RCRA Toxicity Characterization of
Computer CPUs and Other Discarded
Electronic Devices
Prepared By:
Timothy G. Townsend, Principal Investigator
Kevin Vann, Graduate Research Assistant
Sarvesh Mutha, Graduate Research Assistant
Brian Pearson, Graduate Research Assistant
Yong-Chul Jang, Post-Doctoral Associate
Stephen Musson, Graduate Research Assistant
Aaron Jordan, Student Assistant
Department of Environmental Engineering Sciences
University of Florida
Gainesville, Florida
Sponsored By:
United States Environmental Protection Agency,
Region 4 and Region 5
July 15, 2004
-------
This report was prepared by the Department of Environmental Engineering Sciences at the
University of Florida under the direction of Dr. Timothy Townsend. Please direct questions to him
at ttown@ufl.edu.
-------
Table of Contents
EXECUTIVE SUMMARY VIII
1 INTRODUCTION 1-1
1.1 THE MANAGEMENT OF DISCARDED ELECTRONICS 1-1
1.2 THE NEED FOR TOXICITY CHARACTERISTIC
TESTING 1-2
1.3 RESEARCH SCOPE AND REPORT DESCRIPTION 1-5
2 FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FROM COMPUTER CPUS 2-1
2.1 INTRODUCTION 2-1
2.2 MATERIALS AM) METHODS 2-1
2.2.1 Sample Collection and Processing 2-1
2.2.2 Leaching and Analysis Methods 2-2
2.2.3 TCLP on Printed Wire Boards 2-3
2.2.4 TCLP on Synthetic Computer CPU Mix/Particle
Size Study 2-4
2.2.5 Component Impact 2-4
2.2.6 Impact of Head Space 2-5
2.3 RESULTS AND DISCUSSION 2-6
2.3.1 Lead Leachability from PWBs 2-6
2.3.2 Predicting the TC of an Electronic Device 2-7
2.3.3 Synthetic Computer CPU Mix 2-7
2.3.4 Component Impact 2-8
2.3.5 Effects of Particle Size 2-10
2.3.6 Impact of Head Space 2-10
2.3.7 Comparison of Filtered vs. Nonfiltered Samples 2-11
2.4 DISCUSSION 2-12
2.4.1 TCLP of PWBs vs. Whole CPUs 2-12
2.4.2 Impact of CPU Components 2-13
2.4.3 Impact of Extraction Vessel Headspace 2-14
i
-------
3 EVALUATION OF A LARGE-SCALE MODIFIED TCLP
FOR RCRA TOXICITY CHARACTERIZATION OF
COMPUTER CPUS 3-1
3.1 INTRODUCTION 3-1
3.2 MATERIALS AM) METHODS 3-1
3.2.1 Re search Approach 3-1
3.2.2 Sample Collection and Processing 3-2
3.2.3 Large-Scale Leaching Procedure 3-2
3.2.4 Impact of Extractor Speed 3-2
3.2.5 Time Studies 3-3
3.2.6 Methodology Comparison 3-3
3.3 RESULTS 3-5
3.3.1 Impact of Extractor Speed 3-5
3.3.2 Time Studies 3-5
3.3.3 Methodology Comparison 3-9
3.4 DISCUSSION 3-14
4 LEACHING RESULTS FOR VARIOUS ELECTRONIC
DEVICES 4-1
4.1 OVERVIEW OF TESTING PERFORMED 4-1
4.2 PERSONAL COMPUTER CPUS 4-2
4.3 COMPUTER MONITORS 4-5
4.4 LAPTOP COMPUTERS 4-7
4.5 PRINTERS 4-9
4.6 COLOR TELEVISIONS 4-11
4.7 VCRS 4-13
4.8 CELLULAR PHONES 4-15
4.9 KEYBOARDS 4-18
4.10 COMPUTER MICE 4-20
4.11 REMOTE CONTROLS 4-21
4.12 SMOKE DETECTORS 4-23
4.13 FLAT PANEL MONITORS 4-25
5 SUMMARY OF RESULTS 5-1
6 REFERENCES 6-1
ii
-------
List of Tables
Table 1-1. TC Concentrations for Heavy Metals 1-3
Table 2-1. Ferrous Metal Impact Sample Compositions 2-4
Table 2-2. Impact of Vessel Headspace Sample Composition 2-5
Table 2-3. Average TCLP Leachate Concentrations from PWBs 2-7
Table 2-4. Component Impact TCLP Results 2-9
Table 2-5. TCLP Results of Synthetic Computer CPU Mixture 2-10
Table 2-6. Impact of Head Space Results 2-11
Table 3-1. Testing Methodologies 3-4
Table 3-2. Methodology Comparison Results 3-11
Table 4-1. Electronic Devices Tested 4-1
Table 4-2. TCLP Results for CPUs 4-3
Table 4-3. TCLP Results for Computer Monitors 4-6
Table 4-4. TCLP Results for Laptops 4-8
Table 4-5. TCLP Results for Printers 4-10
Table 4-6. TCLP Results for Color TV 4-12
Table 4-7. TCLP Results for VCRs 4-13
Table 4-8. TCLP Results for Cell Phones 4-16
Hi
-------
Table 4-9. TCLP Results for Keyboards 4-19
Table 4-10. TCLP Results for Mice 4-21
Table 4-11. TCLP Results for Remote Controls 4-22
Table 4-13. TCLP Results for Flat Panel Displays 4-26
Table 5-1. Summary of TCLP Pb Leaching Results 5-1
Table A-l. Additional CPU Metals Analysis 6-1
iv
-------
List of Figures
Figure 2-1. Average CPU Composition of 29 Computer CPUs by Weight 2-2
Figure 2-2. Effects of Component Mixture on Lead Leachability 2-8
Figure 2-3. Dissolved Oxygen and ORP Results from Head Space Impact Study 2-11
Figure 2-4. Lead Results from Head Space Impact Study 2-12
Figure 3-1. Impact of Rotation Speed Results 3-5
Figure 3-2. Comparison of Metals Results from TCLP Time Study Experiments 3-7
Figure 3-3. Sample 2 Filtered vs. Nonfiltered Metals Concentrations 3-9
Figure 3-4. Metal Concentrations from Methodology Comparison for CPU #1 3-12
Figure 3-5. Laboratory Measurements from Methodology Comparison for CPU #1 3-13
Figure 4-1. Average CPU Composition of 29 Computer CPUs by Weight 4-2
Figure 4-2. Average Composition of Computer Monitors Tested 4-5
Figure 4-3. Average Composition of Laptops Tested 4-7
Figure 4-5. Average Composition of Color TVs Tested 4-11
Figure 4-6. Average Composition of VCRs Tested 4-13
Figure 4-7. Average Composition of Cell Phones Tested 4-15
Figure 4-8. Average Composition of Keyboards Tested 4-18
Figure 4-9. Average Composition of Mice Tested 4-20
V
-------
Figure 4-10. Average Composition of Remote Controls Tested 4-22
Figure 4-11. Average Composition of Smoke Detectors Tested 4-23
Figure 4-12. Average composition of Flat Panels 4-25
V/'
-------
List of Abbreviations
CPU - Computer Processing Unit
CRT - Cathode Ray Tube
DO - Dissolved Oxygen
HDPE - High Density Polyethylene
MSW - Municipal Solid Waste
PWB - Printed Wire Board
RCRA - Resource Conservation and Recovery Act
RMV - Relative Millivolts
TC - Toxicity Characteristic
TCLP - Toxicity Characteristic Leaching Procedure
US EPA - United States Environmental Protection Agency
WTE - Waste to Energy
vii
-------
Executive Summary
Research was conducted to examine whether discarded electronic devices are likely to
meet the regulatory definition of toxicity characteristic (TC) hazardous waste under provisions of
the Resource Conservation and Recovery Act (RCRA). The test used to determine whether a
solid waste is a TC hazardous waste is the toxicity characteristic leaching procedure (TCLP). If
the leachate produced using the TCLP contains certain elements at concentrations above a
regulatory threshold, the solid waste is a TC hazardous waste (unless otherwise excluded).
Discarded electronic devices have the potential to be TC hazardous wastes because they may
contain elements such as lead, cadmium and mercury. Previous research by the authors found
that color cathode rays tubes (CRTs, from computer monitors and television sets) in most cases
leached enough lead using the TCLP to be TC hazardous wastes. The research presented in this
report examines other devices such as computer central processing units (CPUs,) cellular phones,
and computer peripherals (e.g., keyboards, mice).
Prior to performing the TCLP on a variety of electronic devices, research was first
conducted to examine the factors that impact TCLP results on electronic devices and to evaluate
the use of modified TCLP methods. Because electronic devices are large, bulky, heterogeneous
with respect location of toxic elements, and difficult to size reduce, the researchers felt it
necessary to utilize methods that met the intent of the TCLP but that allowed more rapid and
representative testing of this waste stream. Computer CPUs were the device used as part of
these initial examinations. Lead was the element found to most likely result in a computer CPU
leaching lead above the 5 mg/L TC limit. The presence of ferrous metal was determined to have
a major impact on the concentration of lead measured in the TCLP leachate. The dissolution of
iron and zinc from the ferrous components of a computer CPU created electrochemical
conditions in the leaching solution that acted to reduce the amount of lead present in solution.
A modified TCLP where large devices were disassembled and leached in entirety (or near
entirety) was found to be an effective means of characterizing large, bulky and heterogeneous
samples without extensive and prohibitive sample preparation. Testing conditions remained the
same as the TCLP with the exception of rotation speed and sample size reduction. The large-
scale modified TCLP resulted in greater lead concentrations from computer CPUs relative than
the standard TCLP. The hypothesis for this occurrence was that the size-reduced ferrous metal
in the standard TCLP suppressed lead leaching in the same fashion as observed in the study
described above. The modified TCLP was considered appropriate for further characterization of
electronic devices because 1) it allowed a more representative sample to be tested, 2) it permitted
characterization of devices that would have otherwise been infeasible without elaborate size-
reduction devices, and 3) the devices tested more closely represented the conditions they exist as
when disposed in a landfill.
Twelve different types of electronic devices were tested using either the standard TCLP
or modified versions of the TCLP. In many cases, lead concentrations in the TCLP leachates
exceeded the 5 mg/L Toxicity Characteristic (TC). Table ES-1 presents the results in terms of
-------
the number of devices of a particular device type that exceeded the lead TC threshold
concentration. Every device type leached lead above 5 mg/L in at least one test. Most devices
leached lead above the TC limit in a majority of cases. No other elements were found to exceed
their respective TC limits. The impact of ferrous metal content was evident. Computer CPUs
Table ES-1. Summary of TCLP Pb Leaching Results
Device
Standard TCLP1
Modified Large-Scale
TCLP2
Modified Small-Scale
TCLP3
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
CPUs -
Manual Size
Reduction
0
12
21
41
CPUs-
Mechanical
Size Red.
1
11
Computer
Monitors
9
9
Laptop
Computers
6
6
15
15
Printers
5
9
Color
Televisions
6
6
VCRs
9
10
Cellular
Phones
33
43
15
20
Key Boards
0
3
1
1
Computer
Mice
15
15
Remote
Controls
4
4
6
6
Smoke
Detectors
7
9
Flat Panel
Monitors
3
8
Notes: 1 100-g sample size-reduced by hand (unless otherwise noted)
2 Disassembled device leached in entirety (or near entirety)
3 100-g diassembled sample (not size-reduced)
ix
-------
tended not to leach lead above the TC limit using the standard approach, while they leached
above the TC limit just over 50% of the time using the large-scale modified TCLP. Laptop
computers on the other hand leached lead above the TC limit in all cases using both approaches.
The difference between the two devices was attributed to the greater ferrous metal content of the
computer CPUs (68%) relative to the laptops (7%). Smaller devices that contained larger
amounts of plastic and smaller amounts of ferrous metal (e.g., cellular phones, remote controls)
tended to leach lead above the TC limit at a greater frequency than devices with more ferrous
metal (e.g., printers).
Because of the large and expanding universe of electronic devices, no single study can
hope to characterize every device type, let alone every make and model. The results presented
here do, however, provide sufficient evidence that discarded electronic devices that contain a
color CRT or printer wiring boards with lead-bearing solder have a potential to be RCRA TC
hazardous wastes for lead (unless otherwise excluded) and that generators should assume such
devices are hazardous or should conduct specific testing to determine otherwise.
-------
1 Introduction
1.1 The Management of Discarded Electronics
The management of discarded electronic devices (often referred to as E-Waste) has been
raised as an issue of concern for the solid waste community. Devices such as televisions,
computers (and peripherals), and cellular phones are known to contain small amounts of
chemicals that can exert, upon exposure, negative impacts on human health and the environment.
The toxic chemicals electronics often contain raise questions about their impact on the
environment and their status as Resource Conservation and Recovery Act (RCRA) Toxicity
Characteristic (TC) hazardous wastes. (BAN, 2002; FWI, 2001, GFF, 2001; SVTC et al. 2001).
Lead, for example, may be found in most electronic devices, either in the cathode ray tube
(CRT) or in the printed wire board (PWB, i.e. the circuit board). Reports document that
approximately 6.3% of a typical computer is composed of lead, a majority of which is attributed
to the CRT (MCC, 1996). One recent study showed that color CRTs from televisions and
computer monitors usually exceeded the RCRA TC limits for lead (Musson et al., 2000;
Townsend et al., 1999). A recent Federal Register notice proposed to exclude CRTs and CRT
glass sent for recycling from the definition of a solid waste, thus simplifying RCRA management
requirements (USEPA, 2002b). In this notice, the USEPA stated, "we are not currently aware of
any non-CRT computer components or electronic products that would generally be hazardous
waste." Lead is also found in other components of a computer such as PWBs. Tin/lead solder
(63% tin and 37% lead) is the most common solder alloy used in electronics (NCM, 1995).
While some studies have reported TCLP lead concentrations from PWBs to exceed the 5 mg/L
TC limit (Environment Australia, 1999; Yang, 1993), shredded and whole PWBs that are
reclaimed are excluded from the definition of solid waste (USEPA, 1997; USEPA, 1998).
Although tests performed on PWBs indicate that electronic devices such as computer CPUs have
the potential to fail the TCLP, there are no reports that assess the TC of an entire computer CPU.
The growing consumer demand for these products and the tendency for the "in-service life-
time" of many devices to be short (often just a few years) have resulted in greater amounts of
discarded E-Waste requiring management. According to a study prepared for the US EPA, an
estimated 3 million tons of E-Waste were disposed in landfills in the US in 1997 (GFF, 2001).
Industry experts have projected that more than 20 million personal computers became obsolete
in 1998, and that more than 60 million personal computers will be retired in 2005 (National
Safety Council, 1999). Personal computers, which had an average lifetime of 4 to 5 years in
1992, are projected to have an average lifetime of 2 years by 2005 (NSC, 1999). The National
Safety Council reports that approximately 500 million computers will become obsolete between
1997 and 2007 (NSC, 1999).
CHAPTER 1
1-1
INTRODUCTION
-------
Chapter 1 Introduction
Some have speculated that the disposal of discarded electronic devices with the rest of the
municipal solid waste (MSW) stream will result in measurable impacts on human health and the
environment (Yang, 1993; BAN, 2002; Schmidt 2002), or at the very least may pose operational
or permit problems for waste facilities (e.g. elevated metal concentrations in waste-to-energy
(WTE) ash, thereby limiting ash reuse options). Others argue that no evidence exists that
management of E-waste via traditional waste management systems such as landfills causes any
impact on the environment (Akatiff, 2002). The debate over the true effects of discarded
computers, televisions, and cellular phones remains alive and active. Those charged with
managing solid waste are faced with a very real question: are discarded electronic devices
classified as hazardous wastes under RCRA regulations developed by the United States
Environmental Protection Agency (US EPA)?
1.2 The Need for Toxicity Characteristic Testing
A cornerstone of the federal solid waste regulations in the US is that certain types of solid
wastes possess physical and/or chemical properties that warrant additional management
requirements beyond that required for other household or commercial wastes. Solid wastes
identified as hazardous (either through listing or by expressing a characteristic) are subjected to a
greater degree of control, from the point of generation to the point of final disposition. The added
regulatory requirements for hazardous wastes result in greater management costs. In some cases,
these added costs provide incentives for generators of the waste to alter their process or product
in such a fashion that the wastes produced are no longer classified as hazardous. Also, some
types of solid wastes are excluded from being hazardous wastes at the federal level; household
waste is a notable example since substantial electronic device wastes are generated by
individuals at home.
