&EPA
United States
Environmental Protection
Agency
Release of Micronized Copper Particles from
Pressure-treated Wood Products
-------
-------
Release of Micronized Copper
Particles from Pressure-treated
Wood Products
EPA Report No. EPA/600/R-14/365
Final Report Prepared by:
William E. Flatten, III1, Todd P. Luxton2*, Tammie Gerke1, Steve Harmon2,
Nicholas Sylvest1, Karen Bradham3, and Kim Rogers3
1 Pegasus Technical Services, Inc. Cincinnati, OH
2National Risk Management Research Laboratory, U.S. Environmental Protection Agency
Cincinnati, OH,
3National Exposure Research Laboratory, U. S. Environmental Protection Agency
Research Triangle Park, NC
"Corresponding author: Todd Luxton
Tel: 513-569-7210
E-mail address: Luxton.Todd@epa.gov
-------
Forward
The US Environmental Protection Agency (US EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the Agency
strives to formulate and implement actions leading to a compatible balance between human activities
and the ability of natural systems to support and nurture life. To meet this mandate, US EPA's research
program is providing data and technical support for solving environmental problems today and building
a science knowledge base necessary to manage our ecological resources wisely, understand how
pollutants affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks from
pollution that threaten human health and the environment. The focus of the Laboratory's research
program is on methods and their cost-effectiveness for prevention and control of pollution to air, land,
water, and subsurface resources; protection of water quality in public water systems; remediation of
contaminated sites, sediments and ground water; prevention and control of indoor air pollution; and
restoration of ecosystems. NRMRL collaborates with both public and private sector partners to foster
technologies that reduce the cost of compliance and to anticipate emerging problems. NRMRL's
research provides solutions to environmental problems by: developing and promoting technologies that
protect and improve the environment; advancing scientific and engineering information to support
regulatory and policy decisions; and providing the technical support and information transfer to ensure
implementation of environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by US EPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
Cynthia Sonich-Mullin, Director
National Risk Management Research Laboratory
-------
EPA600/R-14/365
Abstract
Micronized copper pressure-treated lumber (PTL) has recently been introduced to the consumer market as
a replacement for ionized copper PTL. The presence of particulate rather than aqueous copper raises
concerns about the exposure of humans as well as the environment to the particles. Two common pathways
of exposure, leaching during contact with water and transfer during physical contact, were investigated to
gage potential human and environmental exposure during intended use of the product.
Characterization, leaching tests, and wipe tests were conducted on two representative formulations of
micronized copper PTL (micronized copper azole or MCA) to quantify the levels of copper present in the
treated material and the amount of copper released during use as well as to determine the form (particle or
ion) of the copper after it was released. Additionally, an ionized copper pressure-treated wood (alkaline
copper azole or ACA) was tested for comparison. The characterization showed that copper carbonate is the
primary particle form in the MCA treated wood, but other forms are also present, particularly in the MCA-
1 formulation, which contained a large amount of organically complexed copper. Microscopy showed that
MCA-1 contained particles roughly half the size of MCA-2.
The leaching results indicate that mostly (> -95%) ionic copper is released from the MCA wood and that
the particulate copper that was released is attached to cellulose and not free in solution. A small number of
particles were captured separate from the cellulose on a 10-kiloDalton (kDa) filter, but the quantity amounts
to less than 1% of the total copper leached from the wood during the test. Comparing the MCA wood to the
ACA wood, MCA released significantly less copper than the ACA, leading to a lower potential impact on
the environment, though ACA has already been shown to have a negligible impact (Forest Products
Laboratory, 2000). The wipe tests were a surrogate for hand contact with the treated wood, developed by
the Consumer Product Safety Commission for gauging exposure of children using playground equipment
constructed with PTL. The results show that the MCA and ACA wood release approximately the same
amount of copper with each contact and that the amount of copper released is high initially, but decreases
to a constant level after being wiped 2-3 times. The boards which were left outdoors, exposed to the
elements, had a higher plateau level than those tested indoors, leading to the conclusion that the exposure,
likely to precipitation, causes migration of the copper to the surface. During the initial period of high release,
contact with the wood causes far more depletion of the copper than any type of environmental exposure.
During testing, MCA-1 released slightly more copper than MCA-2, possibly due to the smaller size of the
copper particles in the MCA-1 formulation or the increased concentration of organically complexed copper.
Based on previous research on the effect of copper dose and physiological effects, the copper transfer levels
observed in this study could impact children under the age of 8 years, particularly those 1-3 years of age, if
completely ingested; reaching or exceeding the Tolerable Upper Intake Levels (TUIL) recommended by
the Institute of Medicine. Above the TUILs, only mild symptoms could be experienced depending on
dietary levels of copper, while severe acute and chronic symptoms are very unlikely during normal use. It
should be noted that the copper values found from the newer, micronized formulations are comparable to
those found from the older, ionized formulation. In summary, the particulate copper released from
micronized copper PTL constituted a small fraction (< ~5%) of the total released. The total copper released
was less than or comparable to the current aqueous formulation available to consumers.
Keywords
Pressure-treated Wood, Micronized Copper, Nanoparticle, Nanocopper, Copper Toxicity, Copper
Leaching, Copper Exposure, Micronized Copper Azole, Alkaline Copper Azole, Aqueous Copper,
Copper Speciation, CCA Alternatives
-------
This page intentionally left blank.
in
-------
Table of Contents
Forward i
Abstract ii
Keywords ii
Table of Contents iv
Table of Figures vi
List of Tables viii
1 Introduction 1
2 Materials and Methods 3
2.1 Leaching Experiment 3
2.2 Wipe Experiment 4
2.3 Analytical Methods 5
3 Characterization 7
4 Leaching Experiment 14
4.1 Copper Release 14
4.2 Mass Balance 17
4.3 Copper Speciation 19
4.3.1 X-ray Absorption Fine Structure Spectroscopy 19
4.3.2 Microscopy 23
4.4 Environmental Impact 25
5 Wipe Experiment 27
5.1 Copper Release 27
5.2 Copper Speciation 30
iv
-------
5.2.1 Microscopy 30
5.2.2 X-ray Absorption Fine Structure Spectroscopy (XAFS) 35
5.3 Copper Toxicity Assessment 38
6 Conclusion 41
7 References 43
-------
Table of Figures
Figure 3-1: X-Ray diffraction patterns for the as-purchased lumber. The dashed lines labeled 1 and 2
mark the location of diagnostic diffraction peaks for copper carbonate. The dash line labeled 3 indicates
the location of a diffraction peak associated with the cellulosic wood structure 9
Figure 3-2: First derivative XANES spectra for the three wood samples (MCA-1, MCA-2, and ACA) and
reference compounds. The highlighted blue area shows the location of the first derivative peaks that
correspond to copper carboxylate complexes 10
Figure 3-3: Micrographs of (A) BSE SEM MCA-1 and (B) Bright Field TEM MCA-2 11
Figure 3-4: Histogram, cumulative abundance, and box plots for particle length, width and aspect ratio
(length/width) for Cu particulates identified in the SEM and TEM images for MCA-1 and MCA-2,
respectively 13
Figure 4-1: Leaching Experiment: 24-Hour Wood Block Leachate Fractionation 14
Figure 4-2: Leaching Experiment: 72-Hour Wood Block Leachate Fractionation 15
Figure 4-3: Leaching Experiment: 24-Hour Sawdust Leachate Fractionation 15
Figure 4-4: Leaching Experiment: 72-Hour Sawdust Leachate Fractionation 16
Figure 4-5: First derivative Cu Kot XANES spectra for the materials retained on the filter papers after
sequential filtration of the supernatant from a sample of MCA-2 lumber leached with SPLP solution for
72 h. A) Indicates the position of the Cu1+ first derivative peak, B) indicates the low energy shoulder of
copper carbonate (Q^COsCOHh) first derivative peak, and C) indicates a first derivative shoulder and
peak position for copper complexed with organic acids 21
Figure 4-6: Copper speciation and mass of cooper leached from wood blocks leached with the SPLP
solution for 72 h as a function of filter particle cutoff. Data presented is the relative abundance of each
copper phase that was retained on the filter paper after sequential filtration of the supernatant and the
mass per unit area of copper that was leached from the lumber. The first column in each graph refers to
the original copper species distribution in the as-purchased materials 22
Figure 4-7: BSE SEM Micrographs of particulates retained on a 2.5 jam membrane (A) MCA-1 and (B)
MCA-2. The white square outline indicates the location of where EDX spectra were collected to
determine elemental composition of the particles 24
Figure 4-8 BSE SEM Micrographs and EDX chemical maps of particulates retained on a 0.45 mm
membrane (A) MCA-1, (B) EDX copper map, (C) MCA-2, (D) EDX copper map 24
Figure 4-9 BSE SEM Micrographs of particulates retained on a 10 kDa membrane from the MCA-2
sample 25
Figure 5-1: Copper Released by Wiping - Outdoor Weathering 27
vi
-------
Figure 5-2: Copper Released by Wiping - Freeze/Thaw Weathering 28
Figure 5-3: Copper Released by Wiping -Weathering Comparison 29
Figure 5-4: Copper Released by Wiping - Outdoor Weathering -After Sanding 30
Figure 5-5: SEM micrographs of materials collected on 0.45 jam filter papers. MCA-1 (A), MCA-2 (B),
andACA(C) 31
Figure 5-6: BSE SEM micrograph of MCA-1 (A), EDS map of Cu from the same sample (B), and
histogram and cumulative abundance plots for particle length, width, and aspect ratio (length/width) (C).
