyl (non-sterile) TriØex 0.15°0.01 78°5 s 0.11°0.01 62°3 s 384°3 (103 ng cm22) Polymer-coated latex Natural latex rubber Boxed Ambi 0.15°0.01 40°3 s 0.15°0.01 42°4 s 386°4 (103 ng cm22) Copolymer Polymerized alkenes Boxed Ambi 0.12°0.01 40°4 s 0.16°0.01 68°5 s 383°5 (102 ng cm22) Disposable vinyl Vinyl Dura-Touch 0.18°0.01 115°7 s 0.18°0.01 140°6 s 385°4 (102 ng cm22) Powdered nitrile Nitrile Boxed Ambi 0.11°0.01 70°2 min 0.12°0.01 77°1 min 382°4 (101 ng cm22) Sol-Vex Nitrile butyl rubber 37±175 0.36°0.01 w4 h 0.38°0.01 w4 h ndd aMean thickness°SD (n~5).bBreakthrough time detected by the change in color of aliphatic amine pads (the ASTM F739 testing procedure).cBreakthrough time detected by the change in color of aliphatic amine pads (the direct testing procedure). dNot detected at 4 h. J. Environ. Monit., 1999, 1, 545±548 547pads exposed to triethylamine can be successfully analyzed under the conditions for the extraction process and GC analysis described (101±103 ng cm22). This procedure offers potential for developing permeation breakthrough indicators for the entire series of alkylamine compounds which may be encountered in workplace exposure situations.A companion study under actual workplace conditions will be conducted in order to determine the validity of laboratory Ændings under dynamic conditions found during workplace exposures. Acknowledgements The authors thank Mr Paul Keane (NIOSH, Morgantown, WV, USA) and Mr Tom Klingner, Colormetric Laboratories (Des Plaines, IL, USA), for their valuable assistance in the preparation of the manuscript. References 1 J. E. Davis, E. R. Stevens and D. C. Staff, Bull. Environ. Contam. Toxicol., 1983, 31, 631. 2 W. F. Durham, H. R. Wolfe and J. W. Elliott, Arch. Environ. Health, 1972, 24, 381. 3 H. R. Wolfe, W. F. Durham and J. F. Armstrong, Arch. Environ. Health, 1967, 14, 622. 4 C. A. Franklin, N. Muir and R. P. Moody, Toxicol. Lett., 1986, 33, 127. 5 OHSA, Analytical Method Manual, Occupational Health and Safety Administration, Washington, DC, 1985. 6 S. A. Ness, in Surface and Dermal Monitoring for Toxic Substances, Van Nostrand Reinhold, New York, 1994, p. 359. 7 E. Vo, S. Berardinelli and R. Hall, Analyst, 1999, 124, 941. 8 S. Berardinelli and E. Moyer, Am. Ind. Hyg. Assoc. J., 1987, 48, 324. Paper 9/06477J 548 J. Environ. Monit., 1999, 1, 545±5

ISSN:0960-7919    

DOI:10.1039/a906477j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

The evaluation of a low resolution Fourier transform infrared (FTIR) gas analyser for monitoring of solvent emission rates under Æeld conditions Jouni Ra»isa»nen* and Raimo Niemela» Finnish Institute of Occupational Health, Finland Received 1st September 1999, Accepted 1st November 1999 The applicability of a low resolution (8 cm21) Fourier transform infrared (FTIR) gas analyser with an absorption path length of 3 m was evaluated for the on-line monitoring of organic solvent mixture emissions in a Øexographic ink manufacturing plant.The on-line monitoring revealed that the highest variations of solvent concentrations, up to three decades, occurred in the exhaust air. The FTIR analyser with a dynamic range of four decades covers well the concentration ranges typically found in the exhaust air and in the workroom air of ink manufacturing plants. The average emission rate of solvent mixture based on a sampling period of two days was 1.8 kg h21 consisting of mainly ethanol (70%), ethyl acetate (15%) and propan-2-ol (11%).The detection limits of the analyser for the solvent compounds ranged from 0.3 to 4.3 mg m23 and the measurement uncertainty was less than 10% in the concentration range of 8±15 000 mg m23.These characteristics make the apparatus appropriate for most industrial hygiene applications. An FTIR spectrophotometer, equipped with an multipoint sampling unit, facilitates rapid identiÆcation of solvent components, real-time display of concentration data relevant to workroom air and environment monitoring as well as process control. Furthermore, the on-line concentration information enabled a rapid selection of representative sampling locations.The spectrophotometer is transportable, rugged and relatively simple to calibrate even in a hostile industrial environment. Introduction The ink manufacturing process has a major potential to release solvent vapours into workrooms and to outdoor air.The process usually has multiple sources of solvent mixtures with wide concentration variations both in time and in space. Knowledge about solvent emission rates is important from the point of view of good indoor air quality and the outdoor environment. Traditionally, exposure to contaminants in workspaces has been regulated by occupational exposure limits. Emissions to the outdoor environment are more strictly controlled.For instance, the EU directive deÆnes the emission limit values for the different organic solvents used in industry including the manufacture of printing inks.1 Obviously, the monitoring tasks for industrial solvents will be increased. Recent studies have shown that FTIR spectroscopy can be a useful method for monitoring mixtures of solvent vapours and gases in air.2±6 Most studies have been devoted to evaluating the sensitivity and accuracy of high-resolution spectrometers at ppm or sub-ppm concentrations under laboratory conditions. 2±5 The successful use of a low-resolution spectrometer with an absorption path length of 9.6 m for measuring the concentrations of solvent mixtures under Æeld conditions has been reported.7 The purpose of this paper is to describe the applicability of a low resolution FTIR analyser with a short absorption path length (3 m) for the determination of solvent mixture emissions and solvent concentrations in workroom air during the manufacture of Øexographic ink.The detection of concentration peaks in different process phases was of particular interest. Materials and methods Ink manufacturing plant Flexographic ink manufacturing consists of four main process phases, i.e., mixing, dispersing, blending and weighing, which are performed in different departments connected to each other by open doors.The plant manufactures approximately 2000 t of Øexographic ink per year and consumes ethanol 400 t, propan-2-ol 150 t, 1-methoxy-propan-2-ol 180 t, and ethyl acetate 100 t.The production rate varied from 100 kg h21 to 300 kg h21 during the study. The manufacturing plant (Øoor area 4700 m2, height 7 m) was equipped with a mechanical ventilation system with two exhaust fans (Fig. 1). The air from the departments was exhausted at Øoor level through ductwork. The supply air diffusers were mounted in the middle of the departments at a height of 4 m.In addition to the general ventilation, all departments were equipped with local exhausts. Calibration of the FTIR gas analyser The measurements of ethanol, ethyl acetate, propan-2-ol and 1- methoxy-propan-2-ol concentrations were carried out using a portable low-resolution (8 cm21) FTIR gas analyser (Gasmet, Temet Inc., Helsinki, Finland), equipped with a Peltier-cooled MCT detector (D*~0.36109) and a 1.1 l gas cell with an absorption path length of 3 m.The spectral response range of the analyser is 4000±950 cm21. The low frequency end is limited by the cut-off frequency of a Peltier-cooled detector. The analyser has a spectral scanning rate of 10 scans s21 and a data acquisition rate of 20 kHz. The interferograms are collected as double-sided.The identiÆcation and quantiÆcation procedure of solvents was performed by CALCMET multicomponent analysis software provided with the analyser.8 The procedure is based on the modiÆed classical least-squares (CLS) analysis algorithm. In this procedure the maximum amount of precomputed information is utilised for making the spectra analysis as simple and as fast as possible. The basic difference compared to the common CLS algorithms is that the algorithm Æts the measured unknown spectrum using a set of single component calibration library spectra.Likewise, there is J. Environ. Monit., 1999, 1, 549±552 549 This journal is # The Royal Society of Chemistry 1999no need to deÆne speciÆc analytical spectral regions for each component where that component is analysed.If necessary, all the data points in the spectra can be used in the Æt. The typical data processing and transfer time is 2±3 s. The gas analyser was calibrated in the laboratory under a nitrogen atmosphere by using the dilution method. The closed loop calibration system had a volume of 10.4 l and contained a dilution glass cuvette and a gas analyser cuvette, and a membrane pump, TeØon tube and stainless-steel valves.The valves were used for connecting the system either to a closed loop circulation (calibration mode) or to an opened circulation (Øushing mode). The reference spectrum was produced by injecting a known mass of liquid analyte into the system. The complete evaporation of analyte was conÆrmed by heating the injector up to 60 �C.After the solvent concentration in the system was stabilised, the IR spectrum was collected and stored on to the reference library on the hard disk of the analyser, together with the concentration information. The reference was then checked by analysing pure samples under a nitrogen atmosphere at Æve different concentrations, which covered the expected range to be measured.The references were also checked under the laboratory air atmosphere (air temperature 20±22 �C, relative humidity 25±35%, concentration of CO2 300±500 ppm) by analysing mixtures of the compounds at Æve different concentrations. The calibration Æts for ethanol and ethyl acetate are shown in Figs. 2 and 3. The limit of detection (LOD) for the solvents was determined in two cases, that of the individual compounds under a nitrogen atmosphere and that of the mixtures under an air atmosphere.In addition to the laboratory calibration, the performance of the analyser was also checked in the Æeld in a closed loop calibration system (loop volume 1.2 l). The checking was performed both at the beginning and at the end of the measurement period at solvent mixture concentrations of 70±90 mg m23.In the laboratory calibrations and LOD determinations 30 spectra were collected using a 20 s spectra averaging time. Spectral overlap from water vapour and carbon dioxide was reduced by including the references of both compounds in the analysis procedure. A background srum of the Æeld conditions was obtained by extracting the workroom air through a charcoal Ælter.The detection limit for a single compound under a nitrogen atmosphere was 1.3 mg m23 for ethanol and 1-methoxypropan- 2-ol, 0.3 mg m23 for ethyl acetate and 0.6 mg m23 for propan-2-ol. For solvent mixtures under an the air atmosphere the detection limits were slightly higher, 4.1 mg m23 for ethanol and 1-methoxy-propan-2-ol, 4.3 mg m23 for ethyl acetate and 2.3 mg m23 for propan-2-ol.Field measurements The air samples from the Æve Æxed sampling points were taken sequentially over a 48 h period by using a computer-controlled multipoint sampling unit. The samples were fed to the analyser via 10±30 m long TeØon tubes (id 4.7 mm) with a sampling Øow rate of 8 l min21 yielding a time constant for the analyser (cuvette volume/sampling Øow rate) of 8 s.For monitoring the solvent emission rates four sampling points were located in the exhaust ducts of each process department. The sampling points were placed downstream from the elbows of the ducts to guarantee a homogeneous concentration distribution at the sampling cross-section. The Æfth sampling point was placed in the centre of the mixing department at a height of 1.6 m to detect solvent vapour concentrations in the air of the workroom. At this point charcoal tube samples were collected in parallel in order to check FTIR values.A sampling time for each sampling line was 3 min, including the spectral scanning time of 20 s, resulting in a total cycle time of 15 min. The solvent emission rates via the individual exhaust ducts were determined by multiplying the solvent concentration and Fig. 1 The layout of the plant, exhaust duct conÆgurations (dashed lines) and the measurement points in the exhaust ducts (1±4) and in the mixing room (5). A~mixing dept., B~dispersing dept., C~blending dept. and D~weighing dept., I and II~the exhaust air fans. &~dispersing mills, $~blending tanks. Fig. 2 The FTIR concentration vs. reference concentration for ethanol and a calibration Æt.The calibration Æts for propan-2-ol (FTIR conc.~0.89 (°0.01)6ref. conc.) and 1-methoxy-propan-2-ol (FTIR conc.~0.89 (°0.03)6ref. conc.) corresponded to the Æt for ethanol. Fig. 3 The FTIR concentration vs. reference concentration for ethyl acetate and a Æt. 550 J. Environ. Monit., 1999, 1, 549±552the air Øow rate. The total emission rates for the individual solvents were obtained by summing the separate emission rates.The volumetric air Øow rates in the exhaust ducts were measured using a standard pitot tube and an electrical micromanometer (Alnor 3KDS Alnor, Inc., Turku, Finland) by a multipoint traversing method according to the ISO 3966 standard.9 During the measurements the exhaust and indoor air temperature varied between 20 and 28 �C and the relative humidity between 40 and 60%.Results In this study the emission rates of solvent mixtures released in the ink manufacturing process was monitored on-line by using a low resolution FTIR gas analyser. The mean total solvent emission over a 48 h measurement period was 1.8 kg h21. However, a notable temporal variation of emission occurred with the peak value of 10 kg h21 (Fig. 4). The strongest emissions were released from the blending department (Fig. 5). The ethanol, ethyl acetate and propan-2-ol concentrations varied in the exhaust air widely, by up to three decades (Table 1), while the concentration of 1-methoxy-propan-2-ol was below the detection limit most of the time. The widest variation was observed for ethanol from a concentration level of a few mg m23 up to 1800 mg m23.The peak concentrations of solvents were 3±26 times higher than the corresponding mean values. The highest concentration peaks in the exhaust air were observed during the ink blending and weighing operations. The average elemental carbon concentration in the exhaust air over the measurement period was 50 mg C m23.The solvent concentrations in the mixing room were at the most 15% of the Finnish occupational exposure limit for the combined effect of solvents.10 The average FTIR concentrations and charcoal tube concentrations in the mixing room are shown in Fig. 6. The total volumetric exhaust air Øow rate, 18 000 m3 h21, consisted of 4700 m3 h21 in the mixing department, 3300 m3 h21 in the dispersing department, 9300 m3 h21 in the blending department and 700 m3 h21 in the weighing department, respectively. Discussion The results of the on-line monitoring showed that the solvent concentrations, excluding 1-methoxy-propan-2-ol, varied in the exhaust air ducts widely, by as much as three decades.The calibrated measurement range, four decades, covered well the concentrations that typically occurred in the exhaust air as well as in the workroom air in this kind of manufacturing plant. The 48 h mean solvent concentrations, excluding propan-2-ol in the mixing room and 1-methoxy-propan-2-ol, were 2±76 times Fig. 4 The temporal variation of the total emission rates of solvents. During the Ærst 20 h the production rate was 100 kg h21 and 20±36 h three-fold.Fig. 5 The mean solvent emissions. Fig. 6 Average FTIR concentration vs. charcoal tube concentration and a regression line. Table 1 The mean and the range of solvent concentrations (mg m23) in the exhaust ducts and in the mixing room over the 48 h measurement period Measurement point Ethanol Ethyl acetate Propan-2-ol Mean Rangea Mean Range Mean Rangea Exhaust air– Mixing dept. 38 12±758 6 1±17 7 LOD±32 Dispersing dept. 38 12±122 5 1±20 16 LOD±108 Blending dept. 90 30±1800 17 3±240 8 4±211 Weighing dept. 125 LOD±1100 32 3±220 53 LOD±455 Work air– Mixing room 16 LOD±140 4 LOD±23 3 LOD±41 aLOD~limit of detection. J. Environ. Monit., 1999, 1, 549±552 551higher than the corresponding detection limits. If necessary, lower LODs may be achieved by increasing the number of scans.However, when the number of scans is increased, the information on temporal variation is lost. Based on the laboratory and Æeld tests, the measurement uncertainty of the gas analyser is estimated to be less than 10% over the concentration range 8±15 000 mg m23. At concentrations close to the detection limit the measurement uncertainty is higher, 10±40%. Consequently, the dynamic range and the accuracy of the gas analyser meets the requirements of the European Standard EN 482 for screening measurements of concentration variation with time.11 The FTIR monitoring showed that solvent emissions varied strongly due to production rate and process phase, as expected.Ethanol contributed approximately 70% of the total emission, ethyl acetate 15% and propan-2-ol 11%, while the emission of 1-methoxy-propan-2-ol was negligible.The mean total solvent emission over a measurement period was approximately 1.8 kg h21 yielding an annual average emission of 8600 kg, i.e., 1% of the solvent input volume. This Ænding is in line with the previously reported emission estimates from the ink manufacturing process, giving 0.3±2% of the input volumes.12 The average elemental carbon concentration in the exhaust air of the plant studied, 50 mg C m23, is lower than the limit of 150 mg C m23 set by the European Union directive for exhaust air in ink manufacture in plants, with solvent consumption of over 1000 t.1 This study shows that the analyser is sensitive and accurate enough for emission determinations from various process phases and for monitoring work air in this type of premises (those using high volumes of solvents).The FTIR spectrophotometer, equipped with a multipoint sampling unit, facilitates rapid identiÆcation of solvent components, realtime display of concentration data relevant to workroom air and environmental monitoring, as well as process control. Furthermore, the on-line concentration information from several sampling points enables a rapid selection of representative sampling locations.The spectrhotometer is transportable, rugged and relatively simple to calibrate even in a hostile industrial environment. A typical set-up time for the analyser in the Æeld was 4 h. These Æeld measurements demonstrated the advantages of the on-line analyser compared to traditional sampling methods.References 1 EU Directive, 1999, 1999/13/EC. 2 C. R. Strang, S. P. Levine and W. F. Herget, Am. Ind. Hyg. Assoc. J., 1989, 50, 70. 3 Y. Li-Shi and S. P. Levine, Am. Ind. Hyg. Assoc. J., 1989, 50, 360. 4 Y. Li-Shi, S. P. Levine, C. R. Strang and W. F. Herget, Am. Ind. Hyg. Assoc. J., 1989, 50, 354. 5 C. H. Lindh, B. A. G. Jo»nsson and H. E. Welinder, Am. Ind. Hyg. Assoc. J., 1996, 57, 832. 6 H.-K. Xiao, S. P. Levine, J. B. D'Arcy, G. Kinnes and D. Almaguer, Am. Ind. Hyg. Assoc. J., 1990, 51, 395. 7 I. Ahonen, H. Riipinen and A. Roos, Analyst, 1996, 121, 1253. 8 P. Saarinen and J. Kauppinen, Appl. Spectrosc., 1991, 45, 953. 9 International Standards Organisation, Measurements of Fluid Flow in Closed Conduits–Velocity Area Method Using Pitot Static Tubes, ISO, Geneva, 1977, ISO 3966. 10 Occupational Exposure Limit Values 1998, Finnish Ministry of Social Affairs and Health, Tampere, 1998, vol. 25. 11 European Committee for Standardization, Workplace Atmosphere –General Requirements for the Performance of Procedures for the Measurement of Chemical Agents, CEN, Brussels, 1994, EN 482: 1994. 12 Air Pollution Engineering Manual, ed. A. J. Buonicore and W. T. Davis, New York, 1992, pp. 451±456. Paper 9/07101F 552 J. Environ. Monit., 1999, 1, 549±552