Discarded electronic devices are not listed as hazardous wastes. The known presence of
several heavy metals in E-waste raises the potential that these devices might be toxicity
characteristic (TC) hazardous wastes. The test for determining whether a solid waste is
hazardous because of the TC (40 CFR 261.34) is the Toxicity Characteristic Leaching Procedure
(TCLP). The EPA publication SW-846, entitled Test Methods for Evaluating Solid Waste,
Physical/Chemical Methods, is the EPA's official compendium of analytical and sampling
methods that have been evaluated and approved for use in complying with the RCRA
regulations. SW-846 Method 1311, the TCLP, is an 18-hour batch-leaching test in which 100
grams of solid waste are leached in the presence of a prescribed leaching fluid designed to
simulate the conditions that might occur in a MSW landfill as the waste decomposes. The TCLP
fluid used depends on the alkalinity of the waste material. Very alkaline waste materials are
leached with a fixed amount of glacial acetic acid without an added buffer (pH 2.88 ± 0.05),
while other waste materials are leached with glacial acetic acid buffered at pH 4.93 ± 0.05 with
1-N sodium hydroxide. Particle size reduction of solid waste is required so that the waste
material is capable of passing through a 9.5-mm standard sieve. Two liters of the leaching fluid
are placed in a container with the lOOg sample. The container is placed in a rotary extraction
vessel and leached for 18 ± 2 hours at 30 ± 2 rpm. After rotation, the leachate is filtered and then
CHAPTER 1
1-2
INTRODUCTION
-------
Chapter 1 Introduction
analyzed for a number of chemicals. Eight heavy metals are on the TC list. They are presented
along with their appropriate TC concentrations in Table 1-1. If the leachate concentration
exceeds the TC concentration, the solid waste is classified as a TC hazardous waste.
Table 1-1. TC Concentrations for Heavy Metals
MI-TAI.
tc m:(;i i.a i i:d
CONCENTRATION
Arsenic
5 mg/1
Barium
100mg/l
Cadmium
1 mg/1
Chromium
5 mg/1
Lead
5 mg/1
Mercury
0.2 mg/1
Selenium
1 mg/1
Silver
5 mg/1
Several of the elements listed in Table 1-1 are found in electronic devices. Most notable are
lead, cadmium, chromium and mercury. If these elements are present in an electronic device, the
device could potentially be classified as a TC hazardous waste when discarded. TCLP results on
manufactured articles like electronic devices are not common, however. This stems from several
factors, including the relatively recent recognition of electronic devices as a possible concern,
and the difficulty of conducting the TCLP on such devices.
While the TCLP is performed routinely in laboratories across the country, the analysis of
wastes like electronic devices poses several problems. The TCLP methodology requires 100 g of
sample sized-reduced to pass a 0.95-cm sieve. A whole device such as a computer must
therefore be ground, shred or cut to the appropriate particle size. Unlike most wastes that either
inherently meet the size requirement (e.g., ash, sludge) or exist in a homogenous physical state
that can be crushed or cut (e.g., cement-stabilized forms, wood), electronic devices do not lend
themselves to ready size reduction. Electronic devices are composed of different material types
each with distinct physical properties (e.g., printed wire boards, plastic, ferrous and non-ferrous
metal, glass). Each of these materials may respond differently to different size reduction
equipment. Common laboratory grinders or mills that are sometimes used in the preparation of
samples for the TCLP are of little use for size reducing something on the scale and heterogeneity
of a printer.
Larger-scale equipment (e.g. shear shredders) that is sometimes used in scrap electronic
debris-processing facilities does not size reduce the material to the necessary size, and is
designed for processing large amounts of material (as opposed to a single device). Without the
CHAPTER 1
1-3
INTRODUCTION
-------
Chapter 1 Introduction
acquisition of specialized equipment or the use of multiple size reduction units, those performing
the test are often left with manual size reduction (e.g., cutting by hand) as the only reasonable
option. In addition to being very labor intensive (and very difficult for metal housings and
components such as hard drives), manual reduction methods may introduce human bias into the
sample collection process as the technician in charge of producing the sample may have to select
which part of a large device should be size-reduced.
One possible approach to determine the TCLP results for a heterogeneous manufactured
device is to use TCLP results from the suspected hazardous component of the device and assume
that the rest of the device does not impact leaching. The first is to calculate a predicted TCLP
concentration for the entire device based on the TCLP results from the hazardous components
only. For example, the suspected hazardous component in a computer CPU is the PWB;
therefore, the TCLP can be performed on the PWB only. A predicted TCLP lead concentration
for the entire CPU can then be calculated based upon the fraction of PWB in the CPU and the
results from the TCLP performed on the PWB. The equation that to calculate a predicted TCLP
concentration for a computer CPU is:
TCT P M
ypy p _ pwB PWB
The TCLPpwb term represents the TCLP lead concentration from the PWB, MPWB is the mass
of PWB in the CPU, and Mtotal is the total mass of the CPU. As an example, research showed
that the PWB comprised 15.8% of the weight of a typical CPU. Therefore this method would
require multiplying the results of the lOOg PWB sample by 0.158 to determine the TCLP results
expected from the CPU.
A second method of predicting the TC of an electronic device is to perform the TCLP on the
relative fraction of the suspected hazardous component only. For example, research showed that
the PWB comprised 15.8% of the weight of a typical CPU. Performance of the TCLP on a
sample of only 15.8 grams of the PWB could predict the lead leachability from a CPU.
However, each of these methods neglects the potential chemical and physical effects of other
components of the CPU in the TCLP
The magnitude of electronic devices in the waste stream, their potential to be a TC hazardous
waste, and the lack of TCLP data for these devices, provide ample incentive for a RCRA TC
study of electronic devices. The objective of the research presented in this report was to
examine the propensity of electronic devices to be characterized as TC hazardous wastes under
RCRA. In some cases, the difficulties encountered in performing the TCLP on large devices
prompted the researchers to conduct modified leaching tests that were designed to meet the
intent of the TCLP testing. As such, some work was conducted to examine the impacts of these
modifications on testing results. The goal of this research was not to characterize all electronic
devices for TC status - that would be unrealistic because of the great variety of devices sold and
CHAPTER 1
1-4
INTRODUCTION
-------
Chapter 1 Introduction
in use. Instead, the goal was to provide the regulators and the regulated community with some
understanding of the likelihood of such devices being TC hazardous wastes.
1.3 Research Scope and Report Description
This report presents the results of TCLP testing of electronic devices. This research was
funded by US EPA Regions 4 and 5 to assist state regulators in making hazardous waste
determinations. The research started with a focus on computer CPUs, but evolved to examine
other devices as well. Much of the research was also devoted to evaluating the parameters that
impact TCLP leaching results from discarded electronic devices and to examine modified
leaching procedures. Modified leaching procedures were designed in an effort to meet the intent
of the TCLP while overcoming some of the obstacles that greatly limited the number of tests that
could be performed. This report does not attempt to interpret the TCLP results with respect to
environmental impacts.
Chapter 2 presents the result of several experiments conducted to examine the factors that
impact lead leaching from computer CPUs during the TCLP. Specifically, Chapter 2 presents
the following test results:
- TCLP testing of PWBs
TCLP tests of a mixture representative of CPU composition
TCLP tests of PWBs with each CPU component individually to determine if each
component can affect TCLP lead leachability from PWBs
TCLP tests of PWBs with varying quantities of the iron metal components to further
characterize its effects on lead leachability
TCLP tests of a representative CPU sample with varying head space above the leaching
solution to determine the effects of head space upon the TCLP applied to electronics.
Finally, Chapter 2 concludes with a comparison of the toxic metal levels contained in the filtered
TCLP solution to those present in the solution prior to filtering (unfiltered).
Chapter 3 presents the results of research utilizing a new large-scale modified TCLP for
leaching whole devices such as computer CPUs. This method was performed to provide a
potential alternative method to performing the TCLP on electronics and other devices not
amenable to size reduction. By testing the whole device, the errors and difficulties discussed
earlier may be avoided. However in designing the new method several parameters required
CHAPTER 1
1-5
INTRODUCTION
-------
Chapter 1 Introduction
testing:
- Does the speed of rotation of the device impact TCLP results?
- Does the large-scale method reach the equilibrium level intended in the TCLP
methodology over the same period of time, 18+2 hours?
- What results are achieved by placing smaller devices such as computer mice and remote
controls into TCLP extraction vessels not size reduced, but only disassembled?
Chapters 2 and 3 were conducted and presented as the major part of a master of engineering
thesis (Vann, 2003).
Chapter 4 presents all of the TCLP results for the 12 types of electronic devices
examined in this research. The devices examined included:
Personal Computer CPUs Cellular Phones
Computer Monitors Key Boards
Laptop Computers Computer Mice
Printers Remote Controls
Color Televisions Smoke Detectors
VCRs Flat Panel Monitors
The average composition of each of the devices is presented, as well as the concentrations of
several elements detected within the TCLP leachates. The results are summarized and compared
in Chapter 5.
CHAPTER 1
1-6
INTRODUCTION
-------
2 Factors Affecting Toxicity Characteristic
Leaching Procedure Lead Leachability from
Computer CPUs
2.1 Introduction
Research was performed to examine the factors that affect lead leachability from computer
CPUs. Computer CPUs consist of several distinct materials with different physical and chemical
properties. An understanding of the processes affecting heavy metal leaching is useful for
developing the appropriate testing procedures and for understanding how results may differ
among devices. Slight changes in the TCLP methodology, which still meet the requirements of
the method, can also affect results. For example, Meng et al. (2001) found that changing the
head space above the TCLP leaching fluid during the test greatly impacted the amount of arsenic
that leached from water treatment sludges.
This chapter provides the details of an experiment that sampled, size-reduced, and mixed
several personal computer CPUs to create a composite CPU mixture that was leached under a
variety of conditions. Lead, iron, copper, and zinc results are presented. Although lead is the
primary chemical of interest, the analysis of iron, copper, and zinc can be used to describe the
processes (reduction by metallic iron and zinc and sorption by hydrous ferric oxide) that have
been documented to impact lead leachability (Kendall, 2003). The objective of the research was
to investigate the factors that impact lead leachability from computer CPUs during the TCLP, not
to provide a definitive study on the hazardous waste characterization of computer CPUs.
2.2 Materials and Methods
2.2.1 Sample Collection and Processing
Computer CPUs were collected from a variety of sources including electronics
demanufacturing facilities, individuals, and a local household hazardous waste collection center.
Twenty-nine computer CPUs were completely disassembled and separated into five material
categories: PWBs, ferrous metals, nonferrous metals, wires/cables, and plastics. The weight of
material for each of these categories was measured. Figure 2-1 presents the average composition
(by weight) of all the CPUs.
From the twenty-nine CPUs, five were selected at random to create a synthetic CPU mixture
to be used in TCLP testing. The synthetic CPU mixture was created to match the average
composition depicted in Figure 2-1. To provide enough material for all the desired tests,
approximately 500 g of PWB, 240 g of plastic, 2,200 g of ferrous metal, 170 g of nonferrous
metal, and 100 g of wires were collected from each of the five CPUs. These pieces were
CHAPTER 2
2-1
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
selected at random from each CPU, and the collected materials were size-reduced by hand (i.e.
using shears) so they would pass a 0.95-cm sieve (a requirement of the TCLP). Materials in the
same category from each CPU were combined and mixed. The "synthetic" CPU samples were
prepared by mixing representative subsamples of each material type. Since the total weight
required for a TCLP is 100 g, the "synthetic" CPU mix was composed of 15.8 g of PWB, 7.5 g
of plastic, 68.2 g of ferrous metal, 5.4 g of nonferrous metal, and 3.1 g of wire/cable.
2.2.2 Leaching and Analysis Methods
The TCLP, USEPA Method 1311, is the USEPA prescribed test for determining whether a
solid waste is a TC hazardous waste (USEPA, 1996). The TCLP, in this study, was performed
by manually size reducing (i.e., hand cutting) the CPU components using shears so they would
pass a 0.95 cm sieve. One hundred gram samples of the size-reduced materials were placed into
2L TCLP extraction vessels composed of high-density polyethylene (HDPE). Two liters of
TCLP extraction fluid #1, which consists of 11.4 mL of glacial acetic acid and 128.6 mL of 1 N
sodium hydroxide solution diluted to 2 L with reagent water, were added to the extraction vessel.
The initial pH of the TCLP extraction fluid was 4.93±0.05. Initial measurements of the pH,
oxidation-reduction potential (ORP), and dissolved oxygen (DO) were recorded. All pH and
ORP measurements were made using an Orion Model 710A+ benchtop meter equipped with an
Orion Model 91-55 combination pH electrode and an Orion Model 91-79 ORP platinum triode.
The pH probe and meter were calibrated with standard buffer solutions (4.0, 7.0, and 10.0) with a
three-point calibration. The ORP probe and meter were calibrated using a reference standard
(475 mV) in the relative millivolt (RMV) mode and all measurements were in RMV. Dissolved
Figure 2-1. Average CPU Composition of 29 Computer CPUs by Weight
Nonferrous Wires
Ferrous Metal
68.2%
Plastic
7.5%
CHAPTER 2
2-2
FACTORS .AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABIIITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
oxygen measurements were collected using an YSI Inc. Model 55 handheld dissolved oxygen
meter.
All samples were performed in triplicate and a TCLP blank was included for each set of
leaching extractions. The samples were rotated at 30±2 rpm for 18±2 hours in a 12-vessel rotary
extractor (Analytical Testing Corporation). After rotation, the final pH, DO, and ORP of the
leachates were recorded. The TCLP leachates were filtered through a glass fiber filter (0.7 [j,m
pore size) using pressure filtration and preserved by adding concentrated nitric acid until the pH
of the filtrate was below 2. In addition to collecting the filtered leachate, samples of nonfiltered
leachates were also collected and preserved. All samples were then stored in HDPE bottles until
acid digestion. Specific experimental details of the TCLP methodologies performed will be
described in the following sections.
Ferrous iron (Fe2+) analysis was performed prior to preserving the samples using a HACH
Model DR/4000 spectrophotometer using the 1,10 phenanthroline method (HACH program
2150). The spectrophotometer was zeroed with reagent water and a 1-mg/L standard was used to
verify the machine calibration. Samples placed in glass vials were allowed to react for three
minutes with ferrous iron reagent powder pillows added to the sample. The glass vials were
cleaned to remove any surface contamination and then placed in the spectrophotometer. The
ferrous iron (Fe2+) concentration was recorded. Readings were taken before and after the
addition of the ferrous iron reagent powder pillows for the nonfiltered samples to account for any
absorbance caused by particulates in the sample. Lead, iron, copper, and zinc analyses were
performed by digesting the samples using the hotplate acid digestion procedure, USEPA Method
301 OA. The digested samples were then analyzed using USEPA Method 601 OB (Inductively
Coupled Plasma-Atomic Emissions Spectrometry) on a Thermo Terrell Ash Trace Analyzer ICP
(USEPA, 1996).
2.2.3 TCLP on Printed Wire Boards
The source of lead in CPUs is the lead-based solder on the PWBs in the CPU. A baseline
measurement of lead from isolated PWBs was necessary to determine the effects of other CPU
materials on lead leachabiiity. The TCLP was performed on the PWBs. One hundred grams of
the size-reduced PWBs were leached in triplicate. PWBs comprised an average of 15.8% of the
weight of a CPU such that in a lOOg CPU sample, 15.8g of PWB would be expected. In addition
to the 100 g samples of PWBs, 15.8 g PWB samples in 2 L of leaching solution would be used to
predict the leachate concentration of lead attributed to PWBs for a CPU. However, this violated
the prescribed 20:1 liquid:solid ratio prescribed by the TCLP. The TCLP was performed on 70
g, 30 g, and 15.8 g PCB samples to examine the effects of a change in the liquid to solid ratio,
provide a set of data for potential samples with varying PWB content, and predict the CPU
leachate lead concentration. The same volume of leaching fluid, 2 L, was used in each case.