32
Figure 5-7 BSE SEM micrograph of MCA-2 (A), EDS map of Cu from the same sample (B), and
histogram and cumulative abundance plots for particle length, width, and aspect ratio (length/width) (C).
33
Figure 5-8: BSE SEM Micrographs and EDX chemical maps of particulates wiped from the surface of
three different pressure-treated lumber samples and deposited onto a 0.45 jam filter paper. MCA-1 (A),
MCA-2 BSE micrograph and EDX maps (B, C, and D), respectively. ACA BSE micrograph and EDX
maps (E, F and G), respectively 35
Figure 5-9: Cu Kot XANES spectra for the as-purchased pressure-treated lumber and the material
dislodged from the wood surface and retained on a 0.45 jam filter after 1 month (Wipe 1) and two months
(Wipe 2). MCA-1 (A), MCA-2 (B), ACA (C). Dashed lines indicate the energy positions associated with
Cu1+ and malachite 36
Figure 5-10: Comparison of the percent distribution of particle size between the as-purchased MCA-2 and
the MCA-2 Wipe 1 samples. Length (A) Width (B) 38
vn
-------
List of Tables
Table 2-1: Leaching Experiment Testing Conditions. Columns indicate the different wood types, form,
chemistry, and duration of the experiments 4
Table 3-1: Chemical composition of the four wood types as determined by microwave assisted acid
digestion and ICP-OES analysis. Averages are based on seven replicate samples. Additional elements
analyzed included Al, Ag, As, Ce, Co, Cr, Fe, K, Mo, Ni, P, Pt, Sb, Se, Si, and V, however there were no
detectable quantities in the samples 7
Table 3-2: Volumetric concentration of copper in the 4 selected wood products. Calculated using the
elemental concentrations from Table 1 and mass measurement of known volumes of lumber 8
Table 3-3: Results from the linear combination fitting of the first derivative XANES spectra. Values are
presented as average/standard deviation 10
Table 3-4: Average particle size for MCA-1 and MCA-2. Values are presented as average/standard
deviation 11
Table 4-1: Summary of Leaching Results with SPLP Water Chemistry 16
Table 4-2: Mass Balance of Copper Released from Wood Blocks during Leaching 18
Table 4-3: Particulate Copper Released and Copper Associated with Each Filtration Fraction 19
Table 4-4: Copper speciation (percent abundance) and mass of cooper leached from wood blocks leached
with the SPLP solution for 72 h as a function of filter particle cutoff. Data presented is the relative
abundance of each copper phase that was retained on the filter paper after sequential filtration of the
supernatant. Unleached refers to the as-purchased materials 22
Table 5-1: Average Release of Copper During Wiping Experiments 29
Table 5-2: Average particle size and aspect ratio of micronized copper particles dislodged from the wood
surface after wiping with a polyester cloth as determined by analysis of BSE SEM micrographs. Previous
data for the as-purchased pressure-treated lumber is presented for comparison. Aspect ratio is equal to the
ratio between the particle length and width 34
Table 5-3: Linear combination fitting results for Cu Kot XANES first derivative spectra for materials
wiped off of the treated wood samples exposed to the environment after 1 and 2 months of exposure
retained by a 0.45 jam filter 37
Table 5-4: Summary of Copper Toxicity Levels 39
Vlll
-------
1 Introduction
Pressure-treating wood is a popular method for preserving lumber from rot caused by exposure to outdoor
conditions, particularly insects and microbial agents. Prior to 2003, the majority of the pressure-treated
lumber (PTL) in the U.S. was produced using Chromated Copper Arsenate (CCA) aqueous formulations
that were forced into the wood under pressure. The copper and arsenic in the solution combined to provide
protection from bacterial, fungal, and insect decay while the chromium acted as a fixing, or binding, agent
and UV protectant. Concerned about human exposure to the chemical residues, mainly arsenic, the
Consumer Product Safety Commission (CPSC) investigated the potential human exposure, focusing mostly
on children through hand contact with pressure-treated lumber at playgrounds. After a series of reports from
the CPSC detailing the potential exposure level to children, the wood preservation industry voluntarily
agreed to remove the CCA formulation from the majority of wood intended for residential use.
Several alternatives to the CCA formulation were developed as a replacement. Based around copper, the
formulations used ionic copper as the primary insecticide and fungicide along with a co-biocide to provide
additional resistance. Examples of these new formulations include acid copper chromate, alkaline copper
quaternary, and aqueous alkaline copper azole (ACA) (Lebow, 2004). These formulations were effective
in preserving the wood, but released significantly more ionic copper into the surrounding environment than
the CCA formulations (Forest Products Laboratory, 2000). The increased release of ionic copper resulted
in the degradation of metal fasteners and other metal in the area of the treated wood. Corrosion of galvanized
surfaces from the ionic copper was targeted as the primary mechanism responsible corrosion of metal
fastener and subsequent structural failure (Forest Products Laboratory, 2000). A newer formulation using
micronized copper particles was developed to reduce the release of ions and, therefore, the corrosive effects,
while still maintaining the desired preservation characteristics.
Micronized copper formulations (micronized copper azole (MCA) and micronized copper quaternary) were
developed to address the issues of the ionic copper formulations. They offer many advantages over the
previous systems including: reduced mass of copper required to produce the same protection as the ACA
treatment, reduced corrosion of the treating plant equipment and metal fasteners, reduced mold growth on
the treated wood, and less copper leaching from the material. For example, research has shown that the
micronized formulas are as effective as the ACA and CCA treated samples for wood preservation while
retaining a greater quantity of copper over the field trial time period (Cookson et al., 2010; Yu etal.,2Q\V).
The difference between the formulations is the use of "micronized copper" as the active ingredient. The
micronized formula is comprised of copper carbonate particles ranging in size from several microns to a
few nanometers in size (Evans et al, 2008).
Micronized copper pressure-treated lumber currently accounts for -80% of the pressure-treated lumber
being produced and sold within the United States and Canada (Cushman, 2009). The use of micronized
copper particles in residential applications may eventually lead to the release of copper nano sized particles
into the environment. The fate and potential impact of these particles is currently unknown. Concerns over
the fate of these particles were raised in a petition by the International Center for Technology Assessment
to the U.S. EPA, requesting further testing of the material. Thus, in a joint effort between the U.S. EPA and
the U.S. CPSC, the release of micronized copper particles and the potential health and environmental effects
of their release was examined.
In an investigation parallel to this report, commissioned by the CPSC, the types of PTL that utilized
micronized copper were determined. The hierarchy of manufacturers, distributors, and retailers is complex
and there are a number of distinct treatment formulations. A single commercial entity can act as a
1
-------
manufacturer, distributor, and/or retailer while also supplying products to other companies, which, in turn,
can participate in any of those roles. At each stage in the manufacturing and distribution of the wood
product, there is potential for a modification or addition to the treatment formula, resulting in a new
variation. Manufacturers vary the concentrations of the solutions and, in addition to the copper and biocide
chemicals, add a number of additives, producing a similar but separately distinct product. During the
investigation, two primary manufacturers were identified as supplying the entire industry with the pure
micronized copper treatment solutions (containing no biocides or additives) Lonza (formerly Arch
Chemical), and Koppers Performance Chemicals (formerly Osmose, Inc. Wood Preservation Group). The
chemical formulations and treatment technologies are then subsequently licensed by secondary
manufacturers who generate the pressure-treated lumber product. Because of the complexity of the industry
and the near impossible task of locating and testing all of the various treatment formulations, the current
investigation focuses on two readily available micronized copper azole pressure-treated lumber sources
from the two different manufacturers of the treatment solution (MCA-1 and MCA-2) and an aqueous
alkaline copper azole treated sample (ACA) and an untreated material (UTW) for comparison.