ISSN:0960-7919    

DOI:10.1039/a907101f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Establishing normal values for nickel in human lung disease Ivar Andersena and Knut Svenesb aFalconbridge A/S, P. O. Box 604, N-4606 Kristiansand S, Norway bCentral Hospital of Vest Agder, N-4600 Kristiansand S, Norway Received 22nd July 1999, Accepted 9th September 1999 People working in the nickel reÆning industry are known to have a higher concentration of nickel in lung tissue than the general population.To be able to evaluate a potential nickel exposure from other sources, e.g., welding, it is important to have sufÆcient data on what is normal for a local population. Several local factors such as the content of nickel in air and soil can have a signiÆcant impact on this so-called normal value. As almost all surgical equipment contains nickel, the sampling process can in itself be a source of contamination.The scope of this work was to investigate if there was any measurable contamination from the sampling instruments routinely used in hospitals, and if the presence of a nickel reÆnery had any effect on the nickel content in the lungs of the general population. Autopsy lung tissue samples were collected in situ from 50 people who had lived in the county of Vest Agder in Norway.Two samples were collected from each person; one with a regular scalpel (Swann±Norton) and forceps, and one with a titanium knife and plastic forceps. None of the persons had any known connection to the nickel reÆnery. The samples were collected at random and no special attention was given to age, sex and place of residence. The autopsies were performed according to Norwegian law and in understanding with the next of kin.The arithmetic mean value °s of nickel was 0.64°0.56 mg g21 and 0.29°0.20 mg g21 dry weight, respectively, for samples collected with a regular scalpel and a titanium knife (Pv0.0001). For people who lived 8 km and closer to the reÆnery by the time of death, the nickel content was 0.41°0.19 mg g21 and for those who had lived between 8 and 70 km away from the reÆnery it was 0.18°0.13 mg g21 (Pv0.015). No statistical difference was established between results for males and females.Previous investigations have shown that the nickel content in lung tissue varies in the so-called normal population. This work has shown that factors such as sampling equipment and place of residence have an impact on the results. It thus demonstrates that reliable background values can presumably only be obtained by collecting samples from individuals not exposed to known environmental nickel sources and to use nickelfree instruments in the sampling process. Introduction Since certain nickel compounds are known to cause lung cancer in humans,1,2 it is of interest to know if accumulated nickel in lung tissue can be used as an indicator of risk.The knowledge of what could be regarded as a normal value for the nickel content in lung tissue is thus of crucial importance. The actual concentration might be dependent on local conditions such as the nickel content in soil and air. Whether or not it is an urban area is also important.3,4 One problem in this regard is that nickel is involved in a lot of occupational operations.In addition to workers in reÆning and plating operations, we also know that stainless steel welders are exposed to nickel.5±8 The fact that subjects involved in such operations are found more frequently in urban areas should be considered, and the task of obtaining a good work history cannot be emphasised strongly enough.Several investigators have reported values for what they found to be normal for the general population.3±11 Kollmeier et al.3,4 emphasise the importance of using nickelfree sampling instruments; they recommend quartz knives, while Raithel et al.5,6 used a copper±beryllium alloy cutting blade. Akslen et al.9 do not identify the equipment they used, except, to state that it was nickel free.The present authors7,8 reported normal nickel concentrations of 0.76 and 0.84 mg g21 dry weight for people who have lived in the county of Vest Agder in Norway. The fact that all samples were collected with a regular scalpel, and that some of the people may have lived in the neighbourhood of the nickel reÆnery, could have resulted in values that are elevated compared to specimens collected more ideally from people who lived a distance from the reÆnery.The purpose of this work was to investigate the possible importance of the contamination from the sampling instruments routinely used, and also if airborne exposure from reÆnery emissions could be detected in human lungs. Experimental Tissue collection and storage It was of great importance that people who had worked in nickel reÆning operations were rejected from this work.As a Ærst screening, patients who come to the hospital are asked about their present employer. After death his or her personal doctor or nurse asks the next of kin about a work history. Finally the names are checked against a personnel Æle that contains 17 000 names. The Æle contains details of people that have been employed as far back as 1910.Of 54 autopsy specimens, 4 were rejected after comparison with this Æle, and the Ænal material consisted of lung specimens from 50 autopsied subjects. The samples were collected in the period 1995±1997. Two of the samples were from children that had died of Sudden Infant Death Syndrome (SIDS). They were only used for comparison of the two collecting procedures, and were not used in the comparison of results of people who lived close to or further away from the nickel reÆnery.Two samples were collected in situ from each person, both from the lower right lobe. One was collected with a regular scalpel and forceps, the other with a custom-made titanium knife and plastic forceps. Twenty-three of the samples were from females, and 27 from males.After collection at Vest Agder County Hospital, the samples were transferred directly to small plastic bags (Local J. Environ. Monit., 1999, 1, 553±555 553 This journal is # The Royal Society of Chemistry 1999Producer) and deep-frozen at 218 �C. The samples were stored in this manner until analysed. Analytical method The samples were analysed at the Medical Centre of the nickel reÆnery.The analyst did not know the identity of the autopsied person and was not aware of which samples were collected with ordinary equipment and which were collected with the titanium knife. Special care was taken to avoid contamination of the samples, and all glass and plastic ware were washed and stored in 5% nitric acid and rinsed in ion exchanged water before use.The specimens were removed from the plastic bags using plastic forceps, and transferred to dried (105 �C), preweighed TeØon reagent tubes (Nalgene) equipped with loosely-Ætted plastic screw stoppers. The tubes were placed in a drying oven for 19 h. The tubes plus the dry samples were reweighed. Then 3 ml of a mixture (4z2z1) of HNO3, HClO4, and H2SO4, (Merck suprapur, Darmstadt, Germany) were added for digeston, and one drop of octanol (Merck) was added to prevent fuming.The tubes were stoppered lightly to allow for vapour to escape, and the samples were digested with increasing temperature. When the samples were completely dissolved, the mixture was transferred to 50 ml polyethylene volumetric Øasks equipped with stoppers and diluted to volume with 5% HNO3. From this solution, aliquots were withdrawn and transferred to 20 ml glass-stoppered Pyrex glass test tubes (custom made).To these aliquots, 0.3 ml of ammonium pyrrolidinedithiocarbamate (0.12 M; BDH-Merck, Poole, Dorset, UK) and one drop of m-Cresol Purple indicator (1 g l21) (Merck) were added. The solutions were shaken and adjusted to pH 9 with 13.6 M ammonia (Merck).A portion (2 ml) of methylisobutyl ketone (Merck; pro analysi) was added and nickel was extracted by shaking the samples for 2 min. After the phases had separated, 0.5 ml of the organic phase was transferred to 3 ml test tubes for analysis by electrothermal atomic absorption spectrometry (ETAAS) Perkin±er (Offenbach, Germany) model 603. Standard reference material for nickel in biological tissue was not available, and the standardisation was done by running of aqueous standard solutions through the whole procedure, including the digestion of the sample. The detection limit for the method is deÆned as three times the standard deviation of the blank, and calculated to be 5 ng g21 for a 300 mg sample; the precision of the method was calculated to be 12.2%.Subject details There were 27 men and 23 women in the cohort. The average age at death for the total cohort was 70.6 years, for females and males the average age of death was 74.6, and 67.8 years, respectively. Twenty-Æve people had lived within 8 km of the nickel reÆnery, and 23 between 8 and 100 km. Unfortunately we do not have smoking data for the complete cohort.Data analysis The statistical analyses include linear regression analysis of the nickel concentration in samples taken with the regular scalpel against those in samples taken with the titanium knife. Pairedsamples t test was employed to compare the means for the two sets of samples. The mean lung tissue nickel levels for people who had lived close to the reÆnery and for those who had lived further away were compared using the independent-samples t test.To meet the condition of normal distribution and comparable population variance, all statistical comparisons involved logarithms of the nickel concentration. Results The nickel concentrations detected in the lung tissue collected with a regular scalpel and in those obtained with a titanium knife with their corresponding conÆdence limits (CLs) are listed in Tables 1 and 2, respectively.Paired analysis of the variance showed a highly signiÆcant difference between the two sampling procedures (f~14.3; Pv0.0001). Tables 3 and 4 show the results for people who lived close to the reÆnery and for those who lived more than 8 km away. One-way analysis of variance of the results shows that people who lived more than 8 km away from the nickel reÆnery had signiÆcantly lower nickel concentrations in the lungs than people who lived closer to the reÆnery (F~6.3, Pv0.015).Linear regression analysis of the samples collected with the regular scalpel and the samples Table 1 Amount of nickel in lung tissue samples collected with a titanium knifea Case no.Ni/mg g21 Case no. Ni/mg g21 1 0.49 26 0.47 2 0.25 27 0.34 3 0.13 28 0.12 4 0.78 29 0.48 5 0.41 30 0.05 6 0.27 31 0.03 7 0.39 32 0.61 8 0.12 33 0.23 9 0.22 34 0.32 10 0.81 35 0.16 11 0.06 36 0.08 12 0.07 37 0.37 13 0.06 38 0.11 14 0.30 39 0.21 15 0.30 40 0.45 16 0.42 41 0.24 17 0.35 42 0.40 18 0.60 43 0.32 19 0.20 44 0.78 20 0.17 45 0.22 21 0.16 46 0.38 22 0.20 47 0.10 23 0.44 48 0.05 24 0.44 49 0.07 25 0.36 50 0.08 aArithmetric mean°s: 0.29°0.20 mg g21.Geometric mean: 0.20 mg g21 with 99% CLs of 0.18±0.30 mg g21. Table 2 Amount of nickel in lung tissue samples collected with a regular scalpela Case no. Ni/mg g21 Case no. Ni/mg g21 1 1.08 26 0.86 2 1.86 27 0.73 3 0.98 28 0.28 4 1.69 29 1.30 5 0.99 30 0.21 6 0.74 31 0.23 7 0.67 32 0.77 8 0.57 33 0.28 9 0.53 34 0.94 10 1.05 35 0.22 11 0.20 36 0.15 12 0.06 37 0.52 13 0.30 38 0.28 14 3.20 39 0.44 15 0.60 40 0.71 16 0.38 41 0.62 17 0.17 42 0.21 18 1.10 43 0.34 19 0.38 44 0.85 20 0.80 45 1.15 21 0.04 46 1.10 22 0.25 47 0.10 23 0.45 48 0.13 24 0.87 49 0.19 25 0.53 50 0.11 aArithmetic mean°s: 0.64°0.56 mg g21.Geometric mean: 0.45 mg g21 with 99% CLs of 0.32±0.64 mg g21. 554 J. Environ.Monit., 1999, 1, 553±555collected with the titanium knife show a grade of linearity (r~0.470, Pv0.001). Finally, one way analysis of the results showed no signi�¡cant difference between male (n~27) and female (n~23) lung tissue (P~2.0, P~0.165). Discussion The analyses of lung tissue samples collected with a titanium knife and plastic forceps show a highly signi�¡cant lower nickel concentration than do samples collected with a regular scalpel and forceps.The average concentration difference for 50 people used in this comparison is 0.35 mg 21 (0.6420.29). Since the samples were taken in situ and immediately transferred to plastic bags without any contact with other items like dishes or bowls, the difference in nickel content must come from the equipment used.The difference will most probably be dependent on the nickel content in the surgical equipment used. Nickel present as an alloy or as a contaminating agent could also be of importance. The average nickel content in �¡ve scalpels similar to those used in this investigation was 0.04%. In a similar pair of forceps the nickel content was 0.14%. In addition to the nickel content in the equipment, the size of the area the scalpel has cut could also be of importance.This could, in addition to contamination from other sources, which cannot be ruled out, explain the variation in the contamination level that was found. The importance of the results in this work depends of course on the use of the results. If the application is to uncover a possible small exposure to nickel, absence of contamination during collection could be of great importance and could give an investigation the necessary statistical power to disclose environmental exposure from re�¡neries or other industries in heavy polluted industrial areas.In earlier publications,4,11 we have reported so-called normal values for nickel which were somewhat higher than those found by other investigators; respectively 0.76 and 0.84 mg g21, which are in agreement with the results reported in this study for samples collected with the regular scalpel and forceps (0.64 mg g21).One major difference between the present investigation and our previous work is that deep-frozen tissue taken in situ were analysed rather than samples that were stored in a buffered formaldehyde solution before the analysis.Whether this has any in�ªuence on the results is not known. However, we have not been able to detect any release of nickel from tissue samples stored in formaldehyde solution even for samples collected from subjects who had been occupationally exposed to nickel. The increased concentration of nickel in lung tissue from people living closer to the re�¡nery could have been caused by their greater likelihood of having worked at the re�¡nery as contractors or as students.They could also have been exposed through welding jobs outside the re�¡nery, but as the concentrations in males and females were of equal magnitude none of these explanations seems likely. The possibility that women in this age group have been working as welders or in other occupations in the industry is very small indeed.Another possible explanation may be that people living closer to the re�¡nery have had family members working at the re�¡nery with subsequent domestic nickel pollution. This we cannot rule out. However, the most probable explanation for the increased lung nickel content in people who lived closer to the re�¡nery, is that nickel in the ambient air might have been higher, especially in the past.We have no data to con�¡rm this theory. Conclusion When analysing for nickel in lung tissue, nickel-free instruments should be used in the sampling process. This is of special importance when only marginal increases from the background concentration is expected. One should also be aware of contamination that is not directly connected to the working environment.Ideally `normal values' of nickel in the lung tissue should be obtained through sampling from a rural population far from nickel-releasing industries and urbanisation. References 1 IARC Monograph on the Evaluation of Carcinogenic Risks to Humans; Chromium Nickel and Welding, International Agency on Cancer, Lyon, 1990, vol. 49, pp. 318¡¾389. 2 International Committee on Nickel Carcinogenesis in Man, Scand. J. Work, Environ. Health, 1990, 16, 1. 3 H. Kollmeier, J. W. Seeman, K.-M. Mu¡í ller, G. Rothe, P. Wittig and V. B. Schejbal, Am. J. Ind. Med, 1987, 11, 659. 4 H. Kollmeier, J. W. Seeman, G. Rothe; ller and P. Wittig, Br. J. Ind. Med., 1990, 47, 682. 5 H. J. Raithel, K. H. Schaller, A.Reith, K. B. Svenes and H. Valentin, Int. Arch. Occup. Environ. Health, 1988, 60, 55. 6 H. J. Raithel, K. H. Schaller, L. A. Akslen, A. O. Myking, O. M¢�rkve and A. Gulsvik, Int. Arch. Occup. Environ. Health., 1989, 61, 507. 7 I. Andersen and K. B. Svenes, Int. Arch. Occup. Environ. Health, 1989, 61, 289. 8 K. B. Svenes and I. Andersen, Int. Arch. Occup. Environ. Health, 1998, 71, 424. 9 L. A. Akslen, A. O. Myking, O. M¢�rkve, A. Gulsvik, H. J. Raithel and K. H. Schaller, Pathol. Res. Pract., 1990, 186, 717. 10 J. Seeman, P. Wittig, H. Kollmeier, K. M. Mu¡í ller and V. Schejbal, Pathol. Res. Pract., 1990, 186, 197. 11 H. J. Raithel, G. Ebner, K. H. Schaller, B. Schellmann and H. Vallentin, Am. J. Ind. Med., 1987, 12, 55. Paper 9/05924E Table 3 Amount of nickel in lung tissue from people who lived 8 km or less from the re�¡nerya Case no. Ni/mg g21 Case no. Ni/mg g21 4 0.78 25 0.36 5 0.41 26 0.47 6 0.27 27 0.34 7 0.39 29 0.48 9 0.22 32 0.61 10 0.81 37 0.37 12 0.07 39 0.21 14 0.30 40 0.45 16 0.42 41 0.24 18 0.60 44 0.78 22 0.20 45 0.22 23 0.44 46 0.38 24 0.44 a Arithmetic mean¡Æs: 0.41¡Æ0.19 mg g21. Geometric mean: 0.37 mg g21 with 99% CLs of 0.27¡¾0.49 mg g21. Table 4 Amount of nickel in lung tissue from people who lived more than 8 km from the re�¡nerya Case no. Ni/mg g21 Case no. Ni/mg g21 1 0.49 33 0.23 2 0.25 34 0.32 3 0.13 35 0.16 8 0.12 36 0.08 11 0.06 38 0.11 13 0.06 42 0.40 15 0.30 43 0.32 17 0.35 47 0.10 19 0.20 48 0.05 21 0.16 49 0.07 28 0.12 50 0.08 30 0.05 aArithmetic mean¡Æs: 0.18¡Æ0.13 mg g21. Geometric mean: 0.15 mg g21 with 99% CLs of 0.10¡¾0.22 mg g21. J. Environ. Monit., 1999, 1, 553¡¾