Thus the liquid-to-solid ratio of the smaller masses leached was greater than the standard 20:1
for TCLP.
CHAPTER 2
2-3
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
2.2.4 TCLP on Synthetic Computer CPU Mix/Particle Size Study
The TCLP was also performed on a 100-g sample of the "synthetic" CPU mixture prepared
in Section 2.2.1 to evaluate lead leachabiiity from a complete CPU. It is noted that the TCLP
does not limit the extent of size reduction of the solid sample. To address the issue of how size
reduction affects lead leachabiiity from CPUs, a second sample of the "synthetic" CPU mixture
was collected and size reduced to pass through a 0.2-cm sieve. This was achieved manually
using shears. The TCLP was then performed to evaluate the impact of sample size on lead
leachabiiity.
2.2.5 Component Impact
An investigation of the effects of computer CPU composition on lead leachabiiity during the
TCLP was performed by leaching different mixtures of the CPU components. Samples were
prepared with 50g of size-reduced PWBs only and 50g of PWBs with 50g of ferrous metal, 50 g
of nonferrous metal, and 50g plastic to examine the individual impact of each material on lead
leachabiiity from the PWBs. The materials were obtained from the CPU components described
in Section 2.2.1.
The impact of CPU composition was also evaluated by conducting TCLPs with a constant
PWB quantity, but with varying fractions of ferrous metal. In addition to the "synthetic" CPU
mixture (68.2% ferrous metal), three other samples were prepared with 40%, 20%, and 0%
ferrous metal. The mass of plastic was increased for the three samples to replace the decreased
mass of ferrous metal such that the total mass of each sample was maintained at 100 g. The
sample masses are presented in Table 2-1.
Table 2-1. Ferrous Metal Impact Sample Compositions
"A> I-'citoiis
Metsil
PWB
(a)
Phistic
(a)
I-'citoiis
Mclal
(a)
NoiiI'citoiis
Mclal
(a)
\\ ircs/Cablcs
(a)
Total
Mass
(a)
68.2
15.8
7.5
68.2
5.4
3.1
100
40
15.8
35.7
40
5.4
3.1
100
20
15.8
55.7
20
5.4
3.1
100
0
15.8
75.7
0
5.4
3.1
100
CHAPTER 2
2-4
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
2.2.6 Impact of Head Space
As discussed in Section 2.1, slight changes in the TCLP methodology can affect results. An
investigation was conducted to evaluate whether the headspace above the TCLP leaching fluid
impacted lead leachabiiity from computer CPUs during the TCLP. The extraction vessel used
during the standard TCLP method had an actual volume of 2.34 L, leaving approximately 0.34 L
of headspace above the 2L of TCLP extraction fluid. The volume occupied by the 100-g CPU
sample itself was 0.023 L. The actual volume of headspace (Va) and extraction fluid (VI) in the
vessel was 0.317 L and 2.023 L, respectively, resulting in an air-to-liquid ratio (volume of air to
volume of liquid (Va/Vl)) of 0.16.
In order to determine if the vessel headspace significantly affected lead leachabiiity, three
additional samples were evaluated in this study with headspace ratios (Va/Vl) of approximately
0, 0.5, and 1. It is noted that the zero headspace samples were not truly free of air due to a small
amount of air leakage into the vessel. The sample materials were obtained from the CPU
components prepared in Section 2.2.1. The liquid-to-solid ratio was maintained at 20:1. The
relative material fractions remained identical to the synthetic CPU sample. The sample masses
were adjusted (considering the volume occupied by the sample itself) to achieve the appropriate
air-to-liquid ratios. The sample compositions and volumes of the TCLP extraction fluid are
presented in Table 2-2.
Table 2-2. Impact of Vessel Headspace Sample Composition
Sample
I'WU
(Si)
ferrous
Metsil
(a)
Noil ferrous
Melsil
(a)
Plastic
(a)
Wires
(a)
Total Msiss
(a)
Solution
(ml)
Va/Vl~0
18.3
79.0
6.3
8.7
3.6
115.8
2315
Va/Vl=0.16
15.8
68.2
5.4
7.5
3.1
100.0
2000
Va/Vl=0.5
12.3
53.2
4.2
5.9
2.4
78.0
1560
Va/Vl=l
9.2
39.9
3.2
4.4
1.8
58.5
1160
CHAPTER 2
2-5
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
2.3 Results and Discussion
2.3.1 Lead Leachabiiity from PWBs
The average TCLP lead concentration from the lOOg samples of PWBs was 151 mg/L. Other
studies have reported TCLP lead concentrations for PWBs to range from 56 mg/L to 1,350 mg/L
(Environment Australia, 1999; Yang, 1993). Typical PWB TCLP lead concentrations measured
in other studies range from 100 mg/L to 200 mg/L (Jang, Y. and T. Townsend, 2003 "Leaching
of Lead from Computer Printed Wire Boards and Cathode Ray Tubes by Municipal Solid Waste
Landfill Leachates," Submitted for publication; Townsend et al., 2001). All of the PWB samples
tested in this study exceeded the RCRA TC limit for lead of 5.0 mg/L. However, the USEPA
excludes PWBs that are being recycled from the definition of a solid waste; they are exempt
from the RCRA hazardous waste regulations.
To determine the expected TCLP leachate lead concentration of a whole CPU, which may
contain varied amounts of PWBs, TCLP testing was performed on PWB samples of 70g, 30g,
and 15.8g. The 15.8 g sample was chosen to correspond with the average PWB content
measured in the CPUs collected. The leaching fluid volume was maintained at 2L for all
samples. The leachate lead concentration increased with an increase in sample mass. The
maximum lead concentration was reached at the 70-g sample. The lead concentration from the
15.8-g PWB sample of 39+7 mg/L provided a prediction for the lead leachabiiity expected from
the "synthetic" CPU mixtures. The lead, iron, copper, and zinc results from the TCLPs that were
performed on the PWBs are presented in Table 2-3.
The final pH did not vary greatly among the samples and did not change greatly from the
initial pH (4.93±0.05). There was no significant difference (t-test, a=0.5 using Microsoft Excel)
in the lead, iron, copper, and zinc concentrations between the 70-g and 100-g samples. The
averages and standard deviations of the iron and zinc concentrations measured in the 30-g and
100-g samples, respectively, were impacted by one sample (an outlier) out of the three that were
analyzed. This caused the standard deviation to be higher than the average concentration of the
three samples. If the outlier is removed and only two samples are used, the average iron
concentration in the 30-g sample is 1 mg/L with a standard deviation of 0.04 mg/L while the
average zinc concentration in the 100-g sample is 0.3 mg/L with a standard deviation of 0.03
mg/L.
CHAPTER 2
2-6
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
Table 2-3. Average TCLP Leachate Concentrations from PWBs
(Average of 3 samples)
M:iss
Liquid/Solid
1 iiiiil
1 .(."ill
Iron (m»/l.)
Copper
Zinc
(a)
Ksilio
pll
(mii/l.)
(mii/l.)
(m»/l.)
15.8
126.6:1
4.91±0.01
39±7
1±0.5
0.3±0.1
0.14±0.1
30
66.7:1
4.92±0.01
61±3
3±4
0.5±0.1
0.2±0.05
70
28.6:1
4.93±0.01
161±6
8±2
0.7±0.2
0.3±0.18
100
20:1
4.91±0.01
151±10
5±0.6
0.5±0.2
3±4
2.3.2 Predicting the TC of an Electronic Device
With the results from Table 2-3, it is possible to predict the TCLP lead concentration from
synthetic CPU samples. As discussed in Chapter 1, two approaches are possible. The first is to
calculate a predicted TCLP concentration for the entire CPU based on the TCLP results from the
PWBs alone using the equation:
TPI p M
ypy p _ 1^1PWB PWB
Using this procedure, the TCLP lead concentration from the PWBs of 151 mg/L, and PWBs
composing 15.8% (MPWB/MTotal=0.158) of the CPU, results in a predicted TCLP lead
concentration of 24 mg/L, which exceeds the TC limit for lead.
The second method to predict the TC of an electronic device is to perform the TCLP on the
relative fraction of the suspected hazardous component only. The results showed that the TCLP
lead concentration from the 15.8-g PWB samples was 39 mg/L. The predicted lead leachabiiity
from the "synthetic" CPU mixture during the TCLP would be expected to be 39 mg/L, assuming
the other components did not impact lead leaching.
2.3.3 Synthetic Computer CPU Mix
The TCLP was performed on a "synthetic" CPU, a mixture of CPU components
representative of the average composition of the CPUs collected. The TCLP leachates from the
synthetic CPU mixture contained an average of 0.3mg/L lead, 19 mg/L iron, and 136 mg/L zinc.
Copper was not detected (MDL=0.05 mg/L). The results are presented in Table 2-4.
CHAPTER 2
2-7
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
As dicussed in section 2.3.2, tests of samples containing only PWBs predicted lead levels of
39 mg/L and 24 mg/L. The 0.3 mg/L of lead leached from the "synthetic" CPU mixture sample
indicate that other components of the CPU affected lead leachabiiity from the PWBs. Figure 2-2
illustrates a comparison of the average lead concentration from the "synthetic" CPU mixture
samples and the predicted TCLP lead concentrations based on PWB testing only. This
confirmed that there was an impact on lead leachabiiity during the TCLP when the other
components of the CPU were mixed with the PWBs. This comparison also demonstrated that
testing the PWBs alone did not accurately predict the lead leachabiiity from a representative
sample of the "synthetic" CPU mixture. A Student's t-test (a=0.5 using Microsoft Excel)
performed on the results shows that the two methods of predicting the TCLP lead concentration
were significantly different for this sample. Predicting the TC of an entire device by testing the
hazardous component only may not be the most reliable option.
100
I
8
i
D
§
u
£
10
0.1
0.01
TCLP <0.95 cm TCLP <0.20 cm
Predicted TCLP
(Method 1)
Predicted TCLP
(Method 2)
Figure 2-2. Effects of Component Mixture on Lead Leachabiiity
2.3.4 Component Impact
Table 2-4 presents a summary of an investigation of material impact on lead leachabiiity
during the TCLP. The TCLP lead, iron, copper, and zinc concentrations from samples
comprised solely of 50g PWB averaged 83 mg/L, 6 mg/L, 0.3 mg/L, and 1 mg/L, respectively.
The lead concentration in the sample that contained 50g of PWB mixed with 50g of ferrous
metal was significantly less (Student's t-test, a=0.5 using Microsoft Excel) (3 mg/L vs. 83
mg/L). This indicated that lead leachabiiity was affected by the ferrous metal component during
the TCLP. Other studies have also documented that iron impacts lead leachabiiity during the
TCLP. Kendall (2003) reported that adding iron metal shavings to brass foundry casting sand
CHAPTER 2 2-8 FACTORS .AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABIIITY
FORhl COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
significantly decreased lead leachabiiity during the TCLP. This explains the prior observations
that lead leachabiiity from the "synthetic" CPU mixture was significantly lower than the results
predicted from the 15.8-g PWB sample and the calculated weighted average TCLP
concentration. Lead leachabiiity was not greatly impacted by the addition of nonferrous metal.
The lead concentration measured in the TCLP leachate was significantly higher (Student's t-test,
a=0.05 using Microsoft Excel) in the sample that contained PWB and plastic. Copper
leachabiiity was also decreased by the addition of ferrous metal. In addition, the final pH of the
PWB/Ferrous metal samples did not differ greatly from the samples of PWB alone. However, the
pH was higher in the samples that contained nonferrous metal and plastic.
Table 2-4. Component Impact TCLP Results
Siimplc C omposition
1 -CiUl
(mji/l.)
1 ron
(mii/l.)
Copper
(mii/l.)
Zinc
(mii/l.)
pll
Material Impact
50g PWB
83±5
6±4
0.3±0.02
1±2
4.74±0.01
50g PWB/50g Ferrous
3±2
21±9
<0.05
110±3
4.78±0.01
50g PWB /50g
Nonferrous
69±8
10±7
<0.05
0.2±0.3
5.10±0.07
50g PWB/50g Plastic
113±6
8±7
1±0.4
0.3±0.1
5.01±0.01
Ferrous Impact
0% Ferrous
44±5
5±9
2±0.2
1±2
4.87±0.02
20% Ferrous
7±2
69±7
BDL
57±4
4.97±0.002
40% Ferrous
2±1
40±9
0.1±0.01
101±10
4.99±0.02
68.2% Ferrous
0.3±0.
2
19±9
BDL
136±24
5.16±0.01
A further investigation of ferrous metal impact on lead leachabiiity during the TCLP was
conducted. The results of this investigation are also presented in Table 2-4. The results showed
that as the fraction of ferrous metal in the "synthetic" CPU mixture decreased the lead
concentration in the TCLP leachate increased. The lead concentrations in these samples ranged
from 0.3 mg/L in the standard "synthetic" CPU mixture (68.2% ferrous metal) to 44 mg/L in the
zero ferrous "synthetic" CPU mixture (0% ferrous). The pH was observed to slightly increase
while the DO and ORP decreased with an increase in the fraction of ferrous metal in the sample.
This indicated that there was some relationship between the redox potential of the solution and
CHAPTER 2
2-9
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
lead and iron leachabiiity during the TCLP.
2.3.5 Effects of Particle Size
Particle size can affect the amount of metals that leach from a waste. As particle size
decreases, the surface area of the sample increases, the time to reach equilibrium is shortened,
and lead results typically increase in time limited tests such as the TCLP . This was observed for
CRT glass (Musson et al. 2000). Two sample sizes (<0.95 cm and <0.20 cm) of the "synthetic"
CPU mixture were tested to evaluate the impact of particle size on lead leachabiiity from CPUs.
No significant difference (Student's t-test, a=0.05 using Microsoft Excel) was observed in the
TCLP lead and zinc concentrations among the particle sizes used to test the "synthetic" CPU
mixture. Iron concentrations, however, were significantly greater in the smaller sample.
The heterogeneity of a device such as a computer CPU often makes obtaining a
representative sample difficult. Size reducing the device helps to increase the surface area for
leaching, homogenize the material, and obtain a representative sample. Size reduction had little
impact, however, on the surface area of the lead in computer CPUs since the lead typically only
exists in small amounts as tin/lead solder used on the surface of the PWBs. Size reduction of the
sample did increase the surface area of the ferrous metal component and, as a result, higher iron
concentrations were measured in the small (<0.20 cm) sample. As seen in the component
testing, the presence of iron in the sample affected lead leachabiiity. Additionally, any increase
in lead leachabiiity in the smaller particle size sample may have been counteracted by the higher
iron concentrations, resulting in a concentration nearly the same as the larger sample. The final
pH, DO, and ORP of the leachate did not differ greatly between the two samples. Table 2-5
summarizes the results.
Table 2-5. TCLP Results of Synthetic Computer CPU Mixture (Average of 3 Samples)
Size
Lead
Iron
Copper
Zinc
PH
(cm)
(mg/L)
(mg/L)
(mg/L)
(mg/L)
<0.95
5.16±0.01
0.3±0.2
19±9
BDL
136±24
<0.20
5.20±0.02
0.3±0.03
52±6
BDL
144±3
2.3.6 Impact of Head Space
Table 2-6 presents the results from an investigation of the impact of headspace above the
leaching fluid on lead leachabiiity. The headspace was measured as a ratio of the volume of air
CHAPTER 2
2-10
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
to the volume of liquid in the extraction vessel (Va/Vl). Testing was performed using samples of
the "synthetic" CPU mixture. In general, the pH, DO, and ORP increased as the headspace
above the leaching fluid increased. The DO and ORP data are presented in Figure 2-3. The
results showed that the leachabiiity of lead and iron increased as the headspace in the leaching
vessel increased. Copper was not detected (MDL=0.05 mg/L). Additionally, an increase in the
percentage of Fe2+ within the total iron was observed as the headspace above the leaching fluid
increased.