-------
2 Materials and Methods
To investigate the release of micronized copper, as-purchased materials were thoroughly characterized and
two experiments were conducted to evaluate the release of copper under different experimental conditions:
leaching and wiping. The leaching test simulated the effects of rainfall and water on the release of copper
from wood. The wipe test simulated direct human contact with the pressure-treated lumber during intended
uses, such as construction using PTL, sitting, walking, or climbing on the PTL wood.
Four wood types were used in the current studies. Two micronized copper treated samples (MCA-1 and
MCA-2), an aqueous copper treated sample (ACA), and an untreated wood sample (Untreated/UTW). All
three pressure-treated woods types utilized Southern Yellow Pine as the wood type, and a Spruce wood for
the untreated sample. Southern Yellow Pine and Spruce woods used for construction do not refer to a
specific wood species, but are used as generic terms for different wood types with similar structural
properties (U.S. Forest Service, 1936). The four wood samples were obtained from national hardware
retailers and wood suppliers within 50 miles of Cincinnati, OH. Market research was used to select two
micronized copper formulations from different manufacturers of the active biocidal component, copper
carbonate, and an aqueous copper pressure-treated lumber source. Untreated Southern Yellow Pine was not
available in the area; therefore an untreated indoor wood source (spruce) was used as the untreated material.
Samples were purchased in bulk to help minimize variation between individual boards and prevent
variations that might result from a change in product formulation or manufacturer. All of the wood samples
used in the current studies were recommended for above ground use by the manufacturer. Pressure-treated
lumber rated for ground contact use has a significantly higher concentration of copper (approximately 5x).
This is related to the increased exposure to insects, mold, or fungus with the wood. While the concentration
is greater for the ground contact materials, the above ground material represents a more direct route of
contact and exposure. Further, the market share of lumber sold is above ground use material. Therefore,
based on market share and more importantly direct exposure route, the current study focused on materials
rated for above ground use. The biocidal ingredients listed on the manufacture's label included copper
carbonate (Q^CC^OFfh), tebuconazole (CieF^ClNsO), and propiconazole (CisHivChNsCh). The release
of the two organic azole pesticides listed were not investigated in the current study since the focus was on
the potential release of nanomaterials and copper from lumber.
2.1 Leaching Experiment
Leaching experiments were conducted to determine the quantity of total and micronized copper released
when in contact with aqueous solutions. Three different solution chemistries were examined: 0.01 M NaCl
solution (pH 7), 0.01 M NaNOs solution (pH 7), and Super Q water adjusted with synthetic precipitation
leaching procedure (SPLP) solution (pH 4.2). The three solutions were chosen to address the impact of
anions (Cl~ and NOs) and simulated acid rain on the wood. The wood samples were tested as wood blocks
and sawdust, to evaluate the role of particle size, at two different mixing durations (24 and 72 h). In total,
48 conditions were tested in triplicate with a fourth replicate set aside for imaging and speciation (see Table
2-1). The sawdust was generated by sanding the cross section of 2x6 boards with a random orbital sander
and 80 grit sand paper. The specific surface area of the sawdust was measured by nitrogen gas adsorption.
The measured specific surface areas were 1.23, 1.21, 1.18, and 1.2 m2 g"1 for MCA-1, MCA-2, ACA, and
Untreated, respectively.
-------
Table 2-1: Leaching Experiment Testing Conditions. Columns indicate the different wood types,
form, chemistry, and duration of the experiments.
Wood Type
MCA-1
MCA-2
ACA
Wood Form
Wood Block
Sawdust
Solution Chemistry
0.01MNaCl-pH7
0.01MNaNO3-pH7
Super Q SPLP - pH 4.2
Duration
24 h
72 h
Untreated Lumber
(UTW)
Wood blocks, approximately 10 g in weight, and sawdust, approximately 1 g, were placed in acid-washed,
250 mL plastic containers and 100 mL of leaching solution was added. The containers were capped and
placed on a reciprocating (back and forth motion) shaker for the desired duration. After the elapsed time
period, the solution was fractionated to determine the form of copper in solution. A 10 mL subsample of
the unfiltered solution was taken for total elemental analysis, and the remainder of the solution was passed
through 2.5 um, 0.45 um and 10 kDa filters, sequentially. After each filtration step, a 10 mL sample was
collected for elemental analysis and the filter membrane was retained for subsequent acid digestion and
elemental analysis. The filtrates and the filters were digested with nitric acid or nitric and hydrochloric acid
following US EPA Methods 3015a and 305 la, respectively, and analyzed via inductively couple plasma
optical emission spectroscopy (ICP-OES) using US EPA Method 6010 (U.S. EPA, 2007).
2.2 Wipe Experiment
The wipe experiment was conducted to determine the release of micronized copper during intended uses.
Three different sets of conditions were tested. A set of boards ("Outdoor") were left outdoors exposed to
environmental conditions at the U.S. EPA Center Hill research facility in Cincinnati, Ohio and sampled at
0, 14, 34, 70, 97, 140, 260, and 399 days. Another set ("Freeze/Thaw") was put through several cycles of
freezing at -80 °C for 24 hours and thawing at room temperature for 48 hours, simulating thermal expansion
cycles from being outdoors in cold climates. The boards were moistened once each cycle, just prior to
returning to the freezer. Samples were collected at 4, 8, 12, and 24 cycles. Procedural and analytical issues
resulted in the loss of reportable data for wipe measurements conducted prior to the first Freeze/Thaw cycle.
A third set of boards ("No Weathering") were wiped repeatedly without having undergone any surface
modification. The boards were wiped a total of 12 times over two days, 6 per day. This set was used to
determine the baseline effect of just wiping the boards. In all instances, eight-foot boards were divided into
four sections and each section was wiped using a polyester cloth (details below). Samples from the first
three sections were digested and analyzed for copper. Samples from the fourth section were preserved for
scanning electron microscopy (SEM) and X-ray adsorption fine structure spectroscopy (XAFS). In the case
of the No Weathering and Freeze/Thaw experiment, imaging samples and XAFS samples were not
collected. Three boards for each type of wood were tested for the Outdoor conditions, one board was tested
for the Freeze/Thaw conditions, and two boards were tested for the No Weathering conditions. The
Freeze/Thaw boards were cut into sections and stored in plastic containers so they could be more easily
moved in and out of the freezer. The Outdoor and No Weathering boards were left intact.
The samples were collected based on the wipe method developed and defined by the CPSC (Cobb, 2003;
Thomas et al., 2004). In brief, a piece of polyester fabric (cloth) (Texwipe TX 1099), approximately a 10
cm square, was placed into a 50 mL conical centrifuge tube followed by 2 mL of 0.9% NaCl solution and
4
-------
left capped overnight to soak. The following day, the cloth was secured to an aluminum disc, 8 cm in
diameter with a mass of 1.1 kg (weight). The effective surface area of the cloth after it was secured to the
weight was 50 cm2. The weight was attached to a sampling apparatus supplied by the CPSC. The apparatus
was secured to the board section at marks to ensure correct placement over the desired wipe area. The cloth
was dragged over a distance of 50 cm for an effective surface area of 450 cm2. A total of 10 wipe cycles
(dragging back and forth once equaled one cycle) were conducted where there were five wipe cycles before
rotating the disc 90° and performing another five wipe cycles. The cloth was removed from the apparatus
and placed back into the 50 mL tube for digestion or microscopy and XAFS analysis.
To simulate wear on the boards, the Outdoor boards were sanded after the day 399 sampling event. To
produce uniform, reproducible results, the CPSC wipe method was adapted using a sanding disk instead of
the cloth. An 80-grit sanding disk was trimmed to fit onto the weight on the sampling apparatus and secured
using double-sided tape. The board was then wiped twice in the same manner described above, for a total
of 20 wipes: five wipes, the weight was rotated 90°, another five wipes, the weight was rotated back to its
original position, five wipes, the weight was rotated 90°, and another five wipes. A new sanding disc was
used for each board. The sawdust produced from the sanding was blown effusing compressed air. Three
sampling events were then conducted to determine the amount of copper released immediately after sanding
and in subsequent wipes.