ISSN:0960-7919    

DOI:10.1039/a905924e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Correlation of urinary nickel excretion with observed `total' and inhalable aerosol exposures of nickel reÆnery workers Mark A. Werner,a Yngvar Thomassen,b Siri Hetland,b Tor Norseth,b Steinar R. Bergec and James H. Vincentd aWisconsin Division of Public Health, 1414 E.Washington Avenue, Madison, WI 53703, USA bNational Institute of Occupational Health, P.O. Box 8149 DEP, N-0033 Oslo, Norway cFalconbridge Nikkelverk A/S, Postbox 457, Kristiansand S N-4601, Norway dDepartment of Environmental Health Sciences, School of Public Health, University of Michigan, 109 S.Observatory, Ann Arbor, MI 48108, USA Received 13th August 1999, Accepted 29th October 1999 An investigation of the relationship between observed nickel aerosol exposures and urinary nickel excretion was undertaken at a Scandinavian nickel reÆnery.The goal of the study was to assess the impact of nickel aerosol speciation, the use of particle size-selective sampling instrumentation and adjustment of urinary levels for creatinine excretion on the usefulness of urinary nickel excretion as a marker for exposure. Urinary nickel measurements and paired `total' and inhalable aerosol exposure measurements were collected each day for one week from reÆnery workers in four process areas.The mean observed urinary nickel concentration was 12 mg L21 (11 mg of Ni per g of creatinine). The strongest relationships between urinary excretion and aerosol exposure were found when urinary nickel levels were adjusted for creatinine excretion and when exposure to only soluble forms of nickel aerosol was considered.No signiÆcant difference was observed between measures of `total' and inhalable aerosol in the ability to predict urinary excretion patterns. In the light of these results, it is recommended that consideration be given to the chemical species distribution of nickel aerosol in the use of urinary nickel measurements as a screening tool for cancer risk in occupationally-exposed populations.Introduction It is widely recognized that the usefulness of occupational aerosol exposure measurements in evaluating workplace health risks is limited by their inability to account sufÆciently for the uptake, distribution and excretion of inhaled particulate matter, including both solid particles as well as liquid droplets, which deposit on the surfaces of the respiratory tract.This limitation often suggests that biological monitoring be conducted in order to more fully assess the health risks faced by exposed workers. An important source of guidance commonly used to aid in the interpretation of results from biological monitoring is the compilation of biological exposure indices (BEIs), published annually by the American Conference of Governmental Industrial Hygienists (ACGIH).1 ACGIH assigns BEIs in order to `...represent the levels of determinants which are most likely to be observed in specimens collected from a healthy worker who has been exposed to chemicals to the same extent as a worker with inhalation exposure to the (threshold limit value) TLV'. ACGIH has identiÆed nickel as a substance for which it seeks to establish a BEI.This is driven by the fact that there has been increasing toxicological and epidemiological evidence in recent decades supporting the role of nickel as a causative agent in lung and nasal cancer in humans. Such Ændings have led to technical efforts aimed at further reductions in exposure levels in the nickel production industry.Uncertainty about the cancer risk at present levels of exposure to airborne nickel, and the possible role of distinct nickel species in nickel toxicity, stimulated a major epidemiological inquiry in the 1980s into nickel-induced carcinogenesis under the auspices of the International Committee on Nickel Carcinogenesis in Man (ICNCM).2 This group examined the relationship between exposure to four nickel species groups and the occurrence of cancer. The species fractions in question were: (a) `sulÆdic' (including arsenides and tellurides); (b) `oxidic' (including silicates); (c) metallic; and (d) water-soluble.The study focused primarily on historical conditions in the nickel primary production and nickel alloy production industries, where the highest occupational exposures to nickel were thought to have occurred.Among its conclusions, ICNCM found it likely that more than one form of nickel gives rise to lung and nasal cancer, and that the risk of cancer is largely limited to (a) those individuals whose mean exposures to soluble forms of nickel aerosol exceeded 1 mg m23, and (b) those whose mean exposures to less soluble and insoluble forms of nickel aerosol exceeded 10 mg m23.The authors also noted, however, the relative scarcity of reliable occupational exposure data on which to base Ærm recommendations for quantitative occupational exposure limits. This paper discusses the results of a biological monitoring exercise carried out in parallel to an aerosol sampling campaign at a Scandinavian nickel reÆnery in September 1995.The goals of the study were threefold: Ærstly, to measure the urinary nickel concentrations of nickel reÆnery workers; secondly, to describe the distribution of airborne nickel among species groups; and thirdly, to characterize the relationship between urinary nickel concentrations and nickel aerosol exposures as measured using `total' and inhalable aerosol sampling instrumentation. The work is complementary to an earlier paper (Thomassen et al., 1999),3 but here the industrial processes were distinctly different, nickel exposure levels were much lower, and many more samples were speciated.It is expected that the results of such work will be useful in assessing the exposure-related cancer risk faced by workers in nickel production industries, and, more widely, in identifying a role for biological monitoring.Materials and methods The worksites The nickel reÆnery in question has about 650 employees and is engaged in the puriÆcation of nickel, copper and cobalt from a J. Environ. Monit., 1999, 1, 557±562 557 This journal is # The Royal Society of Chemistry 1999matte of mixed composition derived from the smelting of the original ore. Raw materials in the form of the pelletized matte are received by ship from nickel smelters in Canada and Botswana.This matte is ground into a Æne powder in the matte grinding area. Ground matte is transferred to the leach tanks in the chlorine leaching plant, along with recycled solution and chlorine gas from subsequent processes. During the leaching step, the bulk of the nickel in the feed is dissolved with copper.The copper is precipitated by adding matte, whereupon the slurry is Æltered to remove copper sulÆde from the pregnant solution. The pregnant nickel solution is then neutralized and oxidized with chlorine to precipitate iron and arsenic. The Æltrate is cooled, re-Æltered for gypsum removal, and pumped to the cobalt reÆning area for extraction with a 15% solution of tri-isooctylamine in an aromatic solvent to remove the cobalt.The rafÆnate from the cobalt extraction is diluted so that lead and other metals can be precipitated by treatment with nickel carbonate and chlorine. The pure nickel solution is then passed to the tank house for electrowinning, with chlorine being generated at the anode and recycled to the chlorine leaching process. Cobalt cathodes are produced from an all-chloride electrolyte in a tankhouse similar to that for nickel.The sulÆde residue is Æltered off after chlorine leaching and cementation, and is transferred to the roasting/smelting plant where it is repulped with water and slurry-fed to Øuidized bed roasters. Sulfur dioxide in the roaster off-gases is recovered as sulfuric acid.The calcine is leached in spent copper anolyte and copper is produced by electrowinning from the puriÆed solution. The residue from the copper leaching is treated further to extract precious metals. All plant workers are provided with respiratory protection equipment, which is required to be worn at all areas where atmospheric nickel exceeds the occupational exposure limit.Biological monitoring Urine samples were collected from workers at the four nickel reÆning processes outlined above, matte grinding, chlorine leaching, electrowinning and roasting/smelting, as part of a surveillance programme being conducted by the occupational health staff at the plant in cooperation with the Norwegian National Institute of Occupational Health.The 20 subjects were selected from a pool of workers who volunteered to participate in a separate aerosol exposure study carried out by other researchers. Each subject was followed for a Æve-day period, with ten subjects followed per week. Four subjects were selected from matte grinding areas, Æve from chlorine leaching, Æve from roasting/smelting and six from electrowinning. Subjects were asked to submit two urine samples per day for each day of the workweek, one in the morning and one in the afternoon.Morning samples were collected by workers at home before leaving for work and delivered directly to researchers at the plant upon arriving at the plant. Afternoon samples were taken at the plant immediately before workers left for the day and after removing their work clothes.Workers were directed to wash their hands before sampling to reduce the potential for contamination. Methods employed in this investigation for the collection and analysis of urine samples have been previously described.3 Urine samples were collected in disposable plastic cups from which sub-samples were poured into a screw-capped vial (Universal Container 25 ml, NUNC, Denmark).Control cups and vials were tested for nickel content to assure the absence of nickel contamination from these sources (i.e., less than the detection limit for the method employed of 0.5 mg L21). Samples were kept frozen at 220 �C prior to analysis. Thawed samples were heated at 95 �C for one hour to redissolve urine precipitates and to prevent the risk of laboratory-acquired infection.Urinary nickel concentration was measured by direct injection of undiluted urine without the use of a chemical modiÆer with electrothermal atomic absorption spectrometry employing a Zeeman-based Perkin Elmer Model 5100 PC/HGA-600 and a Perkin Elmer SIMAA 6000/THGA graphite atomizer calibrated with urine-matched standard solutions (calibration range 0±50 mg L21).The accuracy and precision of measurements were assessed continuously with human urine quality control materials produced by Sero Ltd. (Asker, Norway; Seronorm STE 101021 and 403125). Day-to-day variability of nickel measurements in reference materials was typically of the order of 10%. The average measured nickel concentrations of STE 101021 and 403125 were within °10% of the values given by the manufacturer.The creatinine content of urine samples was measured using a Beckman creatinine analyzer (which operated on the basis of the Jaffe reaction). Aerosol sampling Measurements were made of two indices of worker aerosol exposure as determined using two different personal aerosol samplers. The Ærst involved the measurement of inhalable aerosol, deÆned as the fraction of aerosol that is inhaled through the nose and/or mouth during breathing, reØecting the effect of particle size on what enters.This index is consistent with the criterion for nickel exposure as speciÆed by ACGIH in relation to its TLVs. This was achieved using a sampler developed speciÆcally for this purpose at the Institute of Occupational Medicine (IOM), Edinburgh, UK.4 This instrument, known as the `IOM personal inhalable aerosol sampler', is widely thought to be currently the best reference sampler for the inhalable fraction.The second approach involved the measurement of so-called `total' aerosol, the concept upon which aerosol exposure assessment and occupational exposure standards are largely based in most countries.This was performed using the 37 mm closed-face plastic cassette, which is widely used in the United States and in many other countries. Recent wind tunnel experiments have shown that whereas the IOM sampler provides a good measure of inhalable aerosol exposure, matching closely the particle size-selective inhalability criterion speciÆed by ACGIH, the 37 mm sampler undersamples signiÆcantly with respect to that fraction.5 For each volunteer, two personal aerosol samples were taken simultaneously, one for each sampler type mounted in the worker's breathing zone.Such samples were taken for a full workshift whenever possible, with all samples taken for at least four hours. All samplers were operated at a Øow rate of 2 L min21. At the conclusion of the sampling campaign, the samples were divided into two groups for analysis. The Ærst set, comprising 43 sample pairs, was assayed for 30 elements by inductively coupled plasma absorption emission spectrophotometry (ICP-AES). A smaller set of 20 sample pairs was submitted for the analysis of nickel species content by the method described by Zatka et al.6 Division of samples between the two groups was done so that the subset of samples for speciation was equally representative of the various processes at the facility and so that the speciation subset included samples from as many individual workers as possible. The Zatka method involves the sequential leaching of aerosol samples with three reagents, each of which extracts a distinct group of chemical species from the sample.In the laboratory, Ælters from the 37 mm cassette were laid out on support membranes upon which the sequence of reagents was applied.Filters from the IOM sampler were arranged similarly, with the addition of a cotton pad containing material wiped from the inside of the IOM sampling capsule from each sample (as is prescribed in the use of this instrument). The sample was leached Ærst with a 0.1 M ammonium citrate solution to extract the soluble nickel component from the sample.A second leach was conducted with a solution containing ammonium citrate and hydrogen peroxide to extract the `sulÆdic' group of nickel 558 J. Environ. Monit., 1999, 1, 557±562species. A solution containing methanol and bromine was added to the remaining sample, extracting the elemental nickel component.The residual material contained the `oxidic' group of nickel species. This residual material and the three leachate solutions were dried, treated with nitric and perchloric acids, and analyzed separately for nickel content by atomic absorption spectrophotometry. Data analysis Correlation and linear regression methods were employed to characterize the relationships between urinary nickel (NiU) measurements and aerosol sampling results.A common problem in biological monitoring is the observation of a high variability in urinary output volume from sample to sample. For this reason, urinary monitoring results are often expressed as the ratio of observed urinary nickel concentration (mg Ni per L urine) and creatinine concentration (g creatinine per L urine).This is referred to hereafter as the `adjusted NiU' concentration. Creatinine serves as an index of kidney function, and its use in the denominator in this context allows for urinary nickel concentrations to be expressed in a manner that is largely independent of the volume of urine excreted. Linear regression procedures were employed to describe the relationship between average NiU concentrations from the sampling week (adjusted and unadjusted) and the mean weekly nickel aerosol exposure levels as measured with both `total' and inhalable aerosol sampling methods, respectively.Estimates of soluble nickel exposure for all samples were obtained by multiplying nickel exposure estimates by a factor reØecting the prevalence of soluble nickel for `total' and inhalable aerosol in each of the four workplaces in the reÆnery.These were performed using the results of the speciation analyses of the 20 sample pairs set aside speciÆcally for this purpose. Estimates of insoluble nickel exposure were likewise calculated, using the sum of observed concentrations for three remaining species groups (sulÆdic, metallic and oxidic).In sum, estimates were obtained for exposures to elemental, soluble and insoluble nickel as measured with both `total' and inhalable sampling methods, yielding a total of six sets of aerosol exposure estimates. Regression procedures were Ærst carried out using each of the six measures of aerosol exposure individually as predictor variables. In addition, two multiple regression procedures were executed employing as predictor variables the soluble and overall insoluble exposure levels for each sampler type.It was expected from previous experience of ourselves and others that the distributions of both urinary nickel excretion values and nickel aerosol exposure measurements would be approximately lognormal. So the NiU and aerosol exposure measurements and estimates were log-transformed prior to analysis in order to normalize them.This enabled them to satisfy the assumptions required for such regression analysis procedures.7 This expectation of lognormality was subsequently conÆrmed upon analysis of the urinary excretion and aerosol exposure data. Results Nickel aerosol species measurements The results of the aerosol exposure assessment exercise are summarized in Table 1.They reØect a wide range of mean inhalable exposure values for the three less-soluble species across the range of plant processes, with mean exposures by process ranging over more than two orders of magnitude. The exception is soluble nickel, where observed mean exposurevalues differ by a factor no greater than four across the four process areas.This pattern is seen for both samplers. Fig. 1 displays the distribution of inhalable nickel exposures in reÆning operations among the `sulÆdic', `oxidic', metallic and water-soluble species groups. It is clear that marked differences exist in the distribution of airborne nickel among the species groups for the four process areas. For example, in three of the four process areas, a single species comprises more than 60% of all collected inhalable nickel aerosol, the `sulÆdic' group in matte grinding, the `oxidic' group in roasting and smelting, and