Table 2-6. Impact of Head Space Results (Average of 3 samples)
Va/Vl
PH
Lead
(mg/L)
Iron
(mg/L)
Copper
(mg/L)
Zinc
(mg/L)
0
5.05±0.01
0.6±0.03
7±0.2
BDL
112±5
0.17
5.16±0.01
0.3±0.2
19±9
BDL
136±24
0.5
5.37±0.01
0.9±0.2
200±15
BDL
138±8
1
5.36±0.03
2.5±0.7
217±39
BDL
134±11
Figure 2-3. Dissolved Oxygen and ORP Results from Head Space Impact Study
0.80 -1
r 200
^ 0.60 -
- 0
§
J, 0.40 -
¦
- -200
O
/
Q 0.20 -
- -400
0.00 -
1
- -600
0.00 0.17 0.50 1.00
Va/Vl
¦DO ORP
2.3.7 Comparison of Filtered vs. Non filtered Samples
Lead, copper, and zinc have been reported to adsorb to hydrous ferric oxide (Kendall, 2003).
Kendall reported that as solution pH increased, concentrations of lead, copper, and zinc
decreased and sorption increased, lead being the most strongly adsorbed. Sorption to hydrous
ferric oxide would cause the adsorbed lead, copper, and zinc to be filtered out during the
filtration process thus resulting in lower concentrations being measured in the leachates of the
CHAPTER 2
2-11
FACTORS AFFECTING TOXICITY CHARACTERISTIC
TEACHING PROCEDURE IE4D LEACHABIIITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
filtered samples. Samples of filtered and nonfiltered leachate were collected during TCLP tests
of the "synthetic" CPU mixtures. An evaluation of the difference in lead, iron, copper, and zinc
concentrations between the filtered and nonfiltered samples was conducted to determine if these
metals were filtered out during the filtration process.
Overall, iron and zinc concentrations differed little between all filtered and nonfiltered
samples. However, during the headspace studies, when the headspace above the leaching fluid
exceeded a headspace ratio (Va/Vl) of 0.5, the lead concentrations in the nonfiltered samples
were significantly higher than filtered samples (Student's t-test, a=0.5 using Microsoft Excel), as
presented in Figure 2-4. Although copper was not detected in the filtered samples, the copper
concentrations in the nonfiltered samples increased from 2.5 mg/L in the Va/Vl=0 to 4.3 mg/L in
the Va/Vl=l sample.
Figure 2-4. Lead Results from Head Space Impact Study
100.00
lb 10.00
8
i
D
§
u
£
1.00
0.10
0.01
E3 Filtered
S Nonfiltered I
H
£
£
o
0.16 Va/Vi 0.5
2.4 Discussion
2.4.1 TCLP of PWBs vs. Whole CPUs
Results of this study indicated that lead leachabiiity from computer CPUs depended on
several factors: mass of PWB in the sample, composition of the CPU, and the headspace above
the leaching fluid in the extraction vessel. Lead leachabiiity from PWBs during the TCLP was
generally in the range of 100 mg/L to 200 mg/L and decreased as the percentage of PWB in the
sample decreased. This indicates that CPUs that contain a high fraction of PWB by weight tend
to leach more lead.
CHAPTER 2 2-12 FACTORS .AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABIIITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
However, the mass of PWB in the CPU was not the only factor that affected lead
leachabiiity during the TCLP. The components of the CPU impact the lead concentration in the
leachate during the TCLP. For example, the TCLP lead concentration from the "synthetic" CPU
mixture was 0.3 mg/L. However, tests performed on PWBs solely had predicted results of 39
mg/L and 24 mg/L. This comparison demonstrated that the composition of the CPU did indeed
impact the lead leachabiiity from the computer CPU.
2.4.2 Impact of CPU Components (Ferrous Metal)
Because CPUs consist of several components (PWBs, ferrous metals, nonferrous metals,
plastics, and wires/cables), it was important to determine how each component impacted lead
leachabiiity during the TCLP. Results indicated that the ferrous metal component of the CPU
significantly decreased lead leachabiiity. Other studies have documented that iron impacts lead
leachabiiity during the TCLP. Kendall found that adding iron or zinc metal shavings to brass
foundry casting sand significantly decreased lead leachabiiity during the TCLP (Kendall, 2003).
Kendall explains that the iron metal caused the TCLP leaching fluid to become more reducing
therefore decreasing lead leachabiiity. In addition, lead and copper ions that are in solution will
be reduced by the zinc and iron, which will cause the concentrations of lead and copper
measured in the TCLP leachate to remain low if metallic zinc or iron is present (Kendall, 2003).
Results of the material impact also indicated that plastic increased lead leachabiiity from the
PWBs. This is believed, however, to be a physical process rather than a chemical reaction since
plastic is generally considered chemically inert with respect to lead leachabiiity.
Further investigation of ferrous metal impact on lead leachabiiity from CPUs indicated that
the fraction of ferrous metal (i.e., steel) in the sample significantly impacted the lead
concentration in the leachate. The galvanic series shows that the greater the difference in
electrode potential the greater the potential for corrosion to occur (Snoeyink et al. 1980). Zinc,
which has an electrode potential of +0.76 with respect to oxidation of the metal to the divalent
ion, reduced the lead that was leached into solution. The source of zinc in the samples is from
the thin layer of zinc applied to the steel during the galvanizing process. This caused lead to
plate out onto the metallic zinc as zinc ions were released into solution. However, the minor
amount of zinc present became depleted.
Iron has a greater electrode potential (+0.44 V) than lead (+0.126 V) with respect to
oxidation of the metal to the divalent ions, but less than zinc. As the zinc was delpleted and iron
dissolved from the steel, Fe2+ ions were released into solution and a negative charge was
produced on the steel by the remaining electrons. The Pb2+ ions that leached into solution were
attracted to the negatively charged steel, which resulted in lead plating out onto the steel.
Lead and copper leachabiiity decreased as the fraction of ferrous metal in the sample
increased. As the fraction of ferrous metal in the sample increased, pH increased and the DO
and ORP of the TCLP leachates decreased. As iron was released into the leachate, DO and H+
CHAPTER 2
2-13
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
Chapter 2 Factors Affecting Toxicity Characteristic Leaching Procedure Lead
Leachabiiity from Computer CPUs
were consumed as Fe2+ ions were released. This confirmed Kendall's findings that the ferrous
metal (i.e., steel) in the sample made TCLP leaching fluid more reducing. Furthermore, metallic
iron and zinc reduced some of the lead and copper ions that were in solution, which also
attributed to the lower concentrations being measured in the leachate as the fraction of ferrous
metal increased.
Zinc leachabiiity was observed to increase as the fraction of ferrous metal in the sample
increased. Based on the electrode potentials, zinc (+0.76 V) can reduce iron (+0.44 V), lead
(+0.126 V), and copper (-0.345 V) (Snoeyink et al. 1980). This explained why the iron
concentration decreased from 69 mg/L in the 20% ferrous sample to 19 mg/L in the 68.2%
ferrous sample. Increasing the amount of ferrous metal (i.e., steel) in the sample also increased
the amount of zinc since the steel was coated with a thin layer of zinc for galvanic protection. It
is believed the metallic zinc prevented the Fe2+ ions from leaching into solution, which resulted
in a decrease in the iron concentrations measured in the leachate as the fraction of steel in the
sample increased. This also confirmed Kendall's findings that zinc was not reduced by iron and
remained at high concentrations in the TCLP leachate (Kendall, 2003).
2.4.3 Impact of Extraction Vessel Headspace
The third factor that can impact lead leachabiiity from computer CPUs during the TCLP is
the headspace above the leaching fluid in the extraction vessel. The results indicated that as the
headspace to liquid ratio (Va/Vl) increased, the pH, DO, and ORP generally tended to increase.
The TCLP leaching fluid became more oxidizing as Va/Vl increased resulting in an increased
leaching of lead and iron. An analysis of the samples showed that as Va/Vl increased the
difference in the lead, iron, and copper concentrations between the filtered and nonfiltered
samples increased. Zinc was not greatly impacted by a change in the headspace above the
leaching fluid. This was expected since prior experiments indicated that all zinc was oxidized
under most conditions.
Although more lead was being leached as Va/Vl increased, more of the lead was sorbed or
precipitated as Va/Vl increased. Lead leachabiiity has been reported to adsorb to hydrous ferric
oxide (Kendall, 2003). Kendall reported that solution concentrations of lead, copper, and zinc
decreased and that sorption increased as pH increased and that lead was the most strongly
adsorbed. These results can be explained by the formation of hydrous ferric oxide. The TCLP
solution, which is oxygenated, is believed to have caused the Fe2+ ions to further oxidize to
Fe3+, which then formed ferric hydroxide. Through hydrolysis, the ferric hydroxide was then
converted to hydrous ferric oxide. The pH values measured during the evaluation of the impact
of headspace ranged from 5.05 to 5.36. This is below the pH at which zinc is expected to adsorb,
explaining why zinc concentrations between filtered and nonfiltered samples were not impacted
by the change in Va/Vl. It appears that the lead and copper adsorbed to the hydrous ferric oxide
and were removed during the filtration process thus resulting in the difference in the lead and
copper concentrations between the filtered and nonfiltered samples.
CHAPTER 2
2-14
FACTORS AFFECTING TOXICITY CHARACTERISTIC
LEACHING PROCEDURE LEAD LEACHABILITY
FORM COMPUTER CPUS
-------
3 Evaluation of a Large-Scale Modified TCLP for
RCRA Toxicity Characterization of Computer
CPUs
3.1 Introduction
As noted previously, the collection of a representative sample from heterogenous materials
such as CPUs and other electronics is very difficult. Chapter 3 assesses an alternative
methodology that would resolve some of the issues faced when conducting a TCLP on a device
such as a CPU. A large-scale TCLP method was developed in which an entire electronic device,
such as a CPU, is placed into a large extraction vessel and leached while maintaining the TCLP
requirements for the liquid-to-solid ratio and the extraction fluid. Size reduction is not
performed in this method; the CPU is simply disassembled and placed into the extraction vessel
and rotated for 18 hours. Leaching the entire device eliminates any human bias that is introduced
during sample processing and collection.
3.2 Materials and Methods
3.2.1 Research Approach
In this method, CPUs were completely disassembled and leached using a methodology
similar to the TCLP. The procedure was modified to allow the entire computer CPU to be tested.
The large-scale extractor rotated at half the speed (13 rpm) of the standard TCLP rotary
extractor. Therefore, a preliminary investigation was conducted to evaluate the impact of the
rotation speed on lead leachability during the TCLP.
A series of experiments that evaluated the leachability of lead, iron, copper, and zinc as a
function of time during both the standard TCLP and the large-scale TCLP methods was
performed. The purpose of these experiments was to provide a set of data that would be used to
evaluate the chemical properties of the leaching solution and metal concentrations measured in
the leachate. This data would then be used to determine if the time requirement of the TCLP (18
hours) was sufficient to achieve chemical equilibrium in the large-scale method and if any
physical or chemical differences existed between the standard TCLP method and the proposed
large-scale TCLP method
Results of this study were used to assess the applicability of a large-scale TCLP for the
hazardous waste determination of computer CPUs. Lead, iron, copper, and zinc results are
presented. Although lead is the primary chemical of interest, the analysis of iron, copper, and
zinc can be used to describe the processes that have been documented to impact lead
leachability: reduction by metallic iron and sorption by hydrous ferric oxide (Kendall, 2003). An
CHAPTER 3
3-1
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
evaluation was conducted to determine the similarities (or lack thereof) between the results of
the large-scale TCLP and standard TCLP methods.
3.2.2 Sample Collection and Processing
Computer CPUs were collected from a demanufacturing facility and a local household
hazardous waste collection center. Forty-three personal computer CPUs were collected. Each
CPU was completely disassembled and separated into five material categories as in Chapter 2:
printed wire boards (PWBs), ferrous metals, nonferrous metals, plastics, and wires/cables. Each
category was weighed to determine the CPU composition and total weight.
3.2.3 Large-Scale Leaching Procedure
The large-scale TCLP was performed by leaching an entire disassembled CPU using a large-
scale version of the TCLP. A 55-gallon extraction vessel (high density polyethylene (HDPE)
drum) was placed on a Morse 1-300 Series, Endover Drum Rotator (Morse Manufacturing).
TCLP extraction fluid #1 was mixed in the extraction vessel by adding 114 mL of glacial acetic
acid and 129 mL of 10 N sodium hydroxide solution diluted to 20 L. The amount of extraction
fluid was dependent on the mass of the solid sample. The TCLP requires a liquid to solid ratio
of 20:1. For example, a 10-kg CPU requires 200 L of extraction fluid. The maximum sample
mass for the large-scale TCLP was 10 kg due to the volume required by the extraction fluid.
CPUs weighing more than 10 kg required representative fractions of each material type for that
particular CPU to be chosen at random to obtain a 10-kg sample. The CPUs were not size
reduced but were disassembled. The extraction fluid was mixed by rotating the solution on the
drum rotator. Initial measurements of the pH, ORP and DO were recorded. The initial pH of
TCLP extraction fluid #1 was 4.93±0.05. Measurements of pH, ORP, and DO, were made using
the same equipment as described in Chapter 2. A blank sample of the TCLP extraction fluid was
collected for each leaching extraction.
The disassembled computer CPU was placed into the extraction fluid and rotated end-over-
end at a speed of 13 rpm for 18 hours. After rotation, the samples were drained from the bottom
of the extraction drum and the final pH, DO, and ORP of the leachates were recorded. As in the
studies described in chapter 2, samples of nonfiltered and filtered leachate were collected. The
TCLP leachates were filtered through a glass fiber filter (0.7 [j,m pore size) using pressure
filtration and preserved by adding concentrated nitric acid until the pH of the filtrate was below
2. Analysis of lead, iron, zinc, and copper concentrations as well as ferrous ion analysis was
performed on the filtered and nonfiltered leachate using the methods described in Chapter 2.
3.2.4 Impact of Extractor Speed
The TCLP requires that the rotary extractor rotate the samples at 30±2 rpm. However, the
CHAPTER 3
3-2
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
rotator used in the large-scale TCLP was only rated at 13 rpm. To evaluate the impact of the
extractor speed on lead leachability during the TCLP three samples of "synthetic" CPU mixture
were leached at 0, 13, and 28 rpm, respectively. The TCLP was performed by manually size
reducing (i.e., hand cutting with shears) 100-g samples of the "synthetic" CPU mixture so they
were capable of passing a 0.95-cm sieve (USEPA, 1996). The reader is referred to Chapter 2 for
a description of the preparation of the "synthetic" CPU mixture. The samples were placed into a
2L extraction vessel. Two liters of TCLP extraction fluid #1 were added to each sample and the
samples were placed on a rotary extractor (Analytical Testing Corporation). The samples were
performed in triplicate and a TCLP blank was included for each set of leaching extractions. The
TCLP leachates were filtered through a glass fiber filter of 0.7 [j,m pore size using the pressure
filtration procedure and preserved by adding concentrated nitric acid until the pH of the filtrate
was below 2. In addition to collecting the filtered leachate, samples of nonfiltered leachates
were also collected. The samples were then placed in HDPE bottles and stored until acid
digestion.
3.2.5 Time Studies
A series of time studies were conducted to investigate lead leachability from computer CPUs
as a function of time using the large-scale TCLP method. Three CPUs (2 different models) were
tested in this series of time studies. Throughout the period of the time study (approximately 90
hours), 2 L of the leachate were collected approximately every 9 hours for metals analysis.
Fresh extraction fluid was not added to the sample to maintain the 20:1 liquid-to-solid ratio, thus
the liquid-to-solid ratio of the sample gradually decreased over the period of the experiment.
3.2.6 Methodology Comparison
Forty CPUs were tested to compare the results between the TCLP and the large-scale TCLP.
The CPUs were comprised of eight model types. Three testing methodologies were performed
on each CPU model: large-scale TCLP on disassembled CPUs, standard TCLP on mechanically
shredded CPUs, and standard TCLP on manually size-reduced (i.e., hand cut) samples of
selected CPU components. Table 3-1 summarizes the testing methodologies. The large-scale
TCLP was performed, as described in Section 3.2.3, on 17 of the CPUs.