Cloth samples were extracted with nitric acid to solubilize copper retained on the cloth. The procedure
involved adding 15 mL of 10% nitric acid solution to each 50 mL conical centrifuge tube followed by
heating in a water bath at 60°C for between 22 and 24 hours. After cooling, the samples were vortexed and
6 mL of each sample was place in an ICP sample tube and diluted with 6 mL of Super Q water. The samples
were mixed for 20 minutes on a swirl shaker and then analyzed by ICP-OES. Cloths from the fourth set
were placed in 50 mL conical centrifuge tubes and 30 mL of Milli-Q water was added to the tube and placed
on an end over end shaker for 30 minutes. The extract was sequentially filtered through a 0.45 jam and 10
kDa filter. The resulting filter papers were analyzed via SEM and/or XAFS.
2.3 Analytical Methods
Analytical Techniques Samples were prepared for metals analysis (mainly total copper) using standard US
EPA methods for microwave assisted acid digestion. Method 305 la (Microwave Assisted Acid Digestion
of Sediments, Sludges, Soils, and Oils) was used for any solids samples using nitric and hydrochloric acids.
Method 3015a (Microwave Assisted Acid Digestion of Aqueous Samples and Extracts) was used for all of
the liquid samples using nitric acid (U.S. EPA, 1986 and 2007). Briefly for the solids, 0.1 g of sawdust, 0.5
g of the 2.5 jam filter membrane, or all of the 0.45 jam and 10 kDa membranes were weighed out and
quantitatively transferred to a Teflon® reaction vessel to which 9 mL of concentrated nitric acid and 3 ml
of concentrated hydrochloric acid were added. The vessels were allowed to react with the wood material or
filter membrane overnight prior to microwave digestion the following day. After digestion, the samples
were quantitatively transferred to 50 mL centrifuge tubes and diluted to 50 mL and stored at 4° C prior to
analysis by ICP-OES. For the solution samples 45 ml of solution and 5 mL of concentrated nitric acid were
added to a Teflon® reaction vessel followed by microwave digestion. If 45 mL of filtrate was not available,
the sample of interest was diluted to 45 mL followed by the addition of nitric acid. After digestion the
samples were quantitatively transferred to 50 mL centrifuge tubes and stored at 4° C prior to analysis by
ICP-OES.
ICP-OES Analysis. The total metal concentration of copper was measured by inductively coupled plasma-
atomic emission spectrometry (Thermo Scientific iCAP 6500) in accordance with EPA Method 601 OB after
digestion. The detection limit for Cu was 4 ug L"1. Accuracy and precision were assessed through triplicates,
5
-------
matrix spikes, and method blanks. Metal recoveries from matrix spikes were within 80-120% of the
expected values.
Solid Phase Characterization. The crystal structures of the micronized copper particles were determined
by X-ray diffraction (XRD) using a PANalytical Xpert pro MPD (Westborough, MA) with CuKa radiation.
Six scans were collected and averaged prior to analysis. Sawdust from the four as-purchased wood types
was compressed into a 3 cm disc prior to analysis to ensure the greatest detection of copper in the samples.
The oxidation state and local bonding environment of Cu were examined using X-ray absorption fine
structure (XAFS) spectroscopy. The copper K-edge spectra were collected at beam line 10-BM (Materials
Research Collaborative Access Team, Advanced Photon Source, Argonne National Laboratory, Argonne
IL). For the as-purchased materials, sawdust samples were compressed into 13 mm pellets using a hand
press and sealed between two strips of Kapton tape prior to analysis. For the leaching and rubbing
experiments, the XAFS spectra were collected directly from material retained on the filter membranes.
Adsorption spectra were collected at the K-edge energies of 8979 eV. Scans were collected from 8779-
9979 eV. Data collection was done in fluorescence mode using a 4-element solid-state Si-detector. The
synchrotron was operated at 7.0 GeV at a nominal 100 mA fill current. The energy of a Si (111) double
crystal monochromator was calibrated using an elemental Cu foil. All spectra were collected under ambient
conditions. A minimum of three scans (and up to 5) were collected for each sample.
Particle size and shape were determined by field emission scanning electron microscopy (FESEM) and
scanning transmission electron microscopy (STEM). FESEM analysis was conducted on a (FESEM) (JEOL
JSM-7600F, Tokyo, Japan) in Secondary Electron Imaging (SEI) and Backscatter Electron Imaging.
Elemental analysis was conducted by Energy Dispersive X-ray Analysis. Data reduction, qualitative
elemental analysis, and x-ray mapping was done with an Oxford 50 mm2 silicon drift detector running Isis
analytical software (Oxford Analytical, Oxfordshire, UK). Prior to analysis of the as-purchased material
samples, sawdust, generated using a random orbital sander with 80 grit sand paper, was compressed into
13 mm discs using a hand press and affixed to 12 mm carbon tabs mounted on 12 mm aluminum stubs. For
the leaching and rubbing samples, the filters were cut and affixed to 12 mm carbon tabs mounted on 12 mm
aluminum stubs. The SEM operated at 15 kV with a 2 nm resolution.
TEM micrographs were collected using a FEI Titan 80-300 probe aberration corrected scanning
transmission electron microscope (STEM) with a monochromator operating at 200 kV. Samples were cut
into matchstick-sized fragments using a razor blade. The fragments were dehydrated by placing them in an
increasing concentration of acetone solution. The dehydration procedure was repeated three times and the
wood samples were transferred to a rotator. The Epon-Araldite epoxy used to embed the wood samples
consisted of 20.9 g nadic methyl anhydride resin hardener (NMA) and 24.8 g Eponate 12 of which 8 g
batches were mixed with 0.3 ml batches of benzyl di-methy amine epoxy cure accelerator (BDMA). One
drop of resin was added to 1 ml of dehydrant every 5 minutes until the mixture was approximately 25%
resin. After 1 hour, 75% of the dehydrant was removed and resin was added until the solution was 75%
resin. The mixture remained in the rotator uncapped for 12 hours. The solutions were replaced with fresh
100% resin every 8 hours for the next 24 hours. Next, the samples were embedded in fresh resin and cured
for 18 hours at 60°C. A microtome (Leica UC7 Cryo Ultramicrotome) with a diamond knife at room
temperature and a maximum cutting speed of 1 mm/sec was used to slice 100-120 nm thick sections.
Sections were imaged at North Carolina State University (Raleigh, NC). .
-------
3 Characterization
The wood samples were thoroughly characterized to determine the chemical composition and particle size
distribution within the lumber. Total elemental composition of the lumber was determined from microwave
assisted acid digestion of sawdust samples generated using the method outlined in Section 2.1 (Table 3-1).
The presence of Si and/or Cr in the sample was used as an indicator of contamination of the sample from
materials dislodged from the sanding disc. The non-detect for Si and Cr indicated that there was a minimal
amount of material transferred from the sanding disc during sanding dust collection. Excluding copper, the
chemical compositions of the different wood types were similar (Table 3-1). Differences between samples
were noted (concentration of Ca and Mg in MCA-1 compared to MCA-2 and ACA), however they were
not substantial. The copper concentration in the ACA sample was twice that of the micronized samples,
which was expected based on the reported manufacturers values listed on the product. One of the advantages
of the micronized copper formulation, as described above, is the ability to exhibit similar biocidal
effectiveness with less copper present. There was a statistically significant difference (p-value =0.013) in
the total copper between the two micronized formulas. The total copper in the lumber was converted to a
volume concentration and compared with manufacturers reported concentration (Table 3-2). The measured
values for the total copper were 10 to 20% less than the manufacturer reported values.
As previously mentioned, the primary copper biocidal ingredient in micronized copper pressure-treated
lumber is copper carbonate (Cu2CO3(OH)2). The presence of copper carbonate was confirmed by X-ray
diffraction (XRD) patterns for MCA-1 and MCA-2 (Figure 3-1). XRD patterns were further analyzed to
determine if additional copper phases were present (copper(II) hydroxide, copper(II) oxide, copper(I) oxide,
Azurite (Q^COsHOFfh), and copper phosphate). Based on diagnostic peak locations, copper carbonate
was identified as the only crystalline copper component in the wood. In Figure 3-1, the dashed lines
numbered 1 and 2 mark the location of diagnostic diffraction peaks for copper carbonate and the dashed
line labeled 3 indicates the location of a diffraction peak associated with the cellulosic wood structure. All
four lumber types exhibit two broad diffraction peaks between 10° and 30° 20 and a smaller more well
defined peak near 34° 20 (Figure 3-1). Figure 3-1B and C provide an enlarged view of the area near the
location of the copper carbonate diffraction peaks. Interestingly, diffraction peaks for copper carbonate
were present in the ACA sample (Figure 3-1C) along with several other peaks. Extensive review of the
International Center for Diffraction Data database did not identify another specific copper compound with
similar diffraction peaks. However, calcium carbonate does have diagnostic diffraction peaks near the peak
locations in the ACA sample.