the soluble component in electrowinning. In the chlorine leach plant, the `sulÆdic' group is the most prevalent, followed closely by soluble nickel.Table 1 Exposure to nickel species groups in nickel reÆning processes: mean exposure estimates from measurements of observed (a) inhalable and (b) `total' aerosol exposures to nickel species in nickel reÆning (species identiÆed by process) (a) Process Soluble/mg m23 `SulÆdic'/mg m23 Metallic/mg m23 `Oxidic'/mg m23 Matte grinding 0.0348 0.3914 0.0339 0.0182 Chlorine leaching 0.0304 0.0367 0.0028 0.0047 Roasting/smelting 0.0114 0.0094 0.0004 0.1010 Electrowinning 0.0531 0.0013 0.0001 0.0054 (b) Process Soluble/mg m23 `SulÆdic'/mg m23 Metallic/mg m23 `Oxidic'/mg m23 Matte grinding 0.0286 0.2022 0.0374 0.0104 Chlorine leaching 0.0133 0.0207 0.0031 0.0030 Roasting/smelting 0.0081 0.0049 0.0008 0.0826 Electrowinning 0.0176 0.0010 0.0012 0.0070 Fig. 1 Distribution of nickel species among reÆning processes: prevalence of four nickel species groups in workplace aerosol among nickel reÆnery process areas.The percentage values shown are from measurements of inhalable nickel (with the IOM personal inhalable aerosol sampler). J. Environ. Monit., 1999, 1, 557±562 559Measurements of urinary nickel concentrations A total of 214 urine samples were collected and analyzed for nickel and creatinine content. This number includes 79 pairs which represent morning and afternoon samples from workers for whom `total' and/or inhalable aerosol sampling results are also available.Table 2 summarizes the results of the urinary nickel concentration measurements. Individual mean weekly NiU concentrations were in the range from 2.7 to 52 mg L21 of urine (unadjusted) and from 1.8 to 44 mg per mmol of creatinine (adjusted).The results indicate that observed urinary concentrations were markedly lower for the roasting/smelting workers than for workers in the remaining three process areas. The observed standard deviation values show considerably greater variability in urinary nickel excretion among matte grinding workers than for workers in other process areas.This may be attributed to the presence in the matte grinding cohort of a single worker, represented by the uppermost point in Figs. 2 and 3, whose urinary nickel output was markedly elevated. While the composite mean of all afternoon samples was higher than the corresponding mean for morning samples, no signiÆcant trends were observed between concentrations measured in the morning versus afternoon measurements.Similarly, no trend in urinary nickel excretion was observed across the work week. Relationship between urinary and air measurements Figs. 2 and 3 display the data for mean weekly NiU concentration as a function of the mean weekly nickel exposures of the participating workers. Table 3 shows the results of linear regression analyses in greater detail, where the strengths of the assumed linear associations are now best represented by R2.These results show that, for both unadjusted and adjusted NiU and for regression procedures employing a single aerosol exposure variable, R2 values are appreciably larger when soluble nickel aerosol exposure is employed as the independent variable. Further, in those cases estimated R2 values are greater for the NiU concentrations adjusted for creatinine content.In the multiple regression analyses, the addition of insoluble aerosol exposure as a predictor variable resulted in a negligible improvement in the observed relationship, with R2-values increasing by less than 0.05. In all cases, coefÆcients for insoluble exposure were found not to be signiÆcantly different from zero.Discussion Previous studies have sought to describe the relationship between observed `total' and inhalable aerosol exposures during primary metals production and to examine the impact Fig. 2 Nickel aerosol exposures and raw urinary nickel concentrations: plot of unadjusted urinary nickel concentrations versus (a) `total' and (b) inhalable measures of soluble nickel aerosol exposure.Lines representing the results from regression analysis are also displayed (see legend). Table 2 Urinary nickel excretion in nickel reÆning processes: summary of mean weekly (a) unadjusted and (b) adjusted urinary nickel concentrations observed in nickel reÆnery workers (a) Mean (range) Standard deviation Median Process Number of workers Number of worker-daysa (mg Ni per L of urine) (mg Ni per L of urine) (mg Ni per L of urine) Matte grinding 4 16 18.3 (3.9±51.5) 22.4 8.9 Chlorine leaching 5 18 13.4 (4.9±21.3) 8.0 15.0 Roasting/smelting 5 19 4.8 (2.7±7.9) 2.2 3.7 Electrowinning 6 25 9.7 (7.9±11.7) 1.4 9.2 Plant-wide 20 78 11.1 (2.7±51.5) 10.8 8.5 (b) Mean(range) Standard deviation Median Process Number of workers Number of worker-daysa (mg Ni per L of urine) (mg Ni per L of urine) (mg Ni per L of urine) Matte grinding 4 16 17.1 (3.8±44.5) 18.7 10.2 Chlorine leaching 5 18 10.9 (2.9±22.1) 8.0 10.4 Roasting/smelting 5 19 3.1 (1.8±4.6) 1.1 4.4 Electrowinning 6 25 9.7 (5.4±17.5) 4.4 8.0 Plant-wide 20 78 9.8 (1.8±44.5) 9.8 7.2 aTwo urine samples collected per day. 560 J. Environ. Monit., 1999, 1, 557±562of the differences on occupational exposures and standards.8±10 The principal Ænding was that inhalable aerosol exposures were roughly twice as great as the corresponding `total' aerosol exposures.A summary of the corresponding intersampler comparison results from the current data set at the nickel reÆnery studied, for nickel and several other metallic elements, are shown in Table 4.11 They are broadly consistent with trends observed in the earlier studies.This paper represents results from the Ærst attempt to correlate urinary nickel excretion with nickel aerosol exposures as measured with sampling instrumentation corresponding to the ACGIH particle size-selective criterion for inhalable aerosol [also widely adopted by other bodies, including the International Standards Organisation (ISO) and the Comite� Europe�en Normalisation (CEN)].In this study, correlation coefÆcients for the relationship between urinary nickel excretion and aerosol exposure levels were quite similar for the two sampling methods (see Table 3). This suggests that there is no substantial difference in the ability to accurately predict urinary nickel levels on the basis of either `total' or inhalable aerosol sampling results.It should be kept in mind, however, that the similarity of these correlation coefÆcients should not be interpreted as suggesting that the two sampling methods are equally suited to the assessment of exposure-related health risks for nickel. Nickel has been shown to be a likely nasal carcinogen and, as such, exposures to nickel-containing aerosol should be assessed on a basis which accounts for all airborne particulate matter which may enter the respiratory tract, including the nose.This, therefore, is part of the rationale behind the adoption of the inhalable fraction as the criterion for nickel aerosol exposure. Furthermore, a number of insoluble nickel species have been implicated as likely carcinogens. However, the results presented in this paper suggest that observed urinary nickel levels are not highly responsive to insoluble nickel.It follows that NiU should not be regarded as a reliable marker for exposure to insoluble forms of nickel. Conversely, however, strong correlations were found between observed exposures to soluble nickel aerosol and NiU concentrations. This is, of course, consistent with expectations based on the relatively high efÆciency with which water-soluble materials may pass from the respiratory tract into the blood and thence into the urine, as has been found by other researchers for occupational situations where appreciable amounts of soluble nickel aerosol were likely to have been present.12–14 Two important factors to consider in interpreting these results are the use of respiratory protection by plant workers and non-occupational exposure to nickel.In the plant, the use of respiratory protection was required, and observed, in areas where nickel aerosol exposures were above applicable occupational exposure limits. So it is likely that there may have been some interferences in the relationships studied. More speciÆcally, it is likely that, for a given measured nickel aerosol exposure level, a more signiÆcant contribution to urinary nickel excretion would be observed for workers in areas with low observed dust exposures, such as electrowinning, than for workers in areas with high dust exposures, such as matte grinding.Differences in the prevalence of use of respiratory protection across work areas may, therefore, have a signiÆcant impact on urinary excretion, and must be considered in drawing conclusions on the basis of this research.In addition, nickel is a natural constituent of many foods and of tobacco. These may represent important non-occupational sources of nickel that may inØuence observed urinary nickel excretion patterns. The Ændings in this study are in interesting contrast to the Ændings of Culver et al.15 in their study of the relationship between observed `total' and inhalable boron exposures and levels of boron in urine and blood for workers in the borate production industry.In that study, regression of blood and urine boron levels on observed inhalable aerosol exposures yielded substantially higher R2 values (0.85 for urine, 0.77 for Fig. 3 Nickel aerosol exposures and creatinine-adjusted urinary nickel concentration: plot of adjusted urinary nickel concentrations versus (a) `total' and (b) inhalable measures of soluble nickel aerosol exposure. Lines representing results from regression analysis are also displayed (see key). Table 3 Relationship between nickel aerosol exposure and urinary nickel excretion: results of regression analysis for (a) unadjusted and (b) adjusted NiU concentrations and various measures of nickel aerosol exposure according to the model ln[NiU]~K6ln[NiA]zb, in which [NiU] represents the urinary nickel concentration and [NiA] the nickel aerosol concentration.The K values in parentheses are for those regression procedures where K was found not to be signiÆcantly different from zero (a~0.05) (a) Sampler Analyte K b P: Kw0 R2 `Total' Elemental Ni (0.108) 2.49 0.35 0.05 Inhalable Elemental Ni (0.104) 2.42 0.39 0.04 `Total' Soluble Ni 0.422 4.11 0.006 0.35 Inhalable Soluble Ni 0.468 4.07 0.005 0.36 `Total' Insoluble Ni (0.024) 2.23 0.76 0.005 Inhalable Insoluble Ni (0.019) 2.20 0.82 0.003 (b) Sampler Analyte K b P: Kw0 R2 `Total' Elemental Ni (0.185) 2.55 0.17 0.10 Inhalable Elemental Ni (0.168) 2.40 0.23 0.08 `Total' Soluble Ni 0.633 4.91 v0.001 0.56 Inhalable Soluble Ni 0.677 4.73 v0.001 0.54 `Total' Insoluble Ni (0.043) 2.11 0.65 0.01 Inhalable Insoluble Ni (0.030) 2.04 0.75 0.006 J.Environ. Monit., 1999, 1, 557±562 561blood) than those from regression of the same biological indices on corresponding observed `total' aerosol exposures (0.49 for both urine and blood).This contrast is likely to be attributable to the relative complexity of the toxicokinetics of nickel in comparison with boron, which is excreted from the body largely unmetabolized. The results from the speciation of nickel aerosol in the workers' exposures show the presence of a wide variation in the prevalence of individual nickel species groups in the various nickel re�¡ning processes of this plant.Given the observation that only soluble nickel aerosol was found to signi�¡cantly affect observed urinary nickel levels, the results suggest that the measurement of individual nickel species groups in aerosol samples may provide useful information in the interpretation of urinary nickel measurements.The observation that more than one nickel species group may be abundantly present in an industrial process has been noted elsewhere.16,17 Taken as a whole, these �¡ndings suggest that the characterization of the exposures of nickel industry workers can be greatly enhanced by the consideration of the species distribution of nickel in workplace aerosol. With this in mind, we are currently conducting research to `�¡ngerprint' nickel production industry workplaces in this way.An interesting methodological issue is raised in the comparison between the exposures to airborne nickel and NiU, as re�ªected in the two separate indices, that is, `raw' urinary nickel concentration and values adjusted with respect to urinary creatinine concentrations, respectively.A recent review of urinary nickel concentrations for occupationally exposed populations has shown that urinary nickel concentrations have been commonly reported as either raw concentrations or as creatinine-adjusted levels,18 that is, one or the other. In our study, as shown in Table 3, adjustment of urinary nickel measurements for creatinine resulted in correlation coef�¡cients and R2 values that were consistently higher. This suggests that, where available, the use of urinary creatinine concentrations as a standardization tool may reduce the amount of random variability in urinary nickel measurements and allow for urinary excretion to be better explained on the basis of measurements of nickel aerosol exposure.Conclusions This work was stimulated primarily by the recognition of the need to characterize workplace nickel aerosol exposures in terms of the inhalable fraction.The �¡ndings, however, are particularly interesting in the light of the recent proposal by the ACGIH to express its threshold limit value (TLV) for nickel in terms of a number of speciation groups, as well as continuing interest in identifying the degree to which biological levels of nickel may be indicative of exposure-related health risk.An important consideration in light of these results is the role that ing might play in the assessment of health risks posed by nickel exposures in the primary nickel production industries. Given that a primary health risk of concern for nickel-exposed workers is the development of cancers of the respiratory tract, presumably near sites at which nickel-containing particulate is deposited, it appears that assessments of workplace exposure to inhalable aerosol are likely to better re�ªect health risk for nickel than consideration of nickel levels in urine or plasma.This should be considered in future deliberations, by ACGIH or other bodies, about the development of reference urinary concentrations for nickel.It should also emphasize the need for caution in using the results of biological monitoring to make inferences about the health status of nickel industry workers. Acknowledgements The authors wish to thank the Nickel Producers Environmental Research Association (NiPERA), the National Institute for Occupational Health (NIOH) in Oslo, Norway, and the Torske Klubben of Minneapolis¡¾St.Paul for their support of this research. We also wish to thank Ole Synnes and Kari Dahl of NIOH for their valuable �¡eld and analytical assistance. Finally, we wish to thank the plant management and workers who agreed to participate in the research. References 1 American Conference of Governmental Industrial Hygienists (ACGIH), Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices, ACGIH, Cincinnati, OH, 1998. 2 R. M. Doll, A. A. Andersen, W. C. Cooper, I. Cosmatos, D. L. Cragle and D. Easton and 16 others, Scand. J. Work Environ. Health, 1990, 16 (Suppl. 1), 1. 3 Y. Thomassen, E. Nieboer, D. Ellingson, S. Hetland, T. Norseth, J. E. Odland, N. Romanova, S. Chernova and V. P.Tchachtchine, J. Environ. Monit., 1999, 1, 15. 4 D. Mark and J. H. Vincent, Ann. Occup. Hyg., 1986, 30, 89. 5 L. C. Kenny, R. J. Aitken, C. Chalmers, J. F. Fabries, E. Gonzalez- Fernandez, H. Kromhout, G. Liden, D. Mark, G. Riediger and V. Prodi, Ann. Occup. Hyg., 1997, 41, 135. 6 V. J. Zatka, J. S. Warner and D. Maskery, Environ. Sci. Technol., 1992, 26, 138. 7 D. G. Kleinbaum, K.L. Kupper and K. E. Muller, Applied Regression Analysis and Other Multivariate Methods, Duxbury Press, Belmont, CA, USA, 2nd edn., 1988. 8 P. J. Tsai, J. H. Vincent, G. Wahl and G. Maldonado, Occup. Environ. Med., 1995, 52, 793. 9 M. A. Werner, T. M. Spear and J. H. Vincent, Analyst, 1996, 121, 1207. 10 T. M. Spear, M. A. Werner, J. M. Bootland, A. Harbour, E. P. Murray, R.Rossi and J. H. Vincent, Am. Ind. Hyg. Assoc. J., 1997, 58, 893. 11 M. A. Werner, J. H. Vincent, Y. Thomassen, S. Hetland and S. Berge, Occup. Hyg., 1999, in the press. 12 A. C. H¢�getveit, R. T. Barton and C. O. Kost¢�l, Ann. Occup. Hyg., 1978, 21, 113. 13 W. Torjussen and I. Andersen, Ann. Clin. Lab. Sci., 1978, 8, 184. 14 L. Ulrich, M. Sulcova¡� , L. Spacek, E. Neumanova¡� and M. Vlada¡�r, Sci. Total Environ., 1991, 101, 91. 15 B. D. Culver, P. T. Shen, T. H. Taylor, A. Lee-Feldstein, H. Anton-Culver and P. Strong, Environ. Health Persp., 1994, 102 (Suppl 7), 133. 16 M. Kiilunen, J. Utela, T. Rantanen, H. Norppa, A. Tossavainen, M. Koponen, H. Paakkulainen and A. Aitio, Ann. Occup. Hyg., 1997, 41, 167. 17 J. H. Vincent, P. J. Tsai and J. S. Warner, Analyst, 1995, 120, 675. 18 M. Kiilunen, Occupational Exposure to Chromium and Nickel in Finland and Its Estimation by Biological Monitoring, Doctoral Dissertation, Kuopio University, Kuopio, Finland, 1994. Paper 9/6597K Table 4 Intersampler ratios for inhalable and 'total' nickel aerosol exposures in nickel re�¡ning: summary of results from analysis of intersampler comparison data from nickel re�¡ning as reported by Werner et al.11 The S-values listed are coef�¡cients from the model EIOM~S6E37, where EIOM represents inhalable nickel aerosol exposure (as measured using the IOM personal inhalable aerosol sampler) and E37 represents `total' nickel aerosol exposure (as measured using the 37 mm sampler in closed-face mode). Results in italics are for analyses with designated outliers included in the dataset; LCL is the lower bound of the 95% con�¡dence interval and UCL is the upper bound of 95% con�¡dence interval Worksite S Standard error R2 n LCL 95% UCL 95% Matte grinding 1.80 0.29 0.79 11 1.14 2.46 2.29 0.56 0.61 12 1.04 3.54 Chlorine leaching 1.65 0.13 0.93 14 1.37 1.93 Roasting/smelting 2.29 0.33 0.82 12 1.55 3.03 Electrowinning 1.54 0.12 0.94 13 1.28 1.80 Plant-wide 1.81 0.12 0.83 50 1.57 2.05 1.92 0.16 0.74 51 1.60 2.24 562 J. Environ. Monit., 1999,