CHAPTER 3
3-3
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
Table 3-1. Testing Methodologies
Model
Number of C PI s Tested
Standard TCI.I*
Standard TCI.I*
l.ar«e Scale 1 C I.I»
Mechanically
1 land Cut
Disassembled
Shredded
1
3
2
3
2
2
1
1
3
1
1
2
4
1
2
1
5
2
2
4
6
1
1
2
7
1
2
1
8
-
1
3
Totals
11
12
17
Two techniques for conducting the standard TCLP, each meeting the requirements of the
method, were performed on 23 of the CPUs tested in this study. Eleven of the samples were
prepared by shredding the entire CPU in an industrial shear shredder equipped with 2-inch
blades (SSI Series 40H Model 2000-H), which was located at an electronic equipment
demanufacturing facility in Largo, FL. Since the materials did not meet the TCLP size
requirements after being passed through this shredder, each CPU was passed through a second
shear shredder, which was located at SSI Shredding Systems Inc. headquarters in Oregon and
was capable of size reducing the material down to 3/4 inch (SSI Series 22Q Model Q55ED (40)).
Each CPU was stored in plastic storage containers and transported to the laboratory. Each
sample was then placed on a 0.95-cm sieve and the material retained on the sieve was further
processed by manually size reducing (i.e., hand cutting) the pieces until they were capable of
passing the 0.95-cm sieve as required by the TCLP method. The remaining 12 CPUs were
disassembled and representative fractions of the major material types were selected at random.
The materials were then manually size reduced (hand cut) using shears to a size capable of
passing the 0.95-cm sieve.
Each sample was placed into a 2L extraction vessel. Two liters of the TCLP extraction fluid
#1 were added to each sample. The samples were placed on a rotary extractor and rotated for 18
hours. The leachates were then filtered through a 0.7-[j,m glass fiber filter using the pressure
filtration technique and preserved with nitric acid for metal analysis (USEPA, 1996)
CHAPTER 3
3-4
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
3.3 Results
3.3.1 Impact of Extractor Speed
Results of the extractor speed study are presented in Figure 3-1. Although the TCLP requires
samples to be rotated at 30±2 rpm, the extractor used in the large-scale TCLP was only capable
of 13 rpm. A Student's t-test (a=0.5 using Microsoft Excel) performed on the results indicated
that lead and iron concentrations in the TCLP leachate were not significantly different between
the samples rotated at 30 rpm and the samples rotated at 13 rpm. Therefore, the speed of the
large-scale extractor was not expected to be a factor. Results also showed that the lead
concentration measured in the TCLP leachate was significantly higher in the sample that was not
rotated (0 rpm). This was attributed to the fact that the iron concentration in 0-rpm sample was
significantly lower (Student's t-test, a=0.05 using Microsoft Excel) than the 13-rpm and 28-rpm
samples. As shown in Chapter 2, iron impacts lead leachability from computer CPUs during the
TCLP. The lack of mechanical agitation may have allowed protective oxidation layers to form
on the ferrous metal and reduced exposure of the bare metal that would otherwise occur from the
abrasion.
100.00
1000
'1 1.00
h
u 0.10
0.01
3.3.2 Time Studies
The lead concentrations measured during the time studies ranged from 1 mg/L to 10 mg/L
with iron ranging from 13 mg/L to 341 mg/L. Lead concentrations peaked in the three filtered
Figure 3-1. Impact of Rotation Speed Results
.wwwwwv
.wwwwwv
.wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwv
wwv
wwv
wwv
wwv
wwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
wwwwwv
hNNVVVW.Y
-.wwn
-.wwn
-.wwn
-.wwn
.wwwwwv
lWWWWWN
lWWWWWN
lWWWWWN
lWWWWWN
lWWWWWN
lWWWWWN
baaaaaa
0 RPM
13 RPM
28 RPM
~ Pb
E3 Fe
CHAPTER 3
3-5
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
leachate samples from 6 mg/L to 10 mg/L at 18 hours to 27 hours. After 27 hours, the lead
concentration decreased below the 5 mg/L TC limit. The lead concentration increased again
after 80 hours of rotation. Iron concentrations in the filtered leachates increased with time and
peaked between 45 hours and 60 hours at concentrations ranging from 292 mg/L to 341 mg/L.
After approximately 60 hours, the iron concentrations in all three filtered samples decreased with
time. The lead and iron results of the large-scale TCLP time study samples are presented in
Figure 3-2. This figure provides a comparison of the filtered and nonfiltered results.
The TCLP leachate of all of the samples was visually observed to change from a gray color
initially to an orange (rust) color as time progressed, which indicated that the iron was being
oxidized. The copper concentrations measured in the leachate of all of the filtered samples
generally were not detected (MDL=0.1 mg/L). The zinc concentration measured in the filtered
and nonfiltered leachate of the large-scale TCLP time study samples varied between 114mg/L
and 181 mg/L and did not change greatly with time.
The pH of the TCLP leachate of the large-scale samples ranged from 5.01 to 5.47 with the
highest values of each of the three samples (5.47, 5.42, and 5.44) occurring between 45 hours
and 60 hours. After 60 hours, the pH decreased with time to values ranging from 5.22 to 5.30 at
approximately 90 hours. The ORP of the leaching solution tended to fluctuate throughout the
time of the study and peaked between 9 hours and 27 hours. Results of these samples indicated
that the TCLP leaching fluid in the large-scale TCLP remained oxidizing during the duration of
the experiment.
CHAPTER 3
3-6
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
Figure 3-2. Comparison of Metals Results from TCLP Time Study Experiments
10000
10000
40 60 80
Hours
10000
0 20 40 60 80 100
Hours
C
20 40 60 80
Hours
100
Pb Concentration Filtered Samples
Pb Concentration Nonfiltered Samples
Fe Concentration Filtered Samples
Fe Concentration Nonfiltered Samples
A) Sample 1. B) Sample 2. C) Sample 3.
Results presented in Figure 3-2 indicate lead leachability was limited by an oxidation-
CHAPTER 3
3-7
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
reduction process (reduction by metallic zinc and iron) in the beginning of the time studies.
Precipitation or adsorption is not indicated due to the small difference in the concentrations of
lead and iron measured in the leachates of the filtered and nonfiltered samples. As established in
Chapter 2, metallic iron and zinc reduce lead during the TCLP. As time progressed beyond
approximately 30 hours, results indicate that lead leachability was primarily impacted by the
adsorption to hydrous ferric oxide. The observed color change from gray to orange of the
solutions provides evidence that ferric oxide and hydrous ferric oxide were formed during the
TCLP. As time progressed, iron continued to oxidize and form hydrous ferric oxide which
adsorbed the lead that was leached and allowed additional lead to leach into solution as
evidenced by the nearly constant lead concentration measured in the leachates of the filtered
samples. Results indicated lead was adsorbed to the hydrous ferric oxide and was removed
during the filtration process.
The lead, iron, copper, and zinc concentrations from Sample 2 are presented in Figure 3-3.
These results show that as time increased, the difference in the lead, iron, and copper
concentrations between the filtered and nonfiltered samples increased. It is noted that copper in
the filtered samples was only detected in the 88-hour sample. This further confirmed Kendall's
findings. The iron was oxidized over the entire duration of the experiment and formed hydrous
ferric oxide, which absorbed a portion of the lead and copper that was leached. The hydrous
ferric oxide was filtered out during the filtration procedure, thus removing the adsorbed lead and
copper from the leachate. As time progressed, lead and copper continued to adsorb to the
hydrous ferric oxide, which resulted in the large difference in concentrations of the filtered and
nonfiltered samples at the end of the experiments. The zinc concentrations measured in the
filtered and nonfiltered samples did not greatly differ, indicating that zinc was not adsorbed.
This was expected since the pH of the solution was below 7, the pH above which significant
adsorption of zinc commences under TCLP conditions (Kendall, 2003).
CHAPTER 3
3-8
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
Figure 3-3. Sample 2 Filtered vs. Nonfiltered Metals Concentrations.
o.oi
0 20 40 60
Hours
200
180
tJ 160
oo
N
120
100
Zinc
D
;'b
: G..CD
#
#
«#
m
J
M
(l
0
80 100
Filtered
O" Nonfiltered
0 20 40 60
Hours
80 100
3.3.3 Methodology Comparison
The lead concentrations measured in all (TCLP w/shredded, TCLP w/hand-cut, and large-
scale TCLP) leachates ranged from 0.2 mg/L to 21.4 mg/L with 14 of the 40 CPUs tested
exceeding the 5 mg/L TC limit. Of the 14 CPUs that exceed the TC lead limit, 13 were tested
using the large-scale TCLP method and one was tested using the standard TCLP method w/
mechanical shredding. In addition, the results show that shredding the CPUs did not greatly
impact the lead concentration in the leachate when compared to the samples that were hand cut.
For all samples, iron concentrations ranged from 6 mg/L to 255 mg/L, copper concentrations
ranged from below the detection limit (0.05 mg/L) to 0.31 mg/L, and zinc concentrations ranged
CHAPTER 3
3-9
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
from 27 mg/L to 156 mg/L. Ferrous ion analysis was limited to 32 of the 40 filtered samples and
15 unfiltered samples. Based on the ferrous iron (Fe2+) analysis, the iron that was measured in
the leachate ranged from 2% Fe2+ to 100% Fe2+ and on a majority of the samples was greater
than 50% Fe2+. Results of the TCLP methodology comparison are presented in Table 3-2.
The pH measurements of the leachate from the standard TCLP ranged from 4.99 to 5.26 and
from 5.03 to 5.32 in the large-scale TCLP. The ORP and DO measurements indicated that the
large-scale TCLP produced a more oxidizing environment than the standard TCLP method. The
ORP measurements of the leachate from the standard TCLP ranged from 20.3 to -387 RMV and
the DO from 0.14 mg/L to 1.22 mg/L. The ORP measurements from the large-scale TCLP
ranged from 170 RMV to -84 RMV and the DO from 2.95 mg/L to 4.92 mg/L. The ORP
measurements from the large-scale TCLP were positive in a majority of occasions.
The lead, iron, and zinc data from CPU #1 are presented in Figure 3-4. These figures
represent a typical example of the data from the TCLP methodology comparison. The lead
concentrations measured in the leachate of samples that were leached using the large-scale TCLP
method were higher than those leached using the standard TCLP method. The iron and zinc
concentrations measured in the leachate from the large-scale TCLP were equal to or greater than
the standard TCLP methods in a majority of occasions. Copper was not detected (MDL=0.05) in
any of the filtered samples for CPU #1. The zinc concentrations varied among all of the samples
that were tested and was not impacted by the testing methodology.
The pH, ORP, and DO data from CPU #1 are presented in Figure 3-5. The pH measured in
the leachates from CPU #1 ranged from 5.04 to 5.19 and did not greatly differ between the
testing methods. However, DO and ORP were impacted by the testing methodology. The DO
measurements for CPU #1 from the large-scale TCLP ranged from 2.95 mg/L to 3.85 mg/L,
while the DO from the standard TCLP tests ranged from 0.18 mg/L to 0.67 mg/L. The ORP
ranged from 14.5 RMV to 124.1 RMV in the large-scale TCLP and from -90 RMV to -310 RMV
in the standard TCLP tests. These results indicated that the leaching solution in the large-scale
TCLP test was more oxidizing than the leachate of the standard TCLP.
CHAPTER 3
3-10
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
PI
T~
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
5
Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
able 3-2. Methodology Comparison Results
Filtered)
Model
Processing/ TCI.P Met hod
l.esul
(ni»/l.)
I ron
(ni»/l.)
Copper
(m»/l.)
Shredded/ Standard
1.4
92
BDL
Shredded/ Standard
6.0
38
BDL
Shredded/ Standard
1.00
86
BDL
Disassembled/ Large
9.0
104
BDL
Disassembled/ Large
9.0
94
BDL
Disassembled/ Large
5.0
93
BDL
Manual/ Standard
0.5
50
BDL
Manual/ Standard
0.4
11
BDL
Shredded/ Standard
1.1
106
BDL
Shredded/ Standard
0.9
85
BDL
Disassembled/ Large
5.5
255
BDL
Manual/ Standard
0.3
18
BDL
Shredded/ Standard
3.2
84
BDL
Disassembled/ Large
21.4
117
0.05
Disassembled/ Large
16.4
132
0.2
Manual/ Standard
2.3
20
BDL
Shredded/ Standard
1.0
119
BDL
Disassembled/ Large
9.5
127
BDL
Manual/ Standard
0.4
24
BDL
Manual/ Standard
0.5
31
BDL
Shredded/ Standard
3.6
96
BDL
Shredded/ Standard
1.5
136
BDL
Disassembled/ Large
5.3
65
BDL
Disassembled/ Large
3.1
24
0.06
Disassembled/ Large
15.5
131
0.08
Disassembled/ Large
4.0
62
0.05
Manual/ Standard
0.3
BDL
Manual/ Standard
3.1
59
0.06
Shredded/ Standard
1.3
111
BDL
Disassembled/ Large
0.6
44
BDL
Disassembled/ Large
0.5
50
BDL
Manual/ Standard
0.3
35
BDL
Shredded/ Standard
0.5
147
BDL
Disassembled/ Large
9.1
189
0.07
Manual/ Standard
0.2
BDL
Manual/ Standard
0.1
19
0.32
Disassembled/ Large
1.4
201
0.08
Disassembled/ Large
7.1
253
0.13
3-11
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
CPU
Model
Processing/ TCLP Method
Lead
(mg/L)
Iron
(mg/L)
Copper
(mg/L)
Zinc
(mg/L)
39
8
Disassembled/Large
6.6
267
0.14
134
40
8
Manual/ Standard
0.5
44
BDL
220
Figure 3-4. Metal Concentrations from Methodology Comparison for CPU #1.
Shredded Large Hand Cut
A) Lead Concentration. B) Iron Concentration. C) Zinc Concentration.
CHAPTER 3
3-12
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
Figure 3-5. Laboratory Measurements from Methodology Comparison for CPU #1.
5.20
5.15
% 5.10 ¦ &
PL,-100
o
5 05 ' -200
5.00 -300
Shredded Large Hand Cut Shredded Large Hand Cut
CD
A 2
Shredded Large Hand Cut
A) Final pH. B) Final ORP. C) Final DO.
Previous work (Chapter 2) evaluating the factors that affect lead leachability from computer
CPUs during the TCLP showed that the headspace above the leaching fluid can impact lead
leachability from computer CPUs. Results from that study showed that as the head space-to-
liquid ratio (Va/Vl) increased, the pH, DO, and ORP increased, thus indicating that the TCLP
leaching fluid became more oxidizing as Va/Vl increased. The leachability of lead and iron
increased as the environment of the leaching fluid became more oxidizing. However, the Va/Vl
ratio in the large-scale TCLP method was not greatly different from the standard Va/Vl ratio of
0.16. For example, the mass of CPU #1 was 8,418 g, which resulted in 168.4 L of the TCLP
leaching fluid being used in the extraction based on a 20:1 liquid-to-solid ratio. The Va/Vl ratio
of the large-scale TCLP method was then estimated assuming the total volume of the extraction
drum to be 55 gallons (208 L) and the volume taken up by the CPU itself to be 1.85 L. This
resulted in a Va/Vl ratio of approximately 0.22.
In the TCLP methodology, the CPU components are size reduced so that they are capable of
passing through a 0.95-cm sieve, while the components of the large-scale modified TCLP are not
size reduced. Size reduction of the components greatly increased the surface area of the iron that
was exposed to the leaching solution, which would otherwise have been galvanically protected
CHAPTER 3
3-13
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
Chapter 3 Evaluation of a Large-Scale Modified TCLP for RCRA Toxicity
Characterization of Computer CPUs
by the zinc coating. Exposing the raw iron to the leaching solution allowed it to dissolve more
readily than the non-size reduced components. In this study, metallic iron appears to have
caused the leachate of the TCLP to become more reducing. This was evident by the negative
ORP measurements and the relatively low DO concentrations in the samples that were tested
using the TCLP. The leachates of the large-scale modified TCLP, however, were more oxidizing
than the standard TCLP as evidenced by the positive ORP measurements and relatively high DO
concentrations. This may be attributed to the fact that the components were not size reduced,
which limited the amount of exposed iron. An evaluation of the impact of size reduction on lead
and iron leachability during the large-scale modified TCLP would be beneficial.
Analysis of the nonfiltered samples indicated that the lead, iron, and zinc concentrations did
not greatly differ, on the majority of occasions, from the concentration measured in the filtered
samples during both TCLP methods evaluated in this study. Copper, on the other hand, tended
to be present only in the nonfiltered samples. This indicated that although the large-scale
modified TCLP produced a more oxidizing environment, lead leachability appears to not have
been impacted by adsorption by hydrous ferric oxide. This was expected since the time study
results indicated that adsorption by hydrous ferric oxide did not occur until after 30 hours of
rotation.
3.4 Discussion
A large-scale TCLP approach was examined because of the inherent difficulties in
performing the TCLP on manufactured articles such as electronic devices. A major difficulty is
obtaining a representative size-reduced sample. Size reduction of electronic devices can be
difficult because of the bulky nature of the components. For example, computer CPUs are
composed of a high percentage of steel and other metals that are difficult to cut. The use of
industrial shredders for processing electronic devices is not practical because they often lead to
sample loss and cross contamination in addition to not meeting the size requirements. Manual
size reduction (i.e., hand cutting) using small-scale laboratory equipment is the only reasonable
option but is very time consuming. It is often left to the technician performing the test to select
the components and process them for testing, which can introduce human bias into the results.
The large-scale approach offers several advantages. It permits an entire electronic device to
be tested, eliminating human bias introduced when collecting the sample and processing it.
Obtaining a 100-g representative sample (as required by the TCLP) that contains representative
fractions of the components is not easily accomplished. The lack of size reduction also offers the
advantage of time efficiency. The test was designed to meet the intent of the TCLP in terms of
measuring the maximum equilibrium concentration occurring with the TCLP fluid at a 20:1
liquid to solid ratio. It was originally hypothesized that the modified procedure would result in
an underestimation of lead leaching because of the larger particle sizes. The lead was at greater
concentrations than expected because of the factors described earlier. It is noted that the large-
scale method requires the use of special equipment (a drum rotator) and requires more chemicals.
CHAPTER 3
3-14
EVALUATION OF A LARGE-SCALE MODIFIED
TCLP FOR RCRA TOXICITY
CHARACTERIZATION OF COMPUTER CPUS
-------
4 Leaching Results for Various Electronic Devices
4.1 Overview of Testing Performed
This chapter presents data for the testing of several common electronic devices. In the
preceding chapters, two leaching methods were explored: The standard TCLP and a large-scale
TCLP. A third method, the modified small scale method, was also employed in this chapter
where small electronic devices were disassembled and placed in their entirety (not size reduced)
into the 2-L extraction vessels. With the exception of size reduction, all other parameters of the
TCLP remained the same. This third method was used to measure lead leachate levels in
computer mice and remote controls. A total of 12 devices were leached in some fashion. Table
4-1 shows the type and number of each device tested and the test performed on those devices.
Table 4-1. Electronic Devices Tested
Device
Standard TCLP
(True)
Large-Scale TCLP
(L)
Modified Small-Scale
TCLP
(MS)
Personal Computer CPUs
23
41
Computer Monitors
9
Laptop Computers
6
9
Printers
9
Color Televisions
6
VCRs
8
Cellular Phones
38
14
Key Boards
3
1
Computer Mice
15
Remote Controls
4
6
Smoke Detectors
9
Flat Panel Monitors
8
The results are presented in separate sections and include the average composition of the
materials tested and the TCLP concentrations for several predominant elements. Mercury was
analyzed in the laptops since their composition includes a fluorescent light. Of all RCRA
regulated metals, only lead was found to have exceeded the TC levels at any time, as
demonstrated in Table A-l of the Appendix for CPUs. Therefore each section presents the
leachate levels for lead and for the components presented earlier that affected lead leachability
from electronic devices: zinc, iron, and copper. In each of the following tables, the method of
testing used for each sample is indicated as "True" (standard TCLP), "L" (large-scale TCLP), or
4-1
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
"M.S." (small-scale modified TCLP). Computer CPU testing is also designated "S.S." for CPU
samples tested using the standard TCLP on mechanically shredded samples.
Figure 2.1 is reproduced as Figure 4.1. This figure illustrates the average composition of
CPUs determined from the disassembly of twenty-nine computer CPUs separated into five major
material classes. Five of these CPU's were used to create the "synthetic" CPU mixture studied
in Chapters 2 and 3. The remaining twenty-four CRTs were utilized to perform additional large-
scale TCLP analysis for lead, iron, zinc, and copper. These results and the results of the forty
CPUs tested in Chapter 3 are listed in Table 4-2.
CPUs tested using the large-scale TCLP method resulted in 21 of 41 CPUs exceeding the TC
for lead with an average lead concentration of 5.4 mg/L overall. Samples prepared using the
mechanical shredding method resulted in 1 of 11 samples exceeding the TC for lead with an
average of 2.0 mg/L. The average lead concentration for the 12 CPUs tested using the standard
TCLP was 0.7 mg/L, none exceeding the TC for lead. The average for all CPUs tested in all
methods was 4.0 mg/L.
4.2 Personal Computer CPUs
Nonferrous
(" Metals, 5.40%
Wires, 3.10%
Printed Wire
Boards, 15.80%
Plastic, 7.50%
Ferrous Metals,
68.20%
Figure 4-1. Average CPU Composition of 29 Computer CPUs by Weight
4-2
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-2. TCLP Results for CPUs
S;ini|ik-
Make
Model Year
Mclliod
nii
\lclal ( onccnlmlion in
\unilvi'
1
vacluik- (my 1.)
Cu
Iv
Pli
Zn
1
Sun Microsystems
SPARK Station 2/
(SS.)
5.1
0.02
92
1.4
112
2
Sun Microsystems
SPARK Station 2/
(SS.)
5.1
0.01
38
6.0
107
3
Sun Microsystems
SPARK Station 2/
(SS.)
5.1
0.02
86
1.0
103
4
Sun Microsystems
SPARK Station 2/
(L)
5.2
0.02
104
9.0
143
5
Sun Microsystems
SPARK Station 2/
(L)
5.1
0.02
94
9.0
153
6
Sun Microsystems
SPARK Station 2/
(L)
5.1
0.02
93
8.0
156
7
Sun Microsystems
SPARK Station 2/
(True)
5.1
0.01
50
0.5
105
8
Sun Microsystems
SPARK Station 2/
(True)
5.0
0.01
11
0.4
118
9
Compaq
ProLinea 4/66
(SS.)
5.1
0.02
106
1.1
84
10
Compaq
ProLinea 4/66
(SS.)
5.2
0.01
85
1.0
122
11
Compaq
ProLinea 4/66/ 1995
(L)
5.3
0.03
255
5.5
128
12
Compaq
ProLinea 4/66/ 1994
(True)
5.2
0.01
18
0.3
147
13
IBM PS2
55SX/ 1987
(SS.)
5.2
0.03
84
3.2
128
14
IBM PS2
55SX/ 1989
(L)
5.1
0.05
117
21
81
15
IBM PS2
55SX/ 1991
(L)
5.1
0.20
132
16
92
16
IBM PS2
55SX
(True)
5.3
0.01
20
2.3
147
17
NCR
6020
(SS.)
5.2
0.03
119
1.0
99
18
NCR
6020/1994
(L)
5.1
0.04
127
10
103
19
NCR
6020/1993
(True)
5.1
0.01
24
0.4
130
20
NCR
6020
(True)
5.1
0.01
31
0.5
122
21
Oli
M4 Module 464/ 1994
(SS.)
5.1
0.01
96
3.6
43
22
Oli
M4 Module 464/ 1995
(SS.)
5.1
0.01
136
1.5
52
23
Oli
M4 Module 464/ 1994
(L)
5.0
0.02
65
5.3
21
24
Oli
M4 Module 464/ 1995
(L)
5.0
0.06
24
3.1
33
25
Oli
M4 Module 464/ 1994
(L)
5.1
0.08
131
16
27
26
Oli
M4 Module 464/ 1994
(L)
5.1
0.05
62
4.0
34
27
Oli
M4 Module 464/ 1994
(True)
5.2
0.01
5
0.3
172
28
Oli
M4 Module 464/ 1994
(True)
5.1
0.06
59
3.1
115
29
Network General
Sniffer Server/ 1995
(SS.)
5.1
0.02
111
1.3
111
30
Network General
Sniffer Server/ 1991
(L)
5.0
0.02
44
0.6
99
31
Network General
Sniffer Server/ 1996
(L)
5.1
0.02
50
0.5
101
32
Network General
Sniffer Server/ 1995
(True)
5.2
0.01
35
0.3
106
33
ATT Globalyst 550
9011/1994
(SS.)
5.2
0.02
147
0.5
111
34
ATT Globalyst 550
9011/1994
(L)
5.2
0.07
189
9
114
35
ATT Globalyst 550
9011/1994
(True)
5.0
0.04
6
0.2
168
36
ATT Globalyst 550
9011/1994
(True)
5.0
0.32
19
0.1
129
37
Compaq
Prolinea4/33 1994
(L)
5.3
0.08
201
8
215
4-3
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Siimplc
Make
Model Year
MciIukI
nii
\lclal ( onccnlmlion in
\unilvi'
1
vacluik- (niij 1.)
Cu
Iv
Ph
Zn
38
Compaq
Prolinea4/33 1993
(L)
5.2
0.13
253
7
160
39
Compaq
Prolinea4/33 1994
(L)
5.2
0.14
267
6.6
134
40
Compaq
Prolinea4/33 1994
(True)
5.1
0.05
44
0.5
220
41
Fortiva 5000
CPC-2803/1994
(L)
5.0
0.02
6
0.7
95
42
Digital DEC pc
806WW/1994
(L)
5.0
0.04
3
2.8
82
43
IBM
55SX/1989
(L)
5.1
0.08
131
13
70
44
IBM
55SX/1990
(L)
5.1
0.21
155
13
101
45
NCR
1993
(L)
5.2
0.04
150
6.5
120
46
e Machines
366i2/l 999
(L)
5.2
0.05
14
0.2
181
47
Gateway
4DX33/1993
(L)
5.0
0.02
25
0.2
127
48
Dimension XPS
MM8/1996
(L)
5.1
0.01
68
5.0
76
49
Dimension XPS
MM8/1997
(L)
5.2
0.04
5
0.4
85
50
Dimension XPS
MM8/1996
(L)
5.0
0.01
5
0.3
82
51
Dimension XPS
MM8/1997
(L)
5.0
0.02
9
0.3
87
52
Compaq
PS-5151-4A/1996
(L)
5.0
0.00
11
0.2
89
53
Compaq
PS-5151-4A/1996
(L)
5.0
0.01
35
0.2
87
54
Dell
450-L/1993
(L)
5.2
0.14
135
5.0
106
55
Gateway
4D X2-66/1994
(L)
5.1
0.01
62
1.8
113
56
Apple
M3098/1995
(L)
5.0
0.03
122
10.0
55
57
Dell
DCS/1998
(L)
5.1
0.01
11
0.2
191
58
NEC
1996
(L)
5.3
0.03
201
9.0
133
59
Dell
Optiplex GL
(L)
5.1
0.01
73
1.9
98
60
Dell
Optiplex GL
(L)
5.2
1.12
121
11.2
114
61
Dell
Optiplex GN/1998
(L)
5.2
0.01
13
0.3
204
62
Gateway
E 3000/1997
(L)
5.1
0.02
50
1.6
117
63
Dell
Optiplex GN/1998
(L)
5.2
0.01
8
0.3
202
64
Dell
Optiplex GS/1997
(L)
5.1
0.02
8
0.4
199
4-4
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.3 Computer Monitors
One recent study showed that color CRTs from televisions and computer monitors usually
exceeded the RCRA TC limits for lead (Musson et al., 2000; Townsend et al., 1999). However,
this study tested only the CRT glass and did not include the monitor as a whole. Nine computer
monitors were tested using the large scale testing methodology. All of the monitors tested
exceeded the TC limit for lead with an average concentration of 47.7 mg/L. This was expected
since the CRT glass accounted for over 60 percent of the weight of the monitors. Ferrous metal
was not a factor, comprising only 3% of the total weight.
Ferrous Metal 3%
Printed Wire
Nonferrous Metal
Plastic Casing
17%
\ i-Other Plastic 1%
Metal Casing 1%
Wires 4%
CRT 62%
Figure 4-2. Average Composition of Computer Monitors Tested
4-5
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-3. TCLP Results for Computer Monitors
Sample
Numlvi'
Miik-
\loilcl Year
Method
I'll
Melal ( onceiilralion in 1 .eacluile
(my 1.)
C ii
1 e
I'h
Zn
1
Dell
Ultrascanl5FS/9
2
(L)
5.1
4.50
107
13
9
2
Gateway
CS1572DG/1995
(L)
5.1
0.10
127
33
21
3
Packard
2025/1997
(L)
5.1
0.07
63
59
12
4
Gateway
500-069CS/1998
(L)
5.1
0.22
122
40
26
5
Compaq
471/1993
(L)
5.0
0.02
105
13
31
6
Samsung
CJ4681/1990
(L)
5.0
0.41
120
33
28
7
Phillips
1995
(L)
5.0
0.20
94
94
20
8
Gateway
500-069CS/1997
(L)
5.1
0.13
108
50
23
9
Apple
1995
(L)
5.1
0.90
125
95
15
4-6
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.4 Laptop Computers
Figure 4-3 shows the average composition of 15 laptop computers tested. Table 4-4 presents
the results of the tests. The large scale tests were run using samples consisting of 2 identical
model laptops. This was done to maximize the leaching solution volume and thus reduce the
head space in the large-scale vessel. Nine out of nine of the laptop models tested using the large-
scale procedure exceeded the TC limit for lead with an average concentration of 23 mg/L. Six
out of six laptops tested using the standard TCLP exceeded the TC limit for lead with an average
lead concentration of 37 mg/L resulting in an average of 29 mg/L for all laptops tested. Table 4-
4 lists the individual results. Data for copper, zinc, and iron were not available for those laptops
tested using the standard TCLP.
Figure 4-3. Average Composition of Laptops Tested
4-7
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-4. TCLP Results for Laptops
Sample
Make
Model Year
Melhod
nii
Melal ( oncenlralion in l.e
lchale
Numlvi'
(nitj 1.)
C ii
IV
I'h
Zn
"a
1
Compa
Compaq LTE 5000 (series
(L)
5.2
2.9
54
17
47
<0.03
2
IBM
IBM Thinkpad 9545/1995
(L)
5.1
0.3
67
26
16
<0.03
3
NEC
NEC Versa m/100
(L)
5.0
2.0
100
35
21
<0.03
4
Compa
Compaq LTE Lite 4/25E
(L)
5.0
1.1
82
26
15
<0.03
5
Compa
Compaq LTE 5280 (Series
(L)
5.0
2.2
81
21
42
<0.03
6
Zenith
Zenith ZWL-184-97
(L)
5.0
0.0
17
18
53
<0.03
7
Toshib
Toshiba T1200 System Unit #
(L)
5.1
0.1
80
15
55
<0.03
8
Zenith
Zenith ZWL-371-A
(L)
5.1
0.1
182
25
28
<0.03
9
Zenith
Zenith ZWL-371-A
(L)
5.0
0.2
206
28
38
<0.03
10
IBM
Thinkpad 9545
(True)
5.0
0.1
0.8
32
0.5
<0.03
11
Compa
Contura Series, 4/25C (2820)
(True)
5.2
.03
73
42
11
<0.03
12
NEC
VERSA m/100 (PC-580-6552)
(True)
5.0
1.9
46
86
1
<0.03
13
IBM
Thinkpad 350c
(True)
5.0
.01
116
11
51
<0.03
14
Dell
325N
(True)
5.0
.04
91
27
42
<0.03
15
AST
Premium Exec 386 SX/20
(True)
5.0
.03
39
19
152
<0.03
4-8
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.5 Printers
Figure 4-4 provides the average composition of nine computer printers tested using the
large-scale TCLP. The composition of the printers was similar to the CPUs, having a high
percentage of ferrous metals. Iron concentrations in the leachate averaged 147 mg/L. Five of
the nine printers tested exceeded the TC limit for lead. The average lead concentration for all the
printers was 5.0 mg/L. Individual results are shown in Table 4-5.
Non Ferrous
4-9
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-5. TCLP Results for Printers
Sample
Nuinlvr
Make
\loi_ld Year
\IciIukI
nii
\lclal ( oiKvnli'alion in 1 vacliak-
(my 1.)
C ii
IV
I'h
/n
1
OKIDATA
EN2700A
(L)
5.0
0.40
114
6.0
49
2
HP Deskjet
C4567A/1996
(L)
5.1
0.15
48
0.8
98
3
HP office jet
C3801/1999
(L)
5.1
0.46
231
6.6
54
4
HP Deskjet
890C
C5876A/
1998
(L)
5.0
0.12
176
3.6
43
5
HP Desk writer
2279A/1993
(L)
5.0
0.06
101
3.6
54
6
HP Deskjet
550C
C2121A/1993
(L)
5.1
0.22
196
5.8
35
7
HP Deskjet
500C
C2106A
(L)
5.1
0.10
131
6.6
61
8
HP Deskjet
672C
C2164A/1995
(L)
5.0
0.13
7
0.3
46
9
Panasonic Quiet
1991
(L)
5.1
0.05
174
5.5
64
4-10
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.6 Color Televisions
The construction of modern computer monitors and televisions does not differ greatly as seen
by comparison of previous Figure 4.2 for computer monitors with Figure 4.5 for color
televisions. As with computer monitors, all televisions tested using the large-scale modified
TCLP methods exceeded the TC limit for lead with an average value of 29 mg/L. The individual
results of the 6 televisions tested are shown in Table 4-6.
Figure 4-5. Average Composition of Color TVs Tested
4-11
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-6. TCLP Results for Color TV
Sample
Make
Model/Year
Method
pH
Metal Concentration in
Number
Leachate (mg/L)
Cu
Fe
Pb
Zn
1
JVC
C-13WL5/1994
(L)
5.1
0.05
68
31
23
2
Panasonic
CT216-A/1987
(L)
5.3
0.03
40
15
54
3
Philips
RAX 150 WA02/1990
(L)
5.2
0.01
95
15
21
4
Sanyo
1999
(L)
5.1
0.01
124
12
13
5
RCA
2001
(L)
5.0
0.07
140
43
18
6
Zenith
D1312W, 1987
(L)
5.1
0.10
135
29
15
4-12
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.7 VCRS
Eight VCRs were tested using the large-scale TCLP. The average composition of the VCRs
is shown in Figure 4.6. The VCRs, like the CPUs and printers, contained a higher percentage of
ferrous metal, 45%, with an average of 162 mg/L of iron in the leachate. The average leachate
concentration of lead was 9.49 mg/L with 7 of the 8 VCRs exceeding the TC limit for lead.
2%
Figure 4-6. Average Composition of VCRs Tested
4-13
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-7. TCLP Results for VCRs
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in
Leachate (mg/L)
Cu
Fe
Pb
Zn
1
Sylvania
H1600VD/1998
(L)
5.1
0.28
150
11
35
2
Philips
VRA631AT22/1999
(L)
5.5
0.17
56
8
52
3
Memorex
MVR4040/1997
(L)
5.2
0.22
230
10
58
4
JVC
HRDX420/1992
(L)
5.4
0.16
236
13
152
5
JCPenny
686-6054/1988
(L)
5.0
0.05
105
8
47
6
Emerson/Hitachi
V CR75 5/VT-M262A
(L)
5.0
0.01
10
1
92
7
Phillips
VR6615AT01/1992
(L)
5.4
0.05
237
13
130
8
Hitachi/JVC
1993/HR-J600U/1995
(L)
5.2
0.10
107
11
147
4-14
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.8 Cellular Phones
Cell phones, without their battery, were tested using the standard TCLP and a modified
TCLP procedure. The typical mass of a cell phone without its battery averages nearly lOOg.
Therefore, it was not necessary to use the large-scale TCLP extraction vessel as done for TVs,
computer monitors, keyboards, and VCRs. Whole, disassembled cell phones were placed in the
2L TCLP extraction vessels prescribed in the regulatory procedure. These are designated as
(M.S.) in Table 4-8. The average lead leachate concentration using the modified procedure was
34 mg/L with 10 of 14 samples exceeding the TC limit for lead. In addition to the modified
procedure, the standard procedure using samples reduced in size to less than 0.95 cm was
performed on 38 cell phones. These are designated as (True) in Table 4-8. The average
concentration of lead in the leachate was 20 mg/L with 28 of the 38 samples exceeding the TC
limit for lead. The average composition by weight of the cell phones is shown in Figure 4-7.
/ PC Board
/ 40%
LCD
4%
Solar Cell
0%
Magnessium Plate
3%
\ Plastics
\ 45%
Metals
8%
Figure 4-7. Average Composition of Cell Phones Tested
4-15
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-8. TCLP Results for Cell Phones
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in
Leachate (mg/L)
Cu
Fe
Pb
Zn
1
Motorola
SWF4018DF J 160017B
(M.S.)
4.9
0.01
43
141
6
2
Motorola
80148WNBEA
(M.S)
5.0
0.11
87
115
0.4
3
Sprint
QCP-2700
(M.S)
5.0
0.00
5
1.5
17.3
4
Motorola
SWF2049A H7 41843A4C
(M.S)
4.9
0.01
6
1.6
3.0
6
Ericsson
CF768
(M.S)
5.0
1.16
44
115
19.0
7
Nokia
2120
(M.S)
6.3
0.00
2
0.1
12.0
8
Nokia (b)
5160
(M.S)
4.9
0.03
9
11
20.3
9
Nokia (p)
5160
(M.S)
5.0
0.01
2
2
18.3
10
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
107
14
4.5
11
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
93
15
4.0
12
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
119
17
4.1
13
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
113
17
2.7
14
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
125
15
3.1
15
Motorola
Piper 34106NNDPA/1996
(M.S)
4.9
0.00
111
16
0.3
16
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.00
109
20
3.5
17
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.01
109
21
4.1
18
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.01
97
16
3.1
19
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.01
94
17
3.1
20
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.00
105
20
3.3
21
Motorola
Piper 34106NNDPA/1996
(True)
4.9
0.00
83
18
2.5
22
Ericsson
DF688/1997
(True)
5.5
0.01
7
4
4.0
23
NEC
MP5G1A1A-1 A/1996
(True)
5.4
0.02
4
3
7.0
24
Nokia
2160 EFR
(True)
8.8
0.00
0.1
0.0
0.2
25
Ericsson
KH668(2)/1997
(True)
5.6
0.08
14
16
7.7
26
Nokia
3390
(True)
5.0
0.82
56
60
18.9
27
NEC
DM2100 MP5G1B1-1A
(True)
8.4
0.01
0.1
0.0
0.0
28
Nokia
5120i
(True)
5.0
0.08
56
16
16.0
29
Motorola
Piper 763 62NNDBA/1995
(True)
5.1
0.00
121
30
3.1
30
Mitsubishi
GlOOe/1996
(True)
5.1
0.00
20
21
15.4
31
Qualcomm
QCP-820(2)
(True)
5.2
0.03
147
12
25.1
4-16
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in
Leachate (mg/L)
Cu
Fe
Pb
Zn
32
Ericsson
KH688/1997
(True)
5.4
0.01
17
1.1
7.3
33
Qualcomm
QCP-820(1)
(True)
5.0
0.00
0.6
1.5
20.5
34
Nokia
6160/1997
(True)
5.0
0.22
0.3
32
12.8
35
Nokia
2180/1995
(True)
9.0
0.00
0.0
0
0.1
36
NEC
TR5E800-26B Portable
(True)
5.0
0.02
0.1
65
2.4
37
Motorola
FR5 0/2001
(True)
5.0
0.01
0.3
6
85.0
38
Nokia
918+
(True)
8.4
0.00
0.0
0.0
0.1
39
LGIC
LGP-2300W
(True)
5.0
0.13
0.2
23
5.4
40
Nokia
6160i
(True)
5.0
0.17
0.3
31
13.4
41
Nokia
6160(2)/1997
(True)
5.0
0.06
0.4
16
19.6
42
NEC
MP5AF4-1A Portable/1995
(True)
5.0
0.10
0.1
61
15.1
43
Motorola
Micro Elite TAC Lite II
(True)
5.0
0.00
21
21
8.3
44
Motorola
Ultra Classic
(True)
5.0
0.00
0.1
24
1.6
45
Motorola
Profile 300
(True)
5.0
0.09
69
35
0.7
46
Motorola
76254NNFD A/1991
(True)
4.9
0.35
6
47
1.5
47
Motorola
Micro TAC Elite(2)
(True)
5.0
0.03
59
13
6.8
48
Motorola
Micro TAC Elite
(True)
5.0
0.05
78
18
5.8
49
Qualcomm
QCP-820(3)
(True)
5.0
0.00
74
1.4
10.7
50
Motorola
F09HLD8415BG/1991
(True)
5.0
0.00
23
48
15.7
51
Motorola
F09HGD8467CG/1989
(True)
5.0
0.06
20
40
1.5
52
Motorola
34106NNRS A/1995
(True)
5.0
0.01
66
19
6.0
53
Nokia
2160EFR(2)/1994
(True)
4.9
0.00
1.7
0.3
7.8
4-17
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.9 Keyboards
Keyboards were tested using both the large-scale and standard TCLP procedures. The large-
scale procedure was performed by combining four keyboards into a single extraction. Three
individual keyboards were disassembled, separated into their corresponding components, and
size reduced to less than 0.95 cm. A lOOg sample was then generated using the relative weight
percentage of each component. The 100 g samples were tested using the standard TCLP. The
large-scale TCLP test exceeded the 5 mg/L leachate concentration limit for lead. None of the
three keyboards tested using the standard TCLP exceeded the TC limit for lead with an average
lead concentration of 2.4 mg/L. The average composition of the keyboards and the test results
are shown in Figure 4-8 and Table 4-9 respectively.
Wires
PWB
Figure 4-8. Average Composition of Keyboards Tested
4-18
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-9. TCLP Results for Keyboards
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in Leachate
(mg/L)
Cu
Fe
Pb
Zn
1-4
NMB
RT101+
(L)
5.1
0.5
78
20
144
5
NMB
RT101+
(True)
5.1
0.114
16
2.3
164
6
NMB
RT101+
(True)
5.1
0.12
29
4.7
140
7
Gateway
SK-9921
(True)
5.1
0.02
116
0.1
73.6
Note: Sample 1-4 performed as a large scale composite test.
4-19
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.10 Computer Mice
A computer mouse although small does contain a small printed wire board and therefore may
contain lead due to the tin/lead solder used on these boards. Figure 4-9 shows the average
composition of 15 computer mice tested using the standard TCLP. As listed in Table 4-10 all 15
mice exceed the TC limit for lead with an average leachate concentration of 19.8 mg/L.
PWB
Figure 4-9. Average Composition of Mice Tested
4-20
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-10. TCLP Results for Mice
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in Leachate
(mg/L)
Cu
Fe
Pb
Zn
1
Dexxa
9MD/1987
(True)
5.0
3
9
10
16
2
Logitech
3F-HH/1987
(True)
5.0
1.3
75
38
16
3
Logitech
M30 9F
(True)
5.0
0.6
45
16
11
4
Manhatan
N/A
(True)
4.9
0.6
9
8
0.3
5
Microsoft
58264/1994
(True)
5.0
1.5
27
33
0.25
6
Logitech
M-C43
(True)
5.0
0.25
11
41
4.6
7
Logitech
M30 9F
(True)
5.0
0.32
15
6
1.5
8
Microsoft
58266/1994
(True)
5.0
0.68
51
13
0.34
9
Microsoft
92841
(True)
5.0
0.72
55
12
0.16
10
Microsoft
58269/1994
(True)
5.1
0.33
7
31
0.4
11
Microsoft
58269/1994
(True)
5.0
0.33
23
7
24
12
Microsoft
58269/1994
(True)
5.0
1.5
66
28
7.5
13
Microsoft
52695/1994
(True)
5.0
0.23
5
7
4
14
Microsoft
52695/1994
(True)
5.0
0.43
19
15
10
15
Microsoft
58269/1994
(True)
5.0
1.2
17
22
0.24
4.11 Remote Controls
Another small electronic item found in households is remote controls for electronic
equipment. As was performed with the cell phones, ten remote controls (without batteries) were
disassembled and placed in their entirety into 2L extraction vessels. The results of the modified
TCLP are presented in Table 4-11. The average leachate lead concentration was 12 mg/L with
all 10 samples exceeding the TC limit. The average composition of the remote controls is
presented in Figure 4-10.
4-21
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4
Leaching Results for Various Electronic Devices
PWB
Figure 4-10. Average Composition of Remote Controls Tested
Table 4-11. TCLP Results for Remote Controls
Sample
Number
Make
Model/Year
Method
pH
Metal Concentration in Leachate
(mg/L)
Cu
Fe
Pb
Zn
1
Memorex
4900
(True)
5.1
0.18
67
11
0.2
2
JVC
RM-RX130
(True)
5.0
0.23
13
10
3.5
3
SHARP
G0956GE
(True)
5.0
0.12
17
14
1.7
4
RCA
949T
(True)
5.0
0.37
20
33
2.2
5
Hitachi
Vt-RN 361A
(M.S.)
5.0
0.05
14
5.2
1.5
6
JVC
RM-C688
(M.S.)
4.9
0.03
3
7
0.2
7
JVC
RM-C689
(M.S.)
5.0
0.02
19
6
3.3
8
Magnavox
NE001UP
(M.S.)
4.9
0.23
24
5.7
3.3
9
RCA
N/A
(M.S.)
5.0
0.02
1
8
0.1
10
Hitachi
CLU-241
(M.S.)
5.0
0.07
14
20
3.0
4-22
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.12 Smoke Detectors
Smoke detectors are widely used in both businesses and households. Smoke detectors
contain PWBs and other materials as outlined in Figure 4-11. Nine smoke detectors (without
batteries) were tested, all using the standard TCLP procedure. The results are presented in Table
4-12. The average leachate concentration for lead was 23 mg/L with 7 of the 9 smoke detectors
exceeding the TC limit for lead.
Non Ferrous
6%
Ferrous
41%
PWB
53%
Figure 4-11. Average Composition of Smoke Detectors Tested
4-23 LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-12. TCLP Results for Smoke Detectors
Sample
Number
Make
Model
Method
pH
Metal Concentration in Leachate
(mg/L)
Cu
Fe
Pb
Zn
1
First Alert
SA301
(True)
5.0
0.24
28
44
1.0
2
Square D Company
SD631
(True)
5.0
0.15
64
51
8.0
3
FireX
X6-6
(True)
5.0
0.21
5
38
7.2
4
Square D Company
4919
(True)
5.0
0.05
42
14
10
5
Dicon Systems, Inc.
330L
(True)
5.1
0.22
26
17
2.4
6
FireX
G-6
(True)
5.2
0.11
67
17
0.7
7
FireX
FX 1020
(True)
5.2
0.01
5
1.10
190
8
Square D Company
EDG-4S
(True)
5.1
0.01
35
0.12
79
9
Honeywell
TC805C1000
(True)
4.9
0.12
0.8
24
0.2
4-24
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
4.13 Flat Panel Monitors
Eight different flat panel displays were tested. These devices were difficult to obtain as
they are relatively new on the market. Thus, six of the eight devices tested were of the same
model and manufacturer. All the devices were tested using the large-scale modified
methodology. Table 4-13 summarizes the results of the flat panel displays tested. Three of the
eight devices tested exceeded the TC limit for lead. The average lead concentration was 3.75
mg/L.
Figure 4-12. Average composition of Flat Panels
4-25
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
Chapter 4 Leaching Results for Various Electronic Devices
Table 4-13. TCLP Results for Flat Panel Displays
Sample
Number
Make
Model/Year
Method
pH
Metal Coneentration in Lcaehate (mg/L)
Cu
Fe
Pb
Zn
2
Dell Flatop
1703FP/2003
(L)
5.20
0.06
183
3
72
3
Apple Studio
M4551/1998
(L)
5.02
0.04
37
0.3
85
4
Dell Flatop
1703FP/2003
(L)
5.19
0.10
288
6.4
114
5
Dell Flatop
1703FP/2003
(L)
5.16
0.17
267
6
117
6
IBM
9513-AW1/1999
(L)
5.02
0.06
37
1.2
92
7
Dell Flatop
1703FP/2003
(L)
5.23
0.26
360
4.5
113
8
Dell Flatop
1703FP/2003
(L)
5.20
0.18
274
5.0
62
9
Dell Flatop
1703FP/2003
(L)
5.03
0.12
126
3.6
105
Note: Sample 1 not reported due to extraction vessel leakage of sample.
4-26
LEACHING RESULTS FOR VARIOUS
ELECTRONIC DEVICES
-------
5 Summary of Results
Twelve different types of electronic devices were tested using the TCLP, or one of two
other modified leaching methodologies designed to meet the intent of the TCLP. In many cases,
the lead concentrations in the TCLP leachates exceeded the 5 mg/L TC. Table 5-1 presents the
results in terms of the number of devices of a particular type that exceeded the TC threshold
concentration for lead. It is important to emphasize the objective of the research was not to
provide testing results that would be applicable for an entire class of devices; there are simply
too many manfucturers, makes and models of such devices to do this. The objective was to
provide an indication of whether classes of electronic devices have the potential to be TC
hazardous wastes.
Table 5-1. Summary of TCLP Pb Leaching Results
Device
Standard TCLP
Modified Large-Scale
TCLP
Modified Small-Scale
TCLP
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
Devices
Exceeding
5 mg-Pb/L
Devices
Tested
CPUs - Manual
Size Reduction
0
12
21
41
CPUs-
Mechanical Size
Red.
1
11
Computer
Monitors
9
9
Laptop
Computers
6
6
9
9
Printers
5
9
Color
Televisions
6
6
VCRs
7
8
Cellular Phones
28
38
10
14
Key Boards
0
3
1
1
Computer Mice
15
15
Remote Controls
4
4
6
6
Smoke Detectors
7
9
Flat Panel
Monitors
3
8
5-1
REFERENCES
-------
Chapter 5 Summary of Results
The results in Table 5-1 show that electronic devices containing lead do have the potential to
leach lead concentrations above the toxicity characteristic limit of 5 mg/L when leached using
the TCLP extraction solution. Every device type leached lead above 5 mg/L in at least one test.
It is important to note that the size and heterogeneity of many of the devices tested precluded the
use of the standard TCLP under the constraints of the project scope and budget. Thus modified
testing procedures were used for many of the devices. These modified tests were designed in an
effort to meet the intent of the TCLP. The same leaching solution and the same liquid to solid
ratio were used. These modified tests are not, however, the TCLP as defined in 40 CFR 261. In
the case of the large-scale modified TCLP, the samples were not size reduced. During method
development it was thought that this approach would tend to underestimate the amount of
pollutant leached, as the purpose of the size reduction step is to reach equilibrium concentrations
more quickly. Lead may leach more when the materials are not size reduced because of
chemical conditions resulting from the other components of the device.
The testing of entire color computer monitors and color televisions confirms previous
experiments that show color CRTs to leach lead above the toxicity characteristic concentration.
Small electronic devices that contain a PWB with lead solder often leach above 5 mg/L of lead
using the standard TCLP. When the larger devices were tested using the modified large-scale
method, they often exceeded the TC limit for lead. The amount of ferrous metal present in some
of these devices may result in less lead being leached if the standard TCLP were to be
performed. In the detailed study of computer CPUs (see Chapter 3), size-reduced computer
CPUs leached less lead that the same model of CPU leached with the large-scale modified
method (it should be noted that this particular model did not fail TCLP even for the large
method, while many of the other CPU models tested did). On the other hand, in limited testing
of laptops using the standard approach and the modified approach, the TC limit was always
exceeded for lead. The difference between leaching results for the two devices likely resulted
from the greater ferrous metal content (68%) of the computer CPUs relative to the laptops (7%).
The utility of the modified test results will have to be determined by the regulatory agencies.
One could argue that the modified test procedures are not true TCLP results and thus are not
applicable. However, the modified testing procedures meet the intent of the TCLP which is to
provide a conservative estimate of the amount of pollutant that leaches from a waste in the
simulated landfill leachate designed for the TCLP. At the very least, this research contributes to
the debate regarding the appropriate utilization of leach testing of solid wastes. Finally, it is
noted that this report does not attempt to interpret the TCLP results with respect to
environmental impacts, such as impact on leachate quality at landfills. The authors are
conducting other research to address this issue.
5-2
REFERENCES
-------
6 References
Akatiff, C., 2002, "Is this Ban Really Necessary? A Critical Investigation of the CRT Ban,"
Lake Tahoe, Nevada, Solid Waste Association of North American (SWANA) Western
Regional Symposium.
Basel Action Network (BAN), 2002, "Exporting Harm: The High-Tech Trashing of Asia,"
Seattle, Washington, (PDF copy available at:
http://www.ban.org/E-waste/technotrashfinalcomp.pdf). March, 2002.
Bromine Science and Environmental Forum (BSEF), 2000, "An Introduction to Brominated
Flame Retardants," Brussels, Belgium, (PDF copy available at:
http://www.ebfrip.org/download/weeeqa.pdf). August, 2001.
Environment Australia, 1999, "Hazardous Status of Waste Electrical and Electronic Assembles
or Scrap," Guidance Paper, Department of the Environment and Heritage, Commonwealth of
Australia, (PDF copy available at: http://www.ea.gov.au/industry/hwa/pubs/scrap.pdf).
November 2002.
Five Winds International (FWI), 2001, "Toxic and Hazardous Materials in Electronics: An
Environmental Scan of Toxic and Hazardous Materials in IT and Telecom Products and
Waste," Final Report, submitted to Environment Canada, National Office of Pollution
Prevention and Industry Canada, Computers for Schools Program, 219 ch. Vanier, Aylmer,
Quebec J9HIY5, Canada, (PDF copy at:
http://www.epsc.ca/pdfs/IT Telecom Haz Mats ENG.pdf). September 2001.
Global Futures Foundation (GFF), 2001, "Computer E-Waste and Product Stewardship: Is
California Ready for the Challenge?" Report for the US Environmental Protection Agency
Region IX, San Francisco, California, (PDF copy available at:
http://www.globalfutures.org/e-waste.pdf). November 2002.
Kendall, D., 2003, "Toxicity Characteristic Leaching Procedure and Iron Treatment of Brass
Foundry Waste," Environmental Science and Technology, Vol. 37, pp. 367-371.
Meng, X., G. Korfiatis, C. Jing, C. Christodoulatos, 2001, "Redox Transformations of Arsenic
and Iron in Water Treatment Sludge During Aging and TCLP Extraction," Environmental
Science and Technology, Vol. 35, pp. 3476-3481.
Microelectronics and Computer Technology Corporation (MCC), 1996, Electronics Industry
Environmental Road Map, Document No. MCC-ECESM-001-96, pp. 249, Austin, Texas.
Musson, S., Y. Jang, T. Townsend, I. Chung, 2000, "Characterization of Lead Leachability from
Cathode Ray Tubes Using the Toxicity Characteristic Leaching Procedure," Environmental
6-1
REFERENCES
-------
Chapter 6 References
Science and Technology, Vol. 34, pp. 4376-4381.
National Safety Council (NSC), 1999, "Electronic Product Recovery and Recycling Baseline
Report: Recycling of Selected Electronic Products in the United States," National Safety
Council's Environmental Health Center, Washington, D.C.
Nordic Council of Ministers (NCM), 1995, "Waste from Electrical and Electronic Products: A
Survey of the Contents of Materials and Hazardous Substances in Electric and Electronic
Products," TemaNord 1995:554, pp. 9-52, Copenhagen, Denmark.
Silicon Valley Toxics Coalition (SVTC), Californians Against Waste (CAW), and The Materials
for the Future Foundation, 2001, "Poison PCs and Toxic TVs: California's Biggest
Environmental Crisis That You've Never Heard Of," pp. 2-14, San Jose, California, (PDF
copy available at: http://www.svtc.org/cleancc/pubs/ppc-ttvl.pdf). June, 2001.
Schmidt, C.W., 2002, e-Junk Explosion, Environ Health Perspect, Vol. 110(4), pp. A188-94.
Snoeyink, V, D. Jenkins, 1980, Water Chemistry, John Wiley & Sons, Inc., New York, pp. 363-
378.
Townsend, T., Y. Jang, T. Tolaymat, J. Jambeck, 2001, "Leaching Tests for Evaluating Risk in
Solid Waste Management Decision Making: Year 1," Florida Center for Solid and Hazardous
Waste Management, Gainesville, FL, (PDF copy available at:
http://www.ees.ufl.edu/homepp/townsend/Research/Leach/Leach Yrl PDF). March, 2003.
Townsend, T., S. Musson, Y. Jang, I. Chung, 1999, "Characterization of Lead Leachability from
Cathode Ray Tubes Using the Toxicity Characterization Leaching Procedure," Florida
Center for Solid and Hazardous Waste Management. Report # 99-5, pp. 10-16, Gainesville,
FL, (PDF copy available at: http://www.floridacenter.org/publications/lead teachability 99-
5.pdf). January, 2001.
United States Environmental Protection Agency (USEPA), 1996, "Test Methods for Evaluating
Solid Waste," SW-846, 3rd Ed., Office of Solid Waste and Emergency Response,
Washington, DC.
United States Environmental Protection Agency (USEPA), 1997, Rules and Regulations, Federal
Register, Vol. 62, No. 91, The Code of Federal Regulations, Title 40, Part 261, pp. 25998-
26040, Office of the Federal Register, National Archives and Records Administration,
Washington, DC.
United States Environmental Protection Agency (USEPA), 1998, Rules and Regulations, Federal
Register, Vol. 63, No. 100, The Code of Federal Regulations, Title 40, Part 261, pp. 28556-
28604, Office of the Federal Register, National Archives and Records Administration,
Washington, DC.
United States Environmental Protection Agency (USEPA), 1999, "Identification and Listing of
Hazardous Waste," The Code of Federal Regulations, Title 40, Ch. 1, Part 261, Office of the
6-2
REFERENCES
-------
Chapter 6 References
Federal Register, National Archives and Records Administration, Washington, DC.
United States Environmental Protection Agency (USEPA), 2001, "Land Disposal Restrictions:
Summary Requirements," EPA530-R-01-007, Offices of Solid Waste and Emergency
Response & Enforcement and Compliance Assurance, Washington, DC (PDF copy available
at: http://www.epa.gov/epaoswer/hazwaste/ldr/ldr-sum.pdf). October, 2001.
United States Environmental Protection Agency (USEPA), 2002a, "Municipal Solid Waste in the
United States: 2000 Facts and Figures," EPA530-R-02-001, Office of Solid Waste and
Emergency Response, Washington, DC.
United States Environmental Protection Agency (USEPA), 2002b, Proposed Rules, Federal
Register, Vol. 67, No. 113, pp. 40508-40528, Office of the Federal Register, National
Archives and Records Administration, Washington, DC.
Vann, Kevin N., 2003, Leaching behavior of personal computer central processing units (CPUs)
using standardized and modified toxicity characteristic leaching procedure (TCLP) tests.
[Unpublished Master's Thesis], Gainesville, Florida: University of Florida.
Yang, G., 1993, "Environmental Threats of Discarded Picture Tubes and Printed Circuit
Boards," Journal of Hazardous Materials, Vol. 34, pp. 235-243.
6-3
REFERENCES
-------
Appendix A
Table A-l. Additional CPU Metals Analysis
S;ini|ik'
\unilvi'
¦M?
As
Ba
Cil
( i-
Sc
1
<0.06
<.011
1.456
0.0166
0.056
<0.008
2
<0.06
<.011
2.036
0.0117
0.036
<0.008
3
<0.06
<.011
1.051
0.0158
0.052
<0.008
4
<0.06
<.011
0.070
0.0153
0.061
<0.008
5
<0.06
<.011
0.080
0.0148
0.060
<0.008
6
<0.06
<.011
0.057
0.0142
0.056
<0.008
7
<0.06
<.011
0.078
0.0126
0.020
<0.008
8
<0.06
<.011
0.109
0.0085
0.024
<0.008
9
<0.06
<.011
1.582
0.0171
0.046
<0.008
10
<0.06
<.011
1.964
0.0142
0.037
<0.008
11
<0.06
<.011
1.636
0.1701
0.058
<0.008
12
<0.06
<.011
0.377
0.0188
0.017
<0.008
13
<0.06
<.011
1.291
0.0115
0.038
<0.008
14
<0.06
<.011
0.162
0.0137
0.037
<0.008
15
<0.06
<.011
0.121
0.0136
0.045
<0.008
16
<0.06
<.011
0.479
0.0052
0.182
<0.008
17
<0.06
<.011
1.908
0.0145
0.044
<0.008
18
<0.06
<.011
0.860
0.0215
0.041
<0.008
19
<0.06
<.011
0.352
0.0053
0.017
<0.008
20
<0.06
<.011
0.369
0.0029
0.015
<0.008
21
<0.06
<.011
1.747
0.0225
0.039
<0.008
22
<0.06
<.011
1.918
0.0171
0.052
<0.008
23
<0.06
<.011
1.231
0.0057
0.020
<0.008
24
<0.06
0.1076
1.323
0.0375
0.023
0.022
25
<0.06
0.1036
2.332
0.0048
0.032
0.027
26
<0.06
0.1351
1.505
0.0024
0.032
0.022
27
<0.06
<.011
0.511
0.0008
0.028
<0.008
28
<0.06
<.011
1.366
0.0050
0.058
<0.008
29
<0.06
<.011
1.655
0.0174
0.047
<0.008
30
<0.06
<.011
0.213
0.0061
0.019
<0.008
31
<0.06
<.011
0.452
0.0176
0.030
<0.008
32
<0.06
<.011
0.489
0.0043
0.028
<0.008
33
<0.06
<.011
2.197
0.0186
0.050
<0.008
34
<0.06
<.011
1.185
0.0241
0.029
<0.008
35
<0.06
<.011
0.491
<0.0003
0.027
<0.008
36
<0.06
<.011
0.721
0.0017
0.018
<0.008
37
<0.06
0.0656
1.210
0.0088
0.033
0.036
38
<0.06
0.0930
1.477
0.0073
0.022
0.028
39
<0.06
0.1097
0.312
0.0125
0.025
0.045
40
<0.06
0.0200
1.026
0.0089
0.029
0.020
A-l
Appendix A
-------
Appendix A
Siiniplo
\umlvi'
As
liii
Cd
( i-
Sc
41
<0.06
<.011
2.342
0.0013
0.050
<0.008
42
<0.06
<.011
0.087
0.0025
0.008
<0.008
43
<0.06
<.011
1.452
0.0155
0.044
<0.008
44
<0.06
<.011
1.186
0.0275
0.064
<0.008
45
<0.06
<.011
0.890
0.0176
0.018
<0.008
46
<0.06
<.011
0.482
0.0025
0.011
<0.008
47
<0.06
<.011
0.193
0.0085
0.021
<0.008
48
<0.06
<.011
<0.001
0.0079
0.015
<0.008
49
<0.06
0.0033
0.099
0.0019
0.037
<0.008
50
<0.06
0.0035
0.062
0.0022
0.009
<0.008
51
<0.06
-0.0008
0.094
0.0094
0.009
<0.008
52
<0.06
-0.0009
0.624
0.0014
0.012
<0.008
53
<0.06
-0.0007
1.046
0.0037
0.011
<0.008
54
0.24
-0.0094
0.184
0.0087
0.026
<0.008
55
<0.06
-0.0061
0.179
0.0045
0.025
<0.008
56
<0.06
-0.015
0.593
0.0078
0.021
<0.008
57
<0.06
-0.0011
0.314
0.0007
0.014
<0.008
58
<0.06
-0.0163
0.262
0.0131
0.024
<0.008
59
<0.06
-0.0056
0.141
0.0056
0.025
<0.008
60
<0.06
-0.0091
0.324
0.0079
0.031
<0.008
61
<0.06
0.0011
0.298
0.0014
0.017
<0.008
62
<0.06
-0.0027
0.234
0.0052
0.087
<0.008
63
<0.06
0.0009
0.243
0.0009
0.017
<0.008
64
<0.06
0.0014
0.151
0.0011
0.050
<0.008
A-2
Appendix A
------- |