The intensity and width of XRD peaks can be used to indicate the relative abundance and size, respectively,
of the crystalline structures present. The peak intensity (width at half the maximum peak height) for MCA-
1 is significantly less than MCA-2. The 12% increase in the total copper content of MCA-2 compared with
MCA-1 does not fully account for the difference in peak intensity, suggesting that other non-crystalline
copper phases are present. Crystallites with nano-sized dimensions exhibit broader diffraction peaks
compared to larger particles as a result of the increased influence of the relaxation of surface atoms. The
broad diffraction peaks for copper carbonate for MCA-1, MCA-2, and ACA, indicate the crystallites present
are likely in the sub-micron range.
Table 3-1: Chemical composition of the four wood types as determined by microwave assisted acid
digestion and ICP-OES analysis. Averages are based on seven replicate samples. Additional
-------
elements analyzed included Al, Ag, As, Ce, Co, Cr, Fe, K, Mo, Ni, P, Pt, Sb, Se, Si, and V, however
there were no detectable quantities in the samples.
Elements
Cu
B
Ba
Ca
Mg
Mn
Pb
S
Sr
Ti
Zn
mg kg -1
MCA-1
Average
StDev
Rel St Dev
1327
26.6
2
ND
ND
ND
4.8
0.1
2.1
595.8
8.9
1.5
122.2
2.4
2
48.2
0.6
1.2
ND
ND
ND
79.7
2.3
2.9
2.2
0
1.6
2.5
0.5
18.1
8.1
0.2
o
J
MCA-2
Average
StDev
Rel St Dev
1574
158.6
10.1
ND
ND
ND
6
0.5
8.2
892.8
65.8
7.4
259.9
21.4
8.2
33
2.8
8.5
ND
ND
ND
113.8
8.1
7.1
3.8
0.3
8.5
1.3
0.3
21.3
14
5.4
38.4
ACA
Average
StDev
Rel St Dev
3833
133.6
3.5
65.8
1.6
2.4
2.4
0.1
2.3
908.4
10.8
1.2
253.2
3.9
1.5
20.4
0.6
2.7
2
0.3
13.1
63.7
7.3
11.4
2.1
0
0.7
1.3
0.3
20
11.2
0.3
2.8
UTW
Average
StDev
Rel St Dev
3.9
0.4
11.5
ND
ND
ND
18.3
0.3
1.4
907.8
14.7
1.6
76.2
0.6
0.8
16.8
0.2
1.4
ND
ND
ND
29.2
1.6
5.4
5.5
0.1
1.3
2.3
0.5
23.2
6.7
0.7
10.6
Table 3-2: Volumetric concentration of copper in the 4 selected wood products. Calculated using
the elemental concentrations from Table 1 and mass measurement of known volumes of lumber.
Wood
Type
MCA-1
MCA-2
ACA
UTW
Measured Copper
kg m 3 Ib ft 3
0.61 0.04
0.73 0.045
1.77 0.11
1.8*10-3 1.1*10-4
RSD
%
2
10
3.5
11.5
Reported Copper
kg m 3 Ib ft 3
0.8 0.05
0.8 0.05
N/A N/A
-------
Figure 3-1: X-Ray diffraction patterns for the as-purchased lumber. The dashed lines labeled 1
and 2 mark the location of diagnostic diffraction peaks for copper carbonate. The dash line labeled
3 indicates the location of a diffraction peak associated with the cellulosic wood structure.
u>
g
Untreated
MCA-2
MCA-1
ACA
10 20 30 40 SO 60 70
20
Untreated
MCA-2
MCA-1
The speciation of copper in the pressure-treated samples was further investigated by Cu K-edge XAFS
spectroscopy and linear combination fitting (LCF) of the normalized first derivative of the XANES spectra.
The XANES spectra of the three samples were compared with a variety of reference copper spectra to
identify copper species with similar spectral features (Figure 3-2). Reference spectra evaluated included:
Copper oxide (Q^O and CuO), copper carbonate (Q^COsCOFFh), copper phosphate (Cu5(PO4)2(OH)4),
and Cu complexed with acetate, oxalate, cysteine, histidine, or ferrihydrite. Copper complexed with acetate,
oxalate, cysteine and histidine were used as analogues for copper organic complexes that might be present.
Copper carbonate and copper complexed with acetate (AcO) and oxalate were identified as the three
compounds with the greatest similarities to the samples. Results from the LCF analysis indicate there is a
large variation in the relative abundance of copper species present. For the micronized copper samples, the
relative abundance of copper carbonate in MCA-2 is twice that of the MCA-1. The importance of this
finding becomes apparent when considering the total abundance and speciation of copper that is released
from the wood samples. If copper ions are more readily released in comparison to the crystalline material
then the interpretation of an increase in the copper released from MCA-1, compared to MCA-2, must take
into consideration the relative abundance of the copper species present initially. The average sensitivity of
XANES LCF analysis is approximately 5% (Kelly et a/., 2008; Bunker, 2010). Therefore, any species with
-------
a relative abundance of less than 5% is speculative. The presence of 10% of copper carbonate in the ACA
sample and the diffraction data provide ample evidence that copper carbonate is present in the sample.
Figure 3-2:First derivative XANES spectra for the three wood samples (MCA-1, MCA-2, and ACA)
and reference compounds. The highlighted blue area shows the location of the first derivative peaks
that correspond to copper carboxylate complexes.
Cu-Oxalate
CulAcO),
C^CO/OH),
8980 8990 9000 9010
Energy (eV)
Table 3-3: Results from the linear combination fitting of the first derivative XANES spectra. Values
are presented as average/standard deviation.
Copper Species
Wood Type
Copper „
^ , Cu-orgamc*
Carbonate
Percent/Standard Deviation
MCA-1
MCA-2
ACA
42/3 58/2
88/1 12/1
10/2 90/4
*The summation of Cu-Acetate and Cu-Oxalate
The flow paths used to transport fluids in trees are the same paths that facilitate the distribution of
preservatives within the wood. Previous research has identified that micronized copper is deposited along
the fiber cell walls responsible for longitudinal and radial distribution of fluids and on the pits (valves) that
connect void spaces together (Evans et al., 2008; Stirling et al., 2008; Matsunaga et al., 2009; Evans et al.,
2012). More detailed analysis of the distribution of copper carbonate has revealed that the copper carbonate
deposited in the wood structure does not readily penetrate the cell wall (Stirling et al., 2008; Matsunaga et
al., 2009; Matsunaga et al., 2010; Evans et al., 2012). Understanding the distribution of copper carbonate
10
-------
within the wood microstructure is important for considering potential release of nanoparticles from the
wood.
TEM and SEM electron images of the three pressure-treated wood types were collected. SEM images and
EDS spectra revealed a broad range of particles present in the wood samples in addition to copper carbonate.
Other phases present included iron oxides, calcite, and silicates (data not shown). The presence of non-
copper materials is not unusual. The SEM and TEM micrographs in Figure 3-4 provide examples of the
size and shape of the copper carbonate particles present. The SEM and TEM images of MCA-1 and 2,
respectively, clearly show the rectangular shape of the copper carbonate present in the wood. The
rectangular shape is consistent with the long prismatic or acicular shape of malachite (copper carbonate:
Cu2CO3(OH)2). Also evident in both images is the wide particle size distribution present. Based on the
particle shape, it is difficult to calculate an average particle size value of any significant meaning. Therefore,
the average particle length (longest dimension) and width (shorter dimension) of individual particles were
measured to determine an average length, width, and aspect ratio (length:width). The measured length and
width for copper carbonate in both micronized copper samples varied greatly. The standard deviation for
the average values was at least half the mean value, indicating a high degree of variability. There were also
notable differences between the copper carbonate present in both samples. The copper carbonate materials
present in MCA-1 were half the size of those in MCA-2 (Table 3-4). Histograms and total frequency and
box plots for MCA-1 and 2 are presented in Figure 3-5.
Figure 3-3: Micrographs of (A) BSE SEM MCA-1 and (B) Bright Field TEM MCA-2.
Table 3-4: Average particle size for MCA-1 and MCA-2. Values are presented as average/standard
deviation.
Wood Type
MCA-1
MCA-2
Length
nm
121/65
244/125
Average
Width
nm
56/42
105/58
Aspect Ratio
L/W
2.5/1.2
2.7/1.6
Percent of particles less
than 100 nm
Percentage (L, W)
59 94
14 68
Number of
Particles
Counted
314
370
The histograms for both micronized copper samples are skewed towards lower values, indicating a much
higher concentration of smaller particles. The total frequency plots for both samples show that greater than
50% of the measurements for length and width are less than the mean value calculated. SEM images for
11
-------
MCA-1 and MCA-2 (data not shown) indicate that the copper carbonate in the samples was present in
aggregated forms. There were few copper carbonate particles that were identified as isolated particles.