ISSN:0960-7919    

DOI:10.1039/a906597k

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Cement for stabilisation of industrial residues containing heavy metals Begonƒa Vallejo,a Riansares Munƒoz,b Andre�s Izquierdoa and M. Dolores Luque de Castro*c aDepartment of R&D, Gemasur, Polý�gono de las Quemadas, Parcela 271, Co�rdoba, Spain bDepartment of Analytical Chemistry, Faculty of Chemistry Sciences, University Complutense of Madrid, E-28040 Madrid, Spain cDepartment of Analytical Chemistry, Faculty of Sciences, University of Co�rdoba, E-14004 Co�rdoba, Spain.E-mail: qa1lucam@uco.es; Fax: 957 218606 Received 7th May 1999, Accepted 17th September 1999 A method aimed at decreasing the toxicity of heavy metals [namely, Zn(II) and Cr(III)] in real polluted residues by immobilisation has been developed. The residues were processed either with two cement-type stabilisers or lime.The cement-type stabilisers were Portland cement and Depocrete SM/2 at the self-generated pH (ca. 11) which afforded physical as well as chemical potential for the immobilisation of heavy metals. The other stabiliser, lime, reduced organic compounds, thus favouring the decrease of the chemical oxygen demand (COD) and endowing the residue with better mechanical properties for transport.After leaching the stabilised residues using the standard leaching test [Order 13/10/89, Boletý�n OÆcial del Estado (BOE) 270 10/11/89], three ways for establishing the toxicity of the treated residues were used, namely: (1) the ecotoxicity test using Photobacterium phosphoreum (DIN 38 412); (2) determination of the concentration of heavy metals by atomic absorption spectrometry (AAS); (3) determination of the COD or oxygen required for complete chemical oxidation of a water sample.Portland cement (20%) blended with Depocrete SM/2 (3%) acted as an effective stabiliser for residues containing heavy metals as it increased the ecotoxicity index (EC50) by more than Æve times. Thus the heavy metal concentration in the leaching liquid was lowered to less than 0.1 mg l21.The addition of 5% of lime afforded a residue easily transportable from the place of treatment to the landÆll. The precision of the method was studied in terms of both repeatability and reproducibility. The values found with respect to EC50 and expressed as the relative standard deviation (RSD) were 1.6% and 5.1%, respectively. Introduction Contamination can be deÆned as the increase of the content of any substance in a natural medium with a negative effect due to the impact of human activities.Residues with high concentrations of heavy metals are hazardous. The environmental problem of residues containing heavy metals produced by human activity is of increasing concern.1±5 This is due to the greater understanding of the toxicological importance of these residues on the ecosystem.Therefore, they need special treatment prior to deposition in landÆlls to ensure that leaching liquids will not be toxic when subjected to atmospheric conditions, because chemicals could percolate and leach through the soil to the groundwater. Chromium and zinc are common residues in industrial zones. Chromium occurs in two stable oxidation states,6,7 Cr(III) and Cr(VI).The latter is both the most toxic and difÆcult to Æx into solids, but Cr(VI) is easily reduced to Cr(III) by metals (zinc, iron, aluminium, etc.) or by salts of ferrous iron, sulÆtes, sulÆdes, etc. Alkaline cements, e.g. Portland cement and Depocrete SM/2, are utilised as matrices in order to immobilise heavy metals.These materials are characterised by both a high pH of the matrices prepared from them (pH 11) and a low solubility of their hydration products [e.g. semicrystalline calcium hydrosilicates (CSH)].8 Two types of immobilisation of heavy metals in cement matrices have been distinguished by Kindness et al.:9 physical immobilisation, consisting of the adsorption of ions and particles on the microporous surface of the CSH phase, and chemical immobilisation, which takes place as a result of the formation and precipitation in the matrix of compounds with low solubility products.Each chemical element has a unique chemistry in its interaction with cement, so each must be studied individually. Cement affords a highly alkaline environment (pH 11) which is responsible for the low solubility of heavy metals.The alkaline pH is of paramount importance for stabilisation, particularly for Cd, Co, Mn, Ni, Zn and Cr, due both to adsorption onto oxide speciÆc sites and precipitation of hydroxides or carbonates. Previous studies on the immobilisation of heavy metals in cement matrices8,9 have been developed with success using laboratory-prepared soil samples.In this study, we show the decrease in toxicity, heavy metal concentration and chemical oxygen demand (COD)10,11 of natural residues with Zn(II) or Cr(III) produced by human activity when they are cured with cements and lime. Spanish legislation12 establishes the following method for the characterisation of toxic residues regulated by Order 13/10/89, Boletý�n OÆcial del Estado (BOE) 270 10/11/89: leaching step for 24 h with pH control of the leaching solution, followed by a bioassay based on luminescence using Photobacterium phosphoreum 13 (DIN 38 412).According to this method, a residue is considered as hazardous if the leaching solution shows an EC50 (50% reduction of the luminescence intensity of a sample of bacteria which initially has a luminescence of EC100) equal or lower than 3000 mg l21.The Spanish `Reglamento del Dominio Pu� blico Hidra� ulico' (RDPH)14 established that the concentration of Zn(II), Cr(III) and COD must not be higher than 3, 2 and 160 mg l21, respectively, in leaching liquids to ensure that these solutions are non-hazardous. In this work, the concentrations of heavy metals [Zn(II) and Cr(III)] in the leaching solution are determined by Øame atomic absorption spectrometry (FAAS) and graphite furnace atomic absorption spectrometry J.Environ. Monit., 1999, 1, 563±568 563 This journal is # The Royal Society of Chemistry 1999(GF-AAS), respectively, and a Øow injection (FI) system is used for the COD study. Experimental Reagents and solutions All reagents were of analytical-reagent grade and distilled water was used throughout.Leaching step. A 0.5 mol l21 acetic acid solution (HOAc) (Merck, Darmstadt, Germany) was used. Determination step. Photobacterium phosphoreum bacterium LCK 480 and bacterium reactivation solution (both supplied by Neurtek, Zarautz, Spain), solid NaCl (Merck) and a 2% aqueous NaCl solution of pH 7.0°0.2 were utilised. NaOH (Panreac, Barcelona, Spain) and HCl (Merck) solutions (0.05, 0.5 and 5 mol l21, respectively) were prepared for adjusting the pH of the leaching solutions.Stabilisation step. Portland cement, Depocrete SM/2 (both supplied by Tracoisa, Madrid, Spain) and lime (Cal-Gov, Sevilla, Spain) were used as stabilisation agents. Chemical oxygen demand. Sodium oxalate (Panreac) was used as the COD standard substance by drying it at 110±120 �C for ca. 1 h and using an accurately weighed 8.375 g portion of the dried product dissolved in 100 ml of Milli-Q puriÆed water in order to prepare an 83.75 g l21 stock oxalate solution (or 10 g COD l21). All standard solutions were prepared by appropriate dilution of the stock solution before use. The carrier stream was a 0.6 mol l21 H2SO4 solution (Panreac), which was prepared by dilution of the concentrated acid in Milli-Q puriÆed water.The oxidising solution (KMnO4) was prepared by dissolving 1.6 g of the solid product (Merck) in 1000 ml of 0.6 mol l21 sulfuric acid (Panreac). The solution was heated for ca. 1 h at 80±90 �C. After cooling at room temperature, the solution was Æltered through a glass Ælter, and standardised by titration with standard sodium oxalate solution.Then, the solution was stored in an amber-glass bottle in the dark at 4 �C. An 861024 mol l21 KMnO4 solution was prepared by dilution of the standardised solution with 0.6 mol l21 H2SO4. The certiÆed reference material (CRM) QC PLUSz Demand Quality Control Standard (Dicoeed in order to validate the FI method.A solution of 260 mg COD l21 was prepared by appropriate dilution (1 : 200) of the CRM with 0.6 mol l21 H2SO4. Atomic absorption spectrometry. A 1000 mg l21 stock Zn(II) or Cr(III) solution (Sigma, Madrid, Spain) was used for running the calibration curve. All standard solutions of Zn(II) or Cr(III) were prepared before use by appropriate dilution of the stock solutions in Milli-Q puriÆed water.Instruments and apparatus Ecotoxicity assay. A Vibromatic vibrator (Selecta, Barcelona, Spain), a PJ400 balance (Mettler, Switzerland), a Crison micropH 2001 (pHmeter, Copenhagen, Denmark), a Biohit (range 200±1000 ml) micropipette (Genesys, Granada, Spain) and a Shimadzu F-1501 spectroØuorimeter (Isaza, Sevilla, Spain) with 1 cm glass cells (Genesys) were used.Chemical oxygen demand. A four-channel Gilson (Villiers le Bel, France) Minipuls-3 peristaltic pump Ætted with a rate selector, a Rheodyne model 5041 injection valve (Elkay, Galway, Ireland) and poly(tetraØuoroethylene) (PTFE) tubing of 0.8 mm id (Scharlau, Barcelona, Spain) were used. A Selecta water bath (Barcelona, Spain), equipped with a thermostat, and a Philips PU 8625 UV/VIS spectrophotometer (Cambridge, UK), equipped with a Hellma (Jamaica, NY, USA) 138-QS Øow cell (18 ml inner volume) and a Knauer recorder (Scharlau, Barcelona, Spain), were also used.Atomic absorption spectrometry. A Perkin Elmer 1100 B graphite furnace (Norwalk, CT, USA), with a hollow-cathode Cr lamp, and a Perkin Elmer 2380 spectrometer with air± acetylene Øame, equipped with a hollow-cathode Zn lamp, were used.Procedures Sample treatment. Before stabilisation, about 25 g of homogenised sample (only residue with heavy metal) or, during stabilisation, 25 g of stabilised sample (residue plus a few per cent of stabiliser) was weighed in a 1000 ml Erlenmeyer Øask, mixed with 16 g of distilled water per gram of sample and placed on the vibrator.The suspension was stirred while the pH was adjusted to 5.0°0.2 by adding 0.5 mol l21 HOAc at intervals of 15, 30 and 60 min, passing to the following interval when the pH adjustment was unnecessary (pH equal to or lower than 5.2 pH units). This procedure was repeated for at least 6 h. The Spanish legislation12 establishes that the volume of HOAc added during the leaching step must never exceed 4 ml per gram of solid sample.Leaching was completed after 24 h and the amount of distilled water given by the equation below was added to the Erlenmeyer Øask and the solution was then stirred for 5 min: V~20W{16W{A where V is the volume of distilled water to be added (ml), W is the mass of sample weighed into the Erlenmeyer Øask (g) and A is the volume of 0.5 mol l21 HOAc added during the leaching step (ml).Then, the leached sample was Æltered in order to obtain three 50 ml aliquots: one was utilised in the toxicity assay, one for the determination of heavy metals in the sample by FAAS and GF-AAS and one for the COD study. Ecotoxicity assay. The luminescent bacteria test13 applied to the leaching solution was as follow. (1) 2% NaCl solution was prepared and adjusted to pH 7.0°0.2.(2) The salt content of the sample was adjusted to 2% and the pH to 7.0°0.2. (3) One vial of reactivation solution (RS) was thawed at room temperature in a water bath and mixed thoroughly afterwards. The temperature was adjusted to 15 �C for about 15 min. (4) A sample dilution series was prepared in measuring cuvettes in the sequence 1 : 0, 1 : 1.5, 1 : 2, 1 : 3, 1 : 4, 1 : 6, 1 : 8, 1 : 12, 1 : 16, 1 : 24.The leaching solution was diluted with 2% NaCl. (5) One vial of bacteria was thawed quickly (2 min) in a water bath at room temperature. (6) 0.5 ml of RS was added to the vial of bacteria and mixed by gently shaking until it was fully dissolved and then incubated for 15 min at 15 �C. (7) The content of the vial of bacteria was transferred totally into the remaining RS and the suspension was mixed to homogeneity.(8) 0.5 ml of the suspension was pipetted into each of the 10 measuring cuvettes. (9) Then, the initial luminescence (EC100) of the test batches was measured at preset intervals and 0.5 ml of the corresponding sample dilution was added. (10) After the exact incubation time (15 min, according to the Spanish legislation), the luminescence (EC) of the test batches was measured following the same order and intervals as in (9).The plot of log [(EC100,SECB2ECSEC100,B)/ECSEC100,B] vs. log [sample (%)] for each dilution was a straight line, where EC100 was the initial luminescence of the sample dilution (S) or the blank (B) and EC was the luminescence after 15 min incubation.Its interception point with the x-axis showed the percentage of leaching sample which provided a luminescence of EC50 (Fig. 1). This percentage must be multiplied by a factor 564 J. Environ. Monit., 1999, 1, 563±568F of 500 [percentage (g sample/100 ml solution)6(25 g leached sample/500 ml Ænal volume of leaching liquid)6(103 mg g21)6 (103 ml l21)] to obtain the concentration of toxics in the sample, expressed in mg l21.Chemical oxygen demand. A four-channel Gilson Minipuls- 3 peristaltic pump pumped the 861024 mol l21 acidic potassium permanganate solution and the 0.6 mol l21 sulfuric acid separately at Øow rates of 0.5 ml min21. A volume of 0.5 ml of sample or standard was injected through the injection valve into the 0.6 mol l21 sulfuric acid and merged with the potassium permanganate stream. The product was led to a coiled PTFE reactor, which was heated at 50 �C in a thermostated bath.The reaction mixture then passed through the Øow cell of the spectrophotometer, where the fading of colour of the permanganate solution was measured at 525 nm and the peak absorbance interpolated in a calibration graph run with oxalate solutions in the range 0.8±49.6 mg oxalate l21 (or 0.1±5.9 mg COD l21).Determination of heavy metals. The concentration of Zn(II) in the leaching solution was determined by FAAS with an air± acetylene Øame and a hollow-cathode Zn lamp (l~307.6 nm; slit~0.5 nm). The concentration of Cr(III) was studied by GF-AAS with a hollow-cathode Cr lamp (l~357.9 nm; slit~0.7 nm). The temperature of the graphite furnace was held isothermally at 2400 �C for 3 s, and then was increased to 2500 �C with a ramp time of 1 s and 2 s of hold time.The temperature programme established is shown in Table 1. The signals given by FAAS and GF-AAS were extrapolated in a calibration graph run with Zn(II) solutions in the range 0±1 mg l21 and with Cr(III) solutions in the range 0±60 mg l21, respectively.Stabilisation procedure. Portland cement and Depocrete SM/2 (a type of cement which cures very rapidly) were used to formulate matrices which stabilised heavy metals. Different percentages of these cements (3%, 10%, 20% and 30%) and the residue with heavy metals [Zn(II) or Cr(III)] were mixed until total homogenisation was achieved. The samples were prepared using distilled water during homogenisation with a 3 : 1 water volume to weight of stabiliser ratio and were cured for 24±70 h at 36 �C.After curing, the sample treatment was applied to 25 g of sample (residue plus stabilisation agent) and the three assays were applied to the leaching liquids. Results and discussion A great variety of residues produced by human activity are treated by the residue management company GEMASUR.In this work, two solid samples from physical±chemical treatment plants, one containing 3.4% of Zn(II) and the other containing 1.0% of Cr(III), were utilised as natural samples to study the immobilisation of heavy metals in cement matrices in order to determine the correlation with the increase in the ecotoxicity index (EC50).The samples containing high concentrations of heavy metals [Zn(II) or Cr(III)] were subjected to the leaching procedure and subsequent ecotoxicity assay, determination of heavy metal concentration and COD as explained in the `Experimental' section. The results obtained are shown in Table 2. The method used for the determination of eaching solution was simple and rapid, with a detection limit of 80 mg l21, relative standard deviation of 1.3% (calculated at 4.7 mg COD l21), a sampling rate of 18 samples h21 and excellent tolerance to interferences (chloride ion does not interfere at 8000 mg l21).The method has been exhaustively validated as follows: (1) validation using the CRM QC PLUSz Demand Quality Control Standard; (2) application to leaching liquids also analysed by an authorised laboratory (Hidrocen, Madrid, Spain) using the standard method.15 The COD of the CRM found by the FI method16 was statistically indistinguishable from the certiÆcate value.An interlaboratory exercise was realised with Æve samples16 obtained from physical±chemical treatment plants (the COD values were the average of four determinations).{ A comparison of the results found by the proposed method with those of the authorised laboratory was performed and they were found to be statistically indistinguishable for all samples except sample 10/1338. The residue containing Zn(II) was non-hazardous (EC50~3173 mg l21) and the sample containing Cr(III) was hazardous (EC50~2748 mg l21) according to the 13/10/1989 BOE 270 10/11/89 order.The RDPH to continental waters establishes that the concentrations of Zn(II), Cr(III) and COD must not be higher than 3, 2 and 160 mg l21, respectively. The leaching liquid of both residues had a heavy metal concentration higher than the upper level allowed by the RDPH (see Table 2). Therefore, the residues required stabilisation treatment prior to deposition in landÆlls to ensure that their leaching liquids met the RDPH and were non-hazardous.The low COD values of the residues containing heavy metals indicated a low concentration of organic compounds in the leaching solution from these residues. The leaching solution of the Zn(II) residue gave 3175 mg Zn(II) l21. This high concentration of Zn(II) indicated that the residue was hazardous under these conditions; the EC50 of the Zn(II) residue should be lower than 3000 mg l21.This does not occur possibly due to the large amount of precipitate formed when the pH of the leaching solution is adjusted from 5.0°0.2 to 7.0°0.2 (pH at which the bioassay is developed). The precipitation of Zn(II) as hydroxide could be responsible for the loss of toxicity in the aliquot used for the determination of EC50.Fig. 1 Linearisation of the bioluminescence assay. The interception point with the x-axis shows the percentage of leaching sample which provides an EC50 of 3173 mg l21 and 2748 mg l21 for residues containing Zn(II) and Cr(III), respectively. Table 1 Temperature programme of the graphite furnace for the determination of Cr(III) Step Temperature/�C Ramp time/s Hold time/s 1 120 10 30 2 1400 30 20 3 2400 0 3 4 2500 1 2 {Results available as electronic supplementary information.See http:// www.rsc.org/suppdata/em/1999/563. J. Environ. Monit., 1999, 1, 563±568 565Treatments The residues were cured for intervals between 24 and 70 h at 36 �C with single or mixtures of stabilisers as follows: (1) with different percentages of Portland cement (10%, 20% and 30%) which stabilised heavy metals; (2) with the percentages of Portland cement as in (1) and a constant amount (3%) of Depocrete SM/2 in order to improve the stabilisation of heavy metals; (3) with the percentages of Portland cement as in (1) plus 5% lime in order to endow the residue with better mechanical properties for transport.After curing, the sample treatment was applied to 25 g of sample (residue plus stabilisation agent).Then, the EC50, heavy metal concentration and COD were determined and the results obtained are described below. With Portland cement. The best percentages for the stabilisation of heavy metals were 30% for the residue containing Zn(II) and 20% for the sample containing Cr(III) (see Table 3), because they provided the highest EC50 (EC50 at least Æve times higher than the EC50 of the residues without curing) and the concentrations of heavy metals were lower than 0.1 mg l21 Zn(II) and 0.7 mg l21 Cr(III).The leaching liquids of the Zn(II) residue mixed with 30% of Portland cement and the Cr(III) residue mixed with 20% of Portland cement yielded a heavy metal concentration and COD lower than the maximum level allowed by the RDPH.With this stabilisation treatment, the residues containing heavy metals were non-hazardous according to both the 13/10/1989 BOE 270 10/11/89 order (EC50 higher than 3000 mg l21) and RDPH to continental waters (which establishes that the concentration of Zn(II), Cr(III) and COD must not be higher than 3, 2 and 160 mg l21, respectively).Although 20% of Portland cement was best for the stabilisation of Cr(III), 10% of Portland cement was sufÆcient to obtain a non-hazardous cured product from the residue containing Cr(III) (see Table 3). The concentrations of heavy metals in the leaching solutions from both the Zn(II) residue plus 30% of Portland cement and the Cr(III) residue plus 20% of Portland cement were dramatically lowered with respect to the residue without curing: v0.1 mg l21 vs. 3175 mg l21 for Zn(II) and 0.55 mg l21 vs. 2.34 mg l21 for Cr(III). Thus, the decrease in toxicity was ca. 100% for Zn(II) and 76% for Cr(III). In fact, a dilution effect occurred when the residue was mixed with Portland cement because the amount of sample to be leached (the same weight in all instances) contained a percentage of stabiliser and so a lower amount of the original residue was treated.The concentration of heavy metal in the leaching solution from the stabilised residue was ca. three times lower than that produced by dilution; therefore the decrease in toxicity obtained for the Zn(II) residue plus 30% of Portland cement was ca. 100% and that due to dilution was only 30%, and for the Cr(III) residue plus 20% of Portland cement the real decrease of toxicity was 76% and that expected by dilution was 20%.This decrease was due to the immobilisation of heavy metal in the cement matrices and not only due to a dilution effect produced by the stabiliser. With Portland cement mixed with Depocrete SM/2. The improvement in the stabilisation of heavy metals in matrices of cement was studied by adding 3% of Depocrete SM/2 to a batch of samples containing Portland cement (see Table 4).A residue mixed with 20% cement and without Depocrete SM/2 was used as a control sample in order to establish the effect of Depocrete SM/2 (3%). Depocrete SM/2 (3%) plus Portland cement (20% and 30%) provided a higher stabilisation of heavy metals than the sample mixed with Portland cement (20%) without Depocrete SM/2. This mixture increased EC50 as a consequence of the decrease in the concentration of heavy metals [0.14 mg l21 Cr(III) and lower than 0.1 mg l21 Zn(II)] in the leaching liquid.So, we can conclude that Depocrete SM/2, when mixed with Portland cement, increases the stabilisation of heavy metals. The mixture containing the heavy metal residue, 20% of Portland cement and 3% of Depocrete SM/2 produced a dilution (23%) of the concentration of heavy metal in the stabilised residue.The decrease in toxicity obtained for the stabilised Zn(II) residue was ca. 100% (v0.1 mg l21 vs. 3175 mg l21) and that expected by dilution was 23%, and for the stabilised Cr(III) residue the decrease in toxicity was 94% (0.14 mg l21 vs. 2.34 mg l21) and that due to dilution was 23%. The immobilisation of heavy metals in cement matrices (Portland cement plus Depocrete SM/2) is the main factor responsible for the decrease in their concentration in the leaching solution, the dilution by the stabilisers having a small effect. With Portland cement mixed with lime. The residues stabilised with Portland cement and Depocrete SM/2 had a rock-like aspect made up of a compact mass, whose size was a function of the amount of treated residue.For this reason, the transport of large amounts of residues stabilised by this treatment would be difÆcult. In order to endow the residues with better mechanical properties for transport, lime (5%) and Portland cement (10%, 20% and 30%) were utilised as stabiliser agents.After curing, the samples had a granusize making Table 2 Results of the bioassay, FAAS, GF-AAS and COD Sample EC50/mg l21 Heavy metal/mg l21°RSD (%) COD/mg l21°RSD (%) Residue containing Zn(II) 3173c 3175a°17.39 41.2°1.1 Residue containing Cr(III) 2748 2.34b°0.12 33.4°0.9 aDetermination of the concentration of Zn(II) by FAAS. bDetermination of the concentration of Cr(III) by GF-AAS.cFormation of precipitate by adjusting the pH of the leaching solution from 5.0°0.2 to 7.0°0.2. Table 3 Stabilisation of samples containing Zn(II) or Cr(III) using Portland cement Sample Portland cement (%) EC50/mg l21 Heavy metal/mg l21°RSD (%) COD/mg l21°RSD (%) Zn(II)a 10 3220c – 39.1°1.0 20 3600c – 38.1°1.2 30 16 530 v0.1d 37.8°1.2 Cr(III)b 10 12 559 1.46°0.017 31.3°1.1 20 19 637 0.55°0.066 30.0°0.8 30 19 448 0.68°0.074 30.2°1.2 aCuring time: 24 h at 36 �C.bCuring time: 40 h at 36 �C. cFormation of precipitate by adjusting the pH of the leaching solution from 5.0°0.2 to 7.0°0.2. dConcentrations lower than the detection limit (1 mg l21) are not detected. 566 J. Environ. Monit., 1999, 1, 563±568their transport easier.A residue mixed with 20% Portland cement without lime was used as a control sample in order to establish the effect of lime (5%). Concerning the toxicity, samples with Portland cement plus 5% of lime showed a slight decrease in EC50 compared with the sample mixed with 20% of Portland cement without lime, which could be attributed to an increase in the ionic strength of the leaching solution with the subsequent deactivation of Photobacterium phosphoreum.This decrease in EC50 was overwhelmingly surpassed by the stabilisation effect of the Portland cement, and the samples cured with lime (5%) and Portland cement (20% and 30%) showed an EC50 higher than 3000 mg l21 and the concentrations of Zn(II) and Cr(III) were lower than 0.1 mg l21 (see Table 5). Therefore, the samples thus treated were non-hazardous and had good mechanical properties for transport.Curing time The residues containing Zn(II) or Cr(III) mixed with 20% of Portland cement were cured for 24±40 h and 30±70 h, respectively. The concentration of heavy metal in the leaching solution of the stabilised residues decreased and the EC50 increased with increasing curing time (see Tables 3±5).Therefore, the stabilisation of the heavy metal in the cement matrix increased with increasing curing time. The residue containing Zn(II) cured for 24 h at 36 �C with 30% of Portland cement was non-hazardous [EC50~16 530 mg l21 and concentration of Zn(II) in the leaching solution v0.1 mg l21], as was the Zn(II) residue cured for 40 h at 36 �C with 20% of Portland cement [EC50~25 059 mg l21 and concentration of Zn(II) in the leaching solution v0.1 mg l21] (see Tables 3 and 5).The percentage of Portland cement needed in the stabilisation decreases from 30% to 20% with an increase in curing time from 24 h to 40 h. Without stabilisation, the residue containing heavy metal showed a concentration of Zn(II) or Cr(III) in the leaching solution of 3175 and 2.34 mg l21, respectively (see Table 2).In the residue stabilised with 20% of Portland cement, the concentration of Zn(II) or Cr(III) was v0.1 and 0.55 mg l21, respectively, after a curing time of 40 h at 36 �C (see Tables 3 and 5). The stabilisation with cement led to a greater decrease in the concentration of Zn(II) than Cr(III) in the leaching solution (a decrease of ca. 100% vs. 76%, respectively). Thus, Portland cement is a more effective stabiliser for residues containing Zn(II) than for residues containing Cr(III). A residue containing Cr(III) stabilised with 20% of Portland cement without Depocrete SM/2 and cured for 40 h at 36 �C showed a concentration of Cr(III) in its leaching solution [0.55 mg l21 of Cr(III)] higher than that obtained by stabilisation with 20% of Portland cement mixed with 3% of Depocrete SM/2 and cured for 30 h at 36 �C [0.14 mg l21 of Cr(III)] (see Tables 3 and 4).Depocrete SM/2 is a special cement which, when mixed with Portland cement, cures faster than Portland cement alone and with less curing time (30 h vs. 40 h). The stabilisation of Cr(III) with 3% of Depocrete SM/2 and 20% of Portland cement [0.14 mg l21 of Cr(III)] is 74% higher than when only 20% of Portland cement [0.55 mg l21 of Cr(III)] is added.Precision study The precision of the applied methodology was studied in terms of both repeatability (within-day study) and reproducibility (between-week study). The study of repeatability was performed on Æve samples prepared from a residue containing Zn(II) plus Portland cement (30%) cured at 36 �C for 40 h (Table 6).The percentage relative standard deviation of EC50 Table 4 Stabilisation of samples containing Zn(II) or Cr(III) using Portland cement plus Depocrete SM/2 Sample Stabiliser EC50/mg l21 Heavy metal/mg l21°RSD (%) COD/mg l21°RSD (%) Zn(II)a 10% cementz3% Depocrete SM/2 3624d 1032°30.63 38.6°1.4 20% cementz3% Depocrete SM/2 26 852 v0.1e 37.3°1.1 30% cementz3% Depocrete SM/2 26 826 v0.1e 37.8°1.1 20% cementc 14 420 17.60°0.15 38.3°1.3 Cr(III)b 10% cementz3% Depocrete SM/2 15 811 0.59°0.062 31.0°1.1 20% cementz3% Depocrete SM/2 31 548 0.14°0.080 30.2°1.3 30% cementz3% Depocrete SM/2 30 830 0.25°0.048 29.4°1.5 20% cementc 7194 0.71°0.062 30.8°1.0 aCuring time: 33 h at 36 �C.bCuring time: 30 h at 36 �C. cControl sample.dFormation of precipitate by adjusting the pH of the leaching solution from 5.0°0.2 to 7.0°0.2. eConcentrations lower than the detection limit (1 mg l21) are not detected. Table 5 Stabilisation of samples containing Zn(II) or Cr(III) using Portland cement plus lime Sample Stabiliser EC50/mg l21 Heavy metal/mg l21°RSD (%) COD/mg l21°RSD (%) Zn(II)a 10% cementz5% lime 15 811 5.18°0.21 23.2°0.8 20% cementz5% lime 19 905 v0.1d 22.1°1.4 30% cementz5% lime 19 452 v0.1d 20.8°1.1 20% cementc 25 059 v0.1d 36.8°1.2 Cr(III)b 10% cementz5% lime 32 283 0.11°0.012 22.4°0.9 20% cementz5% lime 43 548 0.04°0.009 21.6°1.0 30% cementz5% lime 42 557 0.11°0.01 21.4°1.0 20% cementc 57 407 0.016°0.009 29.7°1.2 aCuring time: 40 h at 36 �C.bCuring time: 70 h at 36 �C.cControl sample. dConcentrations lower than the detection limit (1 mg l21) are not detected. Table 6 Repeatability study Stabiliser EC50/mg l21 Heavy metalb/ mg l21 COD/mg l21 °RSD (%) 30% cementa 23 016 v0.1 35.5°1.3 23 553 v0.1 36.5°1.5 23 750 v0.1 35.2°1.0 24 030 v0.1 36.8°1.3 23 332 v0.1 34.5°1.1 aFive samples of residue containing Zn(II) plus 30% of Portland cement. Curing time: 40 h at 36 �C.bConcentrations lower than the detection limit (1 mg l21) are not detected. J. Environ. Monit., 1999, 1, 563±568 567was 1.6%. The reproducibility of the method was studied on Æve samples of residue containing Cr(III) plus Portland cement (20%) and Depocrete SM/2 (3%). Every 5 days one sample was cured at 36 �C for 40 h. The value found, expressed as the relative standard deviation of EC50, was 5.1% (Table 7).Conclusions From an analytical point of view, the best results of stabilisation of the residues containing Zn(II) or Cr(III) were obtained when 20% of Portland cement plus 3% of Depocrete SM/2 were used as stabiliser, because the stabilised residue had the highest EC50, the lowest COD and concentration of heavy metal and the shortest curing time.Cement matrices are beneÆcial for reducing the leaching of Cr(III) or Zn(II) because they decrease their solubility by both physical stabilisation (adsorption of ions on the surface of the matrix) and chemical stabilisation [formation and precipitation in the matrix of hydroxides of Zn(II) or Cr(III) due to their high pH value (pH 11)]. From an industrial point of view, the costs of the stabiliser, transport and storage must be evaluated in order to select the percentage of stabiliser.Depocrete SM/2 (0.24 euros kg21) is six times more expensive than Portland cement (0.04 euros kg21). The 3% of Depocrete SM/2 decreases the curing time needeeavy metals in cement matrices and the Ænal residue has a larger structure.The leaching of heavy metals is minimised but the high cost of Depocrete SM/2 increases the price of the stabilisation process. Most industries do not use Depocrete SM/2 for stabilising residues containing heavy metals for economic reasons. Portland cement is a cheaper stabiliser, and 20% or 10% of this cement mixed with residues containing 3.4% of Zn(II) or 1% of Cr(III), followed by curing for 40 h at 36 �C, is sufÆcient to make the residues non-hazardous and thus they can be deposited in landÆlls.When the residues must be transported from the place where they have been treated to a landÆll, 5% of lime is recommended together with 20% of Portland cement to stabilise the residues. Lime endows the residues with better mechanical properties for transport in tankers rather than in individual tanks, thus reducing costs.This treatment also degrades the organic compounds, thus eliminating or minimising unpleasant smells in landÆlls. An increase in the curing time decreases the percentage of Portland cement needed for stabilisation, which leads to a reduction in the cost of stabiliser and the costs of transport and storage because the Ænal amount of sample (residue plus stabiliser) transported to the landÆll decreases. Acknowledgement The Comisio�n Interministerial de Ciencia y Tecnologý�a (CICyT) is thanked for Ænancial support (Project No PB96- 0505).References 1 P. Cambier, Analusis Magazine, 1994, 22, M21. 2 C. M. Davidson and A. L. Duncan, Anal. Chim. Acta, 1998, 363, 45. 3 T. Taylor and B. S. Crannell, Environ. Sci. Technol., 1997, 31, 3330. 4 B. Lothenbach, G. Furrer and R. Schulin, Environ. Sci. Technol., 1997, 31, 1452. 5 C. M. Lytle, F. W. Lytle, N. Yang, J. H. Qian, D. Hansen, A. Zayed and N. Terry, Environ. Sci. Technol., 1998, 32, 3087. 6 A. R. Pratt, D. W. Blowes and C. J. Ptacek, Environ. Sci. Technol., 1997, 31, 2492. 7 D. W. Blowes, C. J. Ptacek and J. L. Jambor, Environ. Sci. Technol., 1997, 31, 3348. 8 A. Bobrowski and J. Malolepszy, Environ. Sci. Technol., 1997, 31, 745. 9 A. Kindness, A. Macias and F. P. Glasser, Waste Management, 1994, 14, 3. 10 T. Korenaga and H. Ikatsu, Analyst, 1981, 106, 653. 11 M. L. Balconi, M. Borgarello, F. Realini and R. Ferraroli, Anal. Chim. Acta, 1992, 261, 295. 12 Ministerio de la Presidencia, Luminescence Bioassay, Boletý�n OÆcial del Estado, Madrid, 1989, Anexo 4, p. 35 220. 13 B. Lange, Operation Manual for Luminescent Bacteria Test with Lumistox, Gmb Berlin Industrience Technik, Dusseldorf, Edition 02/93, 1993, p. 1. 14 Ministerio de Obras Pu� blicas y Urbanismo, Reglamento del Dominio Pu�blico Hidra�ulico, Madrid, Real Decreto 849/1986, p. 419. 15 Association of OfÆcial Analytical Chemists, OfÆcial methods of Analysis, ed. K. Helrich, Arlington, Virginia, 15th edn., 1990, pp. 317±318. 16 B. Vallejo, A. Izquierdo and M. D. Luque de Castro, Analyst, 1999, 124, 1261. Paper 9/03668G Table 7 Reproducibility study Stabiliser EC50/mg l21 Heavy metal/mg l21°RSD (%) COD/mg l21°RSD (%) 20% cement plus 3% Depocretea 32 889 0.08°0.012 29.5°1.0 31 203 0.08°0.009 30.5°1.1 33 528 0.09°0.008 30.2°1.3 34 028 0.09°0.009 29.8°1.5 36 119 0.10°0.008 33.1°1.0 aFive samples of residue containing Cr(III) plus 20% of Portland cement and 3% of Depocrete SM/2. Curing time: 40 h at 36 �C. 568 J. Environ. Monit., 1999, 1

ISSN:0960-7919    

DOI:10.1039/a903668g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Detection of chlorophenolics in efØuents from bleaching processes of rice-straw pulp C. Sharma and S. Kumar* Institute of Paper Technology (U.O.R.), Saharanpur-247 001 (U.P.), India Received 24th June 1999, Accepted 8th October 1999 A number of chlorinated derivatives of phenols, catechols, guaiacols, syringaldehydes, have been detected and their concentrations estimated, using gas chromatography in the chlorination (C) and extraction (E) stage of spent bleach liquor generated in the laboratory by bleaching rice-straw soda pulp.The concentration of various compounds detected have also been compared with their reported 96LC50 values. Aim of investigation The destruction of forests is one major impact of the paper industry in seeking to meet its raw material demand.Dwindling forest cover and greater environmental awareness has reduced the availability of forest-based raw material like bamboo, hardwoods and softwoods. As a consequence the dependence on annual Æbers and agricultural residues for papermaking has increased substantially in India. The pulp produced by digestion with chemicals is brownish in color and requires bleaching to produce pulps of acceptable brightness for further processing.In India and many other countries, the use of chlorine and other chlorinated compounds like calcium or sodium hypochlorite with intermediate caustic extraction is common for all the raw materials. Extensive research has been carried out on the identiÆcation of chlorinated organic materials in bleach plant efØuent. Lower molecular weight chlorinated phenolic compounds (i.e., guaiacols, phenols, catechols, and vanillins) formed during pulp chlorination have been identiÆed in pulp mill bleaching efØuent.1,2 Approximately 300 different compounds have been identiÆed to date in bleached pulp mill efØuents.About 200 of these are chlorinated organic compounds.3±5 Approximately 75±80% of the organically bound chlorine in bleach kraft mill efØuent is present in large molecular weight material which is not easily identiÆed or even characterized.3±5 Some resin acids and fatty acids which originate from the raw materials have also been detected.The chlorophenolics are formed during the chlorination stage (Cstage) of pulp bleaching and these get solubilised in the Ærst extraction stage (Estage).The values and concentration of chlorophenolics formed will depend upon the nature of the lignin in the raw material and bleaching conditions. The concentration of resin acids and fatty acids present in the pulp mill bleaching efØuents depends upon wood species and on the degree of washing of the unbleached pulp. During the last three decades, intensive research effort has been devoted to identifying various compounds2,6±8 in bleach plant efØuents and to investigating their possible biological effects.9±13 These studies have been performed mostly on softwoods, and to some extent on hardwoods.Very little information is available on the nature and the quantities of various compounds present in bleach plant efØuents formed from Indian varieties of hardwoods or agro-residues.In India, a number of agro-residue materials like bagasse, wheat straw and rice straw are used for papermaking. Rice straw is being used by a number of small paper mills. The average Æber length of rice-straw Æber is small and yields a paper of lower strength properties viz., tensile, tear and burst strength. A small amount of long Æber pulp, viz., softwood, is blended with rice-straw pulp to improve its strength properties.In the present investigation we report the results on the detection and quantitative determination of various chlorophenolics formed during chlorination and Ærst extraction stage efØuent obtained from the bleaching of rice-straw pulp. Experimental procedures Chlorophenols were obtained from Aldrich, Milwaukee, WI, USA.Chlorocatechols, chloroguaiacols, chlorovanillins; chlorosyringaldehydes, chlorosyringols, resin acids and chloro fatty acids were supplied by Helix, Richmond, BC, Canada. Solvents, viz., n-hexane, acetone and diethyl ether, used were of HPLC grade. Analytical grade acetic anhydride was used after redistillation. Other reagents used for detection studies were of analytical reagent grade.Standard solutions of chlorophenolics were prepared in 10 : 90 acetone±water. The requirement for bleaching chemicals depends upon the quantity of lignin (a brown colored binding material present in raw materials such as woods/agro-residues) which gets dissolved during the pulping reaction. The quantity of residual lignin in the pulp is determined by carrying out a reaction with acidiÆed KMnO4.The consumption of KMnO4 is expressed as a number called the kappa number. This number is empirically related to the residual lignin content of the pulp. Unbleached washed straw soda pulp was taken from a nearby institution. The kappa number (residual lignin) of the pulp was determined (Tappi Test method T236 cm-85). It was found to be 21. The chlorine demand was calculated from the following formula: chlorine demandÖ%Ü~0:25|kappa number Pulp bleaching is carried out in more than one bleaching stage. 70% of chlorine demand has been applied as elemental chlorine during the chlorination stage. All chemicals are applied as %OD pulp (oven dried; heated to constant weight in an oven at 105 �C for 12 h). Unbleached pulp (equivalent to 35 g OD pulp) was bleached in the laboratory to generate efØuents.The bleaching conditions are shown in Table 1. The volumes of efØuent generated in the Cstage and Estage were 1.81 and 2.02 l, respectively. The efØuents were characterized by determining the pH, total dissolved solids, BOD5 (biochemical oxygen demand), COD (chemical oxygen demand) and color. Dissolved solid (DS), BOD5 and COD (open reØux method) were determined by standard methods.14 Color measurement was performed spectrophotometrically on a Shimadzu (Singapore) spectrophotometer model UV 2100/S.J. Environ. Monit., 1999, 1, 569±572 569 This journal is # The Royal Society of Chemistry 1999250 mg of platinum wire (purity 99.99%) was dissolved in hot aqua regia; nitric acid was removed (as oxides of nitrogen) by repeated evaporation with fresh portions of concentrated hydrochloric acid.The residue was dissolved in distilled water and then 500 mg of CoCl2?6H2O (equivalent to 125 mg Co) was added along with 5 ml conc. HCl and the total volume was made up to 100 ml with distilled water to yield a standard solution of 2500 Pt.Co color units. Solutions of different Pt.Co units were prepared by dilution.The absorbance of different standard solutions adjusted to pH 7.6 were determined at 465 nm and a calibration curve was plotted. The pH of the efØuent was adjusted to 7.6 and the efØuent was centrifuged for 5 min at 1500 rpm to remove suspended particulates. The absorbance of this solution was determined at 465 nm and the color of the efØuent was computed from the calibration curve.Extraction of chlorophenols from the efØuents was performed by simple modiÆcation of the procedure suggested by Lindstrom and Nordin.2 500 ml of Estage efØuent or 1000 ml of Cstage efØuent was adjusted to pH 2 and extracted with 200 ml and 400 ml, respectively of 90 : 10 diethyl ether±acetone mixture for 48 h with intermittent shaking. A schematic presentation of the method is shown in Flow sheet 1.Gas chromatographic (GC) studies were carried out on a Shimadzu gas chromatograph (GC-9A model). The analysis of chlorophenols as acetyl derivatives15 was performed on an Ulbon HR-1 glass capillary column (Shimadzu). The GC conditions are given in Table 2. Retention times (RT) and response factors (RF) were determined by preparing standard solutions of various chlorophenolic compounds (100 mg l21) in 10 : 90 acetone±water solution. 1 ml sample was derivatized to form acetyl derivatives and 1 ml of the derivatized extract was injected onto the column. The retention te and area of the peak were recorded. The response factor was calculated using the following equation: RF~ weight of the sample injected ÖpgÜ area of the peak Derivatization procedure To 4.5 ml of sample (or diluted sample) in a TeØon lined screw capped glass tube, 0.5 ml of 0.5 M Na2HPO4 and 0.05 ml of acetic anhydride were added.After adding 1 ml of n-hexane, the mixture was shaken for 3 min and 0.5±1 ml of the hexane extract was injected onto the HR-1 column. 1000 ml of standard solution containing 4 to 5 different chlorophenolics (RT values not too close to each other) each having a concentration of 0.1 mg l21 was extracted using the procedure outlined in Flow sheet 1 and derivatized as outlined above. 1 ml of solution (100 mg l21) containing 0.1 mg each of chlorophenolics was also derivatized. 0.5 ml of n-hexane extract for both extracted and non-extracted chlorophenolics were injected into HR-1 column to determine the peak areas in both the cases.The extraction efÆciency was calculated from the equation: extraction efficiency Ö%Ü~ peak area of extracted sample peak area of non-extracted sample |100 (3) The various chlorophenolics were detected by matching their retention time (°0.1 min) with those of pure standards. However certain compounds have a very similar RT.In these cases the quantities have been determined by assuming only one compound is present and calculating the average value. For quantitative analyses, the response factors (RF) and extraction efÆciencies (EF) of various compounds were determined. Results The characteristics of efØuents generated in the laboratory are shown in Table 3. The values of retention time (RT) and the Table 1 Bleaching conditions.All chemicals charged as %OD (oven dried) pulp Parameter Cstage Estage Cl2 applied (%) (demand) 70 – NaOH applied (%) (OD pulp) – 3 Consistencya (%) 3.5 10 pH 1.8±2.0 11±12 Temperature/�C 25 60 Time/min 60 75 aConsistency is the ratio of g OD pulp per 100 g pulp suspension, expressed as %. Table 2 GC conditions Column Parameter HR-1 Detector FID Detector range 10� Chart speed/cm min21 5 Sample size/ml 0.5±1 Injection splitless/min 2 Column dimensions 30 m60.32 mm Film thickness 0.25 mm Injection and detector temp./�C 275 Column temperature/�C 80 for 3 min 80±160 at 2 �C min21 160 for 5 min 160±260 at 10 �C min21 260 for 15 min Flow sheet 1 Separation of phenolic compounds from efØuents. 570 J. Environ. Monit., 1999, 1, 569±572concentrations of various detected chlorophenols in laboratory generated spent bleach liquor (SBL) are given in Table 4.The concentrations of chlorocatechols may be regarded as semiquantitative because of the low extraction efÆciency of these compounds. The values of the retention times (Table 4) show that most of the chlorophenolic compounds can be separated on an Ulbon HR-1 glass capillary column.The quantities of various chlorophenolics in bleach plant efØuent, expressed as percentages, are shown in Fig. 1. Discussion The results given in Table 4 show that six categories of chlorophenolics are present in spent bleach liquor obtained from Indian variety of rice-straw pulp. These are simple chlorophenols, chlorocatechols, chloroguaiacols, chlorosyringols, chlorovanillins and chlorosyringaldehydes.The structure of lignin is very complex. It is a polymer formed by enzyme initiated dehydroabietic polymerization of a mixture of three different p-hydroxycinnamyl alcohols (pcoumaryl, coniferyl and sinapyl alcohols). Compared with wood lignin the structure of grass lignin is less well understood and it varies signiÆcantly with source.Some grass lignins are thought to contain mainly p-coumaryl units but other grass lignins appear to approximate to hardwood lignin.16 During pulp chlorination lignin gets chlorinated and broken down to simpler chlorophenolic compounds. The solubility of chlorophenolics is low in acidic media (Cstage) but they are soluble in alkaline media (Estage). The nature and concentration of the different chlorophenolic compounds formed, that ultimately Table 4 Retention time, extraction efÆciency and concentration of various chlorophenolics in the efØuent Concentration/g odt21a Chlorophenolic Retention time/min Extraction efÆciency (%) Response factor/pg CStage EStage 2-Chlorophenol 7.82 105 0.16 0.03 0.02 2,6-Dichlorophenol 12.96 64 0.33 0.09 – 2,5-Dichlorophenol/ 13.97 106 0.35 0.39 6.15 2,4-Dichlorophenol 14.05 87 0.57 3,4-Dichlorophenol 17.51 81 0.25 0.02 0.11 6-Chloroguaiacol 17.94 103 0.31 0.03 0.08 2,4,6-Trichlorophenol 19.10 83 0.35 0.28 6.68 2,3,5-Trichlorophenol 22.30 88 0.38 0.09 0.11 2,3,4-Trichlorophenol 24.61 63 0.23 0.02 0.34 4,6-Dichloroguaiacol 24.88 102 0.68 0.05 0.96 3,4-Dichloroguaiacol 25.44 106 0.36 0.06 0.29 4,5-Dichloroguaiacol 27.96 79 0.54 0.16 3.46 3,6-Dichlorocatechol 28.51 3 0.26 8.86 – 3,5-Dichlorocatechol 29.65 5 0.70 18.38 – 3,4,6-Trichloroguaiacol/ 30.49 70 0.51 0.18 9.56 6-Chlorovanillin 30.58 103 0.50 4,5-Dichlorocatechol 33.36 5 0.41 1.26 0.15 3,4,5-Trichloroguaiacol 34.01 108 0.37 0.46 6.26 4,5,6-Trichloroguaiacol 36.04 100 0.22 – 0.38 3,4,6-Trichlorocatechol 36.29 28 0.39 0.88 – 5,6,-Dichlorovanillin 38.07 73 0.69 0.09 – 2-Chlorosyringaldehyde 38.75 110 1.20 26.59 – Pentachlorophenol 39.21 51 0.35 – 4.39 3,4,5-Trichlorocatechol 39.79 21 0.35 10.46 3.84 Tetrachloroguaiacol 40.58 51 0.50 0.08 2.57 Trichlorosyringol 41.70 110 0.28 0.23 2.59 Tetrachlorocatechol 45.96 33 0.48 4.38 – 2,6-Dichlorosyringaldehyde 46.38 75 2.36 6.78 26.38 aodt : oven dried tonne pulp.Table 3 Characterization of efØuent EfØuent Parameter CStage EStage pH 2.5 10.8 Dissolved solid/mg l21 1088 1286 BOD5/mg l21 96 135 COD/mg l21 481 688 Color (Pt.Co units) 786 3018 Fig. 1 Quantity(g odt21) of chlorophenolic compounds from bleach plant efØuent (CstagezEstage). J. Environ. Monit., 1999, 1, 569±572 571end up in SBL, depends upon the quantity of lignin (i.e., kappa number of pulp), the nature of the lignin, and the bleaching conditions, i.e., chlorine added, pH, temperature and consistency (expressed as g OD (oven dried) pulp/100 g pulp suspension).It has been found that in Cstage efØuent: 3,5-dichlorocatechol (18.38 g odt21) and 2-chlorosyringaldehyde (26.59 g odt21) are the major identiÆed chlorophenolic compounds; 3,4,5- trichlorocatechol (10.46 g odt21), 3,6-dichlorocatechol (8.86 g odt21), 2,6-dichlorosyringaldehyde (6.78 g odt21) and tetrachlorocatechol (4.38 g odt21) are present in signiÆcant quantities; and other chlorophenolic compounds are present in very small quantities.In Estage efØuent: 2,6-dichlorosyringaldehyde (26.38 g odt21) is the major identiÆed chlorophenolic compound; 3,4,6- trichloroguaiacol/6-chlorovanillin (9.56 g odt21), 2,4,6-trichlorophenol (6.68 g odt21), 3,4,5-trichloroguaiacol (6.26 g odt21) and 2,4/2,5-dichlorophenol (6.15 g odt21) form a signiÆcant proportion of the total identiÆed chlorophenolic compounds; pentachlorophenol (4.39 g odt21), 3,4,5-trichlorocatechol (3.84 g odt21), 4,5-dichloroguaiacol (3.46 g odt21), tetrachloroguaiacol (2.59 g odt21) and trichlorosyringol (2.59 g odt21) were identiÆed as minor components; and other chlorophenolic compounds were identiÆed in very small quantities.The proportions as well as the quantities of mono-, di-, tri-, tetra- and pentachlorophenolic compounds (g odt21) in the combined SBL (CstagezEstage) are shown in the pie charts (Fig. 1). Results (Fig. 1 and Table 4) indicate that: dichlorophenolic compounds contribute about 47% of the total detected chlorophenolic compounds; chlorocatechols and others chlorinated phenolics contribute a 75% share; chloroguaiacols are greater in Estage than in Cstage efØuent; and chlorocatechols are predominant in Cstage efØuent.The results are similar to those found in the spent bleaching liquor of wood pulps and are due, presumably, to the low solubility of chlorinated guaiacols and the sorption of theson the Æbers at low pH or the formation of chloroguaiacols only upon alkaline hydrolysis of chlorinated lignin.Toxicity 96LC50 is a parameter which represents the toxicity of a particular compound. It is deÆned as the lethal concentration at which 50% of the test organisms will be killed when the test organism is exposed to the toxicant for a period of 96 h under standard test conditions. To estimate whether the efØuent is toxic, the concentrations of individual pollutants are compared with 96LC50 values.The 96LC50 value describes the toxicity of a particular compound when present alone. However when a number of toxic compounds are present, interfering or synergistic effects may be observed. Substantial evidence now exists which indicates that the threshold concentration is approximately 0.05±0.10 of 96LC50 values.17 At this concentration (or below) no sublethal stresses have been observed in pulp mill efØuent.18,19 Reported 96LC50 values show that chlorophenolics are toxic in nature.On comparing the concentrations of chlorophenolics, detected in Cstage and Estage efØuents with the reported 96LC50 values,1,20 it is found that: the concentration of dichlorocatechol (0.53 mg l21) in Cstage efØuent is higher than the reported 96LC50 value (0.5±1.0 mg l21); concentrations of all other chlorophenolics are below the reported respective 96LC50 values; and the concentrations of some of the chlorophenolics exceed their respective threshold concentrations.From the above it can be inferred that the untreated spent bleach liquor of rice straw generated in the laboratory is of concern from a toxicity point of view. Some more experiments are needed to conÆrm the above Ændings. Financial assistance for this work was provided by Council of ScientiÆc and Industrial Research (CSIR), New Delhi, India. References 1 R.H. Voss, J. T. Wearing, R. D. Mortimer, T. Kovaks and A. Wong, Paper Puu, 1980, 62(12), 809. 2 K. Lindstrom and J. Nordin, J. Chromatogr., 1976, 128, 13. 3 A. Sodergren, Symposium preprints, 2nd IAWPRC symposium on forest waste water, Biological treatment and environmental effects of pulp and paper industry waste water, Tempere, Finland, June 9±12, 1987. 4 L. Suntio, W. Y. Shiu and D.A. Mackay, Chemosphere, 1988, 17(7), 1249. 5 K. P. Kringstad and A. B. Mckague, TAPPI/CPPA International Pulp Bleaching Conference Proceeding, Orlando, June 5±9, 1988, 63. 6 K. P. Kringstad and K. Lindstrom, Environ. Sci. Technol., 1984, 18(8), 236A. 7 J. Knuutinen, J. Chromatogr., 1982, 248, 289. 8 R. H. Voss and A. Rapsomatiotis, J. Chromatogr., 1985, 346, 205. 9 C. C. Walden and T. E. Howard, Tappi, 1977, 60(1), 122. 10 I. H. Rogers, J. C. Davis, G. M. Kruzynski, H. W. Mahood, J. A. Servizi and R. W. Gordon, Tappi, 1975, 58(7), 136. 11 B. Holmbom, R. H. Voss, R. D. Mortimer and A. Wong, Environ. Sci. Technol., 1984, 18(5), 333. 12 D. W. Reeve and P. F. Earl, Pulp Pap. Can., 1989, 90(4), 128. 13 R. Crooks and J. Sikes, Appita, 1990, 43(1), 67. 14 Standard Methods for Examination of Water and Wastewater, American Public Health Association, Washington, DC, 16th edn., 1985. 15 K. Abrahamsson and T. M. Xie, J. Chromatogr., 1983, 279, 199. 16 J. M. Harkin, in Chemistry and Biochemistry of Herbage, ed. G. W. Butler and R. W. Bailey, Academic Press, New York, 1973, vol. 1, ch. 7. 17 Environment Management in Pulp and Paper, UNEP-Industry and Environment Manual Series 1981, vol. 1, pp. I±59 and I±64. 18 L. Landner, L. Lindestrom and O. Linden, Review of today's knowledge, IVL-Pub., 1977, NM-80. 19 M. Waldichuk, Can. Fish Cult., 1962, 31, 3. 20 S. A. Heimburger, S. B. Daniel, H. B. Joseph and D. G. Paul, Tappi, 1988, 71(10), 69. Paper 9/05053A 572 J. Environ. Monit., 1999, 1, 569±572

ISSN:0960-7919    

DOI:10.1039/a905053a

出版商:RSC    

年代:1999

数据来源: RSC