A question that has arisen is whether the materials used in pressure-treated lumber are truly nano in size.
The current general definition for a nanoparticle is a material with at least one dimension less than 100 nm.
As previously mentioned, the particle size distribution is very large. Based on the cumulative count and
histograms it is possible to determine the relative percentages of particles with either 1 or 2 dimensions less
than 100 nm. Table 3-4 provides a summary of the relative abundance of particles with a width of length
less than 100 nm. Based on the general definition of nanoparticles, 94% and 68% of the material have
widths less than 100 nm, for MCA-1 and 2, respectively. In the current study, length was always the longest
of the two dimensions, therefore 59 and 14% of the materials were less than 100 nm in 2 dimensions, for
MCA-1 and 2, respectively. Classifying a material as a nanoparticle becomes difficult with a broad size
distribution, especially when only a percentage of the material is less than 100 nm.
The variation in total elemental composition, crystalline abundance, copper speciation, and particle size
distribution are all factors that must be considered when interpreting the leaching and wipe data.
Interpreting the results based only on the concentration of copper released may easily lead to false
conclusions about the stability, fate, and/or the release mechanism of micronized copper in pressure-treated
lumber. The differences in the chemical and physical properties of the three pressure-treated wood samples
(especially the MCA samples) also offer an opportunity to see how copper speciation and copper carbonate
particle size impact the release of copper.
12
-------
Figure 3-4: Histogram, cumulative abundance, and box plots for particle length, width and aspect ratio (length/width) for Cu particulates
identified in the SEM and TEM images for MCA-1 and MCA-2, respectively.
MCA-1
80-
60-
1
040-
20-
Width
0
n
:
. 150
100
50
•
I rTLr-^ ,
50 100 150 200
Width (nm)
-
-
_
100 70~
60-
80
50-
60 § 40-
o
An ^O-
20-
20
10-
n n.
Length
—
•
•
• •
300
200
100
I
o
0 100 200 300
Length
(nm)
-
"
-
10C on
80
60-
60
§ 4°-
4H f\
20
n n.
Aspect Ratio
:
l
'
'
':
;
•
i
•
.
:
0 2
4
b
4
3
1
•
I
I
0
! I.. | : — i
6 8
•
-
100
80
60
A(\
20
n
10
Aspect Ratio
MCA-2
Cf\
oU-
60-
1
§ 40-
O
20-
Width
0
•
200
\
[
150
\
.
T
B
I
0
-
-
100
40-
80
30-
60 w
c
3
O 20-
40 0 '
20 10-
n n
Length
•
,
t
_
'
j
.
•
'
,
i
, • • • •
500
400
300
200
100
.
I
I ' I ! i
100 200 300 400 0 100 200 300 400 500 600 700
-
-
100
70-
80 60-
50-
60 w
§ 40-
40 0 30-
20-
20
10-
n n
Aspect Ratio
%
I
I
%
\
A '
800 0 2
1
4
• *
5
3
2
L-I
L_ : i
6 8
•
1
-
-
100
80
60
40
20
n
10
Width (nm)
Length (nm)
Aspect Ratio
13
-------
4 Leaching Experiment
4.1 Copper Release
The results for the leaching experiments were grouped into four figures, Figures 4-1 - 4-4, for comparing
the variation between treatment type (MCA v. ACA v. UTW), the wood form (block v. sawdust), the
duration (24 v. 72 hours), and the water chemistry (NaCl v. NaNOS v. SPLP). The figures display the total
amount of copper leached and the copper leached normalized to the surface area of the wood. In the case
of the wood blocks, each block was measured prior to the experiment to determine the surface area, while
1 m2 of surface area, the approximate area measured for each wood type, was used for each gram of sawdust.
For easier comparison, the average and standard deviation of the SPLP leachate results are summarized in
Table 4-1. This example was typical of all the results.
Figure 4-1: Leaching Experiment: 24-Hour Wood Block Leachate Fractionation.
1500
CN 1200
ro 900
_,- 600
° 300
0
80
ro
E
=
O
60
40
20
0
ITTl rm im
«-m rim
Unfiltered
2.5 urn Filtrate
0.45 |im Filtrate
1 0 kDa Filtrate
f|Tl JTjTI rpl | aqpn rpn
MCA-1
MCA-2
ACA
UTW
14
-------
Figure 4-2:
1500 -
CNI 1200 -
en 900 -
_,- 600 -
° 300 -
n
80 -
en 60 -
E
=f 40 -
O
20 -
Figure 4-3:
2.0 -
™ 1.5 -
E
^ 1.0-
^~
0 0.5 -
0.0 -
20 -
|> 15 -
5 1°-
5 -
0 -
Leaching Experiment: 72-Hour Wood Block Leachate Fractionation.
Ill ^^m I InfiltPi-pH
mfffim rtenj-c
I
irn rrn rn rm rm irn
[I
1
L
I
1 1 2.5 urn Filtrate
1 • 0.45 urn Filtrate
1 1 10kDa Filtrate
FT IIT
Jn TT
)
MCA-1 MCA-2 ACA UTW
Leaching Experiment: 24-Hour Sawdust Leachate Fractionation.
Li I
Irn Irnlm HT ITHrrn 1
if
- fc , in, Li II
iTnlrnirn IJ Irnlrn 1
^° v/^ <$* ^ ^ <$* ^
V V
/
rj
BT
1
^
^v
^^B Unfiltered
1 1 2.5 urn Filtrate
1 1 0.45 urn Filtrate
« 1 1 10kDa Filtrate
j-.
/~\ ^) O
. (X __X*J ^/^
N~ sfo^ •?
MCA-1
MCA-2
ACA
UTW
15
-------
Figure 4-4: Leaching Experiment: 72-Hour Sawdust Leachate Fractionation
2.0
™ 1.5 H
1.0-
0 0.5 -
zi
o
0.0
20 -
15 -
10 -
5 -
0
It
Unfiltered
2.5 |im Filtrate
0.45 |im Filtrate
10kDa Filtrate
MCA-1
MCA-2
ACA
UTW
The figures show that the ACA released significantly more copper than the MCAs over all the conditions
tested while MCA-1 released slightly more copper than MCA-2 and the UTW released no copper. These
results are expected, since the ionized copper formulation (ACA) has a higher copper concentration than
the micronized formulations and is known to leach copper significantly. The increase in the quantity of
copper released from MCA-1 compared with MCA-2 is likely related to the copper speciation initially
present in the wood. Based on the LCF of the XANES data, 58% of the copper present in MCA-1 is
organically complexed compared to only 12% in MCA-2. The increased quantity of organically complexed
copper would likely result in increased leaching of copper from the wood. The observed differences are
consistent with results published in the literature as well. Using lab treated specimens, Kartal et al. (2009)
tested specimens that had been treated with both copper sulfate (ionized) as well as a nanocopper
formulation. They found a negligible amount of copper was released from the nanocopper treated wood
while seeing over 20% released from the copper sulfate treated wood.
Table 4-1: Summary of Leaching Results with SPLP Water Chemistry.
Copper Measured in the Unfiltered Fraction After Leaching (nig (nig ni"2))
PTL Type
MCA-1
MCA-2
ACA
Wood
24 Hour
7.9 ±0.8 (162 ±16)
5. 5 ±1.5 (99 ±30)
30.3 ±9.0 (571 ±237)
Block
72 Hour
11.1 ±0.5 (251 ±8)
7.6 ±3.0 (151 ±55)
65.6 ±3.0 (1246 ±39)
Sawdust
24 Hour
3. 5 ±0.7 (0.35 ±0.07)
4.3 ±0.3 (0.43 ±0.02)
12.5 ±2.1 (1.23 ±0.20)
72 Hour
6.5 ±2.3 (0.64 ±0.22)
8.5 +3. 1(0.83 ±0.31)
9.6 +0.5 (0.96 ±0.05)
The wood blocks leached more copper into solution than the sawdust, a result that is somewhat
counterintuitive. A possible explanation involves the re-adsorption of copper by the wood particulates. The
surface area of the sawdust was roughly 1 m2 g"1; equaling 1 m2 in each sample (1 g of sawdust per sample)
the average surface area for the 10 g wood blocks was 4.28*10~3 m2, three orders of magnitude less. The
16
-------
high affinity of copper for organic material combined with the large surface area of the sawdust results
increase adsorption of copper and a subsequent decrease in the solution concentration of copper.
Solution chemistries had little effect on the quantity of copper released. There was no noticeable difference
between the three solution chemistries tested for the MCAs. While there was a difference in the ACA results
for the 24-hour duration, the differences were not observed in the 72-hour test and the variability of the
ACA results in general make it difficult to draw any definitive conclusions. The duration of the experiment
affected the amount of copper released from the wood blocks, but not the sawdust. For example, the ACA
wood block increased from 500-800 mg m~2 at 24 hours up to 1000-1300 mg m~2 at 72 hours, while the
sawdust stayed constant between 1 and 1.5 mg m~2. The increase indicates that 24 hours is not long enough
for the leaching to reach steady-state for the wood blocks, but it is sufficient for the sawdust. The leaching
from the wood blocks would likely be controlled by the migration of the copper to the surface, where it
then would be released into the liquid phase. The significantly larger surface area and smaller particle size
of the sawdust would remove any need for migration through the wood, allowing for the direct transition
of the copper into the liquid phase.
Each of the leachates was filtered to fractionate the solution, as described in Section 2.1. By separating the
different size ranges (above 2.5 urn, between 2.5 um and 0.45 urn, between 0.45 urn and 10 kDa, and below
10 kDa), some conclusions can be drawn about the type of copper released and whether it is free in solution
or attached to cellulose. There was very little difference in the solution copper concentrations between the
4 fractionated solutions (No filtration, 2.5 jam, 0.45 jam, and 10 kDa), indicating the majority of the leached
copper passed through all three filters. In order to pass the 10 kDa filter, the copper would have to be in an
ionic form because any micronized or nanoparticle copper above ~3 nm would have been retained. The
lack of particulate copper likely means that when the copper is released or very soon after it is released, it
dissolves and does not persist as a particle. This conclusion is important since the environmental fate of
copper micro/nanoparticles is currently not well understood and could be potentially harmful. The fate,
exposure routes, and toxicity of ionic copper is better understood (Forest Products Laboratory, 2000). The
environmental impact is discussed in more detail in Section 4.4 of this report.
4.2 Mass Balance
The release of copper through leaching discussed above pertained to the amount of copper measured in the
liquid filtrates. As can been seen in the figures, the deviation in the results make it difficult to see small
differences between the different fractions. Micro/nano-copper may be present, but in amounts too small to
be distinguished from the solutions. During the filtration process, the filter membranes were collected and
analyzed, allowing for the amount of copper in the particulates to be determined and a mass balance to be
performed. The mass balance was only performed on the wood block conditions. It was not practical for
the sawdust conditions because of the large amount of sawdust collected during filtration on the 2.5 nm
filter papers.
Table 4-2 shows the results of the mass balance. The table displays the total amount of copper released and
the ionic fraction, both taken from the solution concentrations; the particulate copper released, determined
from the amount of copper retained on all three filter papers; and the filtration recovery, determined from
the ionic fraction and the particulate copper compared to the total released. The copper in the particulates
could be in a number of forms: ionic copper attached to cellulose fibers, micro/nanoparticle copper attached
to cellulose fibers, or free micro/nanoparticle particles. For the MCA samples, roughly 2-6% of the copper
released is attached in some way to a particle and was retained during filtration.
17
-------
Table 4-2: Mass Balance of Copper Released from Wood Blocks during Leaching.
r» *• /i, \ Solution
Duration (hr) „, . ,
Chemistry
Total
Released
(mg)
Ionic
Released
mg
%
Particulate
Released
mg
%
Filtration
Recovery (%)
MCA-1
24
72
NaCl
NaNO3
SPLP
NaCl
NaNOS
SPLP
0.90
0.67
0.79
1.08
1.08
1.11
0.86
0.74
0.71
1.06
0.99
1.06
95
110
89
98
92
96
0.04
0.02
0.05
0.02
0.04
0.07
4.8
3.6
6.3
1.6
4.1
6.3
100
113
96
100
96
102
MCA-2
24
72
NaCl
NaNOS
SPLP
NaCl
NaNOS
SPLP
0.49
0.57
0.55
1.00
0.71
0.76
0.47
0.51
0.52
0.95
0.63
0.67
95
89
95
94
89
89
0.02
0.02
0.03
0.03
0.03
0.05
4.9
3.6
5.2
2.9
4.5
6.0
100
93
100
97
93
95
ACA
24
72
NaCl
NaNOS
SPLP
NaCl
NaNOS
SPLP
4.81
2.76
3.03
5.92
5.22
6.56
4.66
2.66
2.90
6.07
4.34
6.95
97
96
96
103
83
106
0.06
0.03
0.07
0.08
0.09
0.12
1.2
1.2
2.4
1.3
1.7
1.8
98
98
98
104
85
108
Table 4-3 provides a breakdown of the released copper contained within each size range. The majority of
the copper was found in the material retained on the 2.5 jam filter, between -1-5% of the total copper
released from the MCA and ACA samples. Between 0.5% and 1% was retained on the 0.45 jam filter while
0.1%-0.3% was retained on the 10 kDa filter. The majority of the copper carbonate present in the MCA-1
and 2 samples was significantly smaller than 2.5 jam and, since most of the copper was retained on the 2.5
jam filter, it is likely associated with cellulose particulates and not as free particles. The quantities of copper
found in the smaller two fractions do indicate that a small amount of micro/nanocopper could be released.
The results of the speciation analysis, Section 4.3, expand upon these results and identify the form of copper
present.
18
-------
Table 4-3: Particulate Copper Released and Copper Associated with Each Filtration
Fraction.
Duration (hr) Solution
Chemistry
Particulate >2.5 |4,m
Released
mg
%
%
>0.45 fun
%
>10 kDa
%
MCA-1
24
72
NaCl
NaNO3
SPLP
NaCl
NaNOS
SPLP
0.04
0.02
0.05
0.02
0.04
0.07
4.8
3.6
6.3
1.6
4.1
6.3
3.9
2.9
5.2
0.9
3.4
5.4
0.8
0.5
1.0
0.6
0.6
0.7
0.1
0.2
0.1
0.1
0.2
0.1
MCA-2
24
72
NaCl
NaNOS
SPLP
NaCl
NaNOS
SPLP
0.02
0.02
0.03
0.03
0.03
0.05
4.9
3.6
5.2
2.9
4.5
6.0
4.1
2.7
4.1
2.1
3.7
5.1
0.7
0.8
1.0
0.7
0.6
0.8
0.1
0.1
0.1
0.1
0.2
0.1
ACA
24
72
NaCl
NaNOS
SPLP
NaCl
NaNOS
SPLP
0.06
0.03
0.07
0.08
0.09
0.12
1.2
1.2
2.4
1.3
1.7
1.8
0.8
0.8
1.9
0.9
1.3
1.4
0.3
0.3
0.4
0.3
0.3
0.4
0.1
0.1
0.1
0.1
0.1
0.1
4.3 Copper Speciation
4.3.1 X-ray Absorption Fine Structure Spectroscopy
The mass balance data indicates that only a small fraction of the total copper released was associated with
the solid phase. The speciation of the copper retained by the filters was determined using Cu K-edge XAFS
spectroscopy. XAFS data was collected from the materials that were retained on the three different filters
(2.5 jam, 0.45 jam, and 10 kDa). For the current discussion, >2500 nm refers to the solid phases that were
retained on the 2.5 jam filter after filtration. Subsequently, >450 nm refers to the solids that passed through
the 2.5 jam filter but were retained on the 0.45 jam filter, and finally >~3 nm refers to materials that that
passed through the 0.45 jam filter but were retained by the 10 kDa filter, which is approximately a 3 nm
19
-------
membrane. The previous Copper Release and Mass Balance sections, 4.1 and 4.2, respectively, indicated
that there was little difference in the total quantity of Cu released as a function of the leaching solution
chemistry. Due to the similarity in the total ion and particulate Cu released from the lumber, the speciation
analysis focuses on changes in Cu speciation for wood blocks that were leached with the SPLP solution for
72 h. The SPLP solution was chosen because of its use as an analogue for precipitation, and almost all
pressure-treated wood surfaces are exposed to precipitation. The wood blocks were chosen over the sawdust
samples for speciation analysis for two reasons. First, solid wood surfaces are the only pressure-treated
surface found in the built environment aside from the sawdust generated during construction. Second, in
order to fully understand the potential materials that could be leached from a pressure-treated wood surface
it was important to capture the fraction that would be retained by the 2.5 jam filter. With the sawdust, as
discussed previously, it is impossible to differentiate between materials released from the sawdust and the
sawdust itself with respect to materials retained by the 2.5 jam filter.
The MCA-2 first derivative XANES spectra for the original material and the solid phases retained by the
filters is presented in Figure 4-5 along with four reference compounds as an example for all three wood
types. The reference compounds chosen were based on known copper phases that would be present (copper
carbonate and copper complexed with organic materials) and the similarity in spectral shape with the
samples (Ci^O). Other copper organic complexes were investigated as potential analogues for copper
complexed with organics including sulfur and amine enriched compounds. However, the Cu carboxylate
complexes exhibited the most similarity to the solid phases retained on the filters. In additional to additional
organic phases, other potential inorganic copper phases were compared to the sample spectra, but there
were either no or limited similarities.
Visual inspection of the XANES spectra reveals substantial changes in the Cu speciation for each size
fraction compared with the original material (Figure 4-5). The presence of a Cu1+ species is immediately
evident in the > 2500 and > 450 nm fractions as evidenced by the derivative peak present and its coincidence
with the first derivative peak of Cu2O (Line A). The presence and relative abundance of copper carbonate
in the samples is evidenced by the presence and shape of the derivative peak shoulder near 8986 eV (Line
B). A well-defined derivative peak shoulder present near Line B indicates minimal presence of Cu
complexed with organic carboxylate functional groups. The shoulder would be absent or broad if Cu
complexed organic carboxylate groups were present due to the presence of strong derivative peaks present
in the Cu-Oxalate and Cu-Acetate complexes. The presence of organically complexed copper is evident by
derivative peak near 8983 or 8987 eV (derivative peak locations for Cu-Acetate and Cu-Oxalate complexes)
and the presence of a well-defined peak or shoulder in the derivative spectra near the area highlighted by
Line C. All four of the MCA-2 sample spectra presented exhibit features associated with the reference
compounds. The same trends (presence of Cu1+ (Line A), well defined copper carbonate shoulder (Line B),
and strong shoulder/peak near 8995 eV (Line C)) were present in the other two wood samples—MCA-1
and AC A (data not shown).
20
-------
Figure 4-5: First derivative Cu Ka XANES spectra for the materials retained on the filter papers
after sequential filtration of the supernatant from a sample of MCA-2 lumber leached with SPLP
solution for 72 h. A) Indicates the position of the Cu1+ first derivative peak, B) indicates the low energy
shoulder of copper carbonate (Cu2CO3(OH)2) first derivative peak, and C) indicates a first derivative
shoulder and peak position for copper complexed with organic acids.
TJ
(D
N
"CD
Q)
"TO
'c
0
Q
0)
c/)
Jt:
o
Cu-Oxalate
Cu(AcO)2
Cu2C03(OH)2
Cu20
>10kDa
> 0.45 urn
> 2.5
MCA-2
8980 8990 9000 9010
Energy (eV)
In order to determine the relative abundance of each phase present in the samples, LCF analysis of the first
derivative of the XANES spectra was conducted for each wood type and sample fraction (Table 4-2 and
Figure 4-6). In addition to the copper speciation, the surface normalized concentration of copper particulates
is presented in Table 4-2. Copper(I) oxide was identified in all three of the pressure-treated samples for the
>2500 and >450 nm fractions. The mechanism responsible for the copper reduction is unclear, but previous
research has demonstrated that organic matter can reduce copper (II) to (I) under oxidizing conditions (Leal
and Van Den Berg, 1998; Fulda et al, 2013). In previous research, the mechanism occurred through the
reduction of the Cu(II) ion, suggesting that the formation of the Cu(I) is controlled by the ionic Cu present
and not malachite. The total mass of copper retained by the three membranes for the different wood types
did not exceed 30 mg m~2, of which Cu(I) made up no more than 20% of the total copper present.
There were notable changes in the speciation of copper retained by the membranes and the as-purchased
materials. For the ACA wood there was a decrease in the total organically complexed copper and an increase
in the aqueous copper from the >2500 to >~3 nm. The aqueous copper species is most likely related to
copper that is loosely bound to the organic fraction. Changes in the speciation of copper retained on the
21
-------
membranes for the two micronized treated samples differed. For MCA-1 there was an increase in the
abundance of copper carbonate with smaller size fractions, while the reverse was true for MCA-2. The LCF
results for MCA-1 would seem to indicate that copper carbonate may exist in solution as free particles since
there is an increase in the abundance of material with smaller size fractions. However, a similar argument
may be made for the opposite based on the MCA-2 results. If there was a strong association of copper
carbonate with wood fibers then there would a decrease in the relative abundance copper carbonate in the
smaller size fraction.
Table 4-4: Copper speciation (percent abundance) and mass of cooper leached from wood blocks
leached with the SPLP solution for 72 h as a function of filter particle cutoff. Data presented is the
relative abundance of each copper phase that was retained on the filter paper after sequential
filtration of the supernatant. Unleached refers to the as-purchased materials.
Copper Species
Wood Type
Unleached > 2.5 n,m > 0.45 n,m > 10 kDa
MCA-1
Cu Released (mg (mg rri2))
Cu2O
Cu2CO3(OH)2 42%
CuQ2 (aq)
Cu-Organic* 58%
5.4(13.6) 0.8(1.9)
9% 23%
45% 34%
46%
45%
0.1 (0.3)
71%
29%
MCA-2
Cu Released (mg (mg rri2))
Cu2O
Cu2CO3(OH)2 88%
CuCl2 (aq)
Cu-Organic* 12%
5.0(7.6)
13%
71%
16%
0.8 (1.3)
9%
62%
29%
0.1 (0.2)
37%
63%
ACA
Cu Released (mg (mg rri2))
Cu2O
Cu2CO3(OH)2 10%
CuCl2 (aq)
Cu-Organic* 90%
1.4(17.3)
8%
29%
63%
0.4(4.6)
14%
24%
62%
0.1 (0.9)
22%
78%
*The summation of Cu-Acetate and Cu-Oxalate
Figure 4-6: Copper speciation and mass of cooper leached from wood blocks leached with the SPLP
solution for 72 h as a function of filter particle cutoff. Data presented is the relative abundance of
each copper phase that was retained on the filter paper after sequential filtration of the supernatant
22
-------
and the mass per unit area of copper that was leached from the lumber. The first column in each
graph refers to the original copper species distribution in the as-purchased materials.
ICuOgonc •CuO. „ •Cu,COl(OH)> ••CuO
4.3.2 Microscopy
SEM images for the materials that were retained on the 2.5 jam, 0.45 jam and 10 kDa filter membranes for
wood blocks leached for 72 h in the SPLP solution are presented in Figures 4-7, 8, and 9. In the BSE
SEM images, the brighter objects are the copper particles present in the samples. The 2.5 and 0.45 jam
membranes were Whatman® Nuclepore® Polycarbonate Track Etched membranes. These membranes
have a smooth surface and are designed for environmental sample collection and subsequent microscopy
analysis. The larger particle present in the images do not resemble the SEM micrographs provided by the
manufacturer, indicating the particles must be bound/associated with cellulose that was dislodged from
the wood surface during leaching. Based on the small number of particles identified, it was not possible to
determine an average particle size for the copper carbonate. The difference in the scale bars between
Figures 4-7 and 8 indicate that smaller copper carbonate particles were present on the 0.45 jam membrane
compared with the 2.5 jam. While all of the particles imaged are associated with cellulose, the enrichment
of smaller particles retained on the 0.45 jam filter would suggest that copper carbonate may exist as free
particle/aggregate in solution. Analysis of the 10 kDa filters resulted in finding a single aggregate of
copper particles for the MCA-2 sample (Figure 4-9). The very low concentration of copper present made
it extremely difficult to isolate particles. Unlike the images for the 2.5 and 0.45 mm fraction, the particles
are not well defined making it difficult to visually distinguish individual particles. The overall aggregate
is several microns in size indicating that the aggregation of the materials likely occurred after filtration.
23
-------
Figure 4-7: BSE SEM Micrographs of particulates retained on a 2.5 (J,m membrane (A) MCA-1 and
(B) MCA-2. The white square outline indicates the location of where EDX spectra were collected to
determine elemental composition of the particles.
Figure 4-8 BSE SEM Micrographs and EDX chemical maps of particulates retained on a 0.45 mm
membrane (A) MCA-1, (B) EDX copper map, (C) MCA-2, (D) EDX copper map.
24
-------
Figure 4-9 BSE SEM Micrographs of particulates retained on a 10 kDa membrane from the MCA-
2 sample.
4.4 Environmental Impact
The leaching results show that there is no significant release of copper micro/nanoparticles into the
environment, but that copper is released by contact with water. However, the results cannot be used
directly for predicting the MCA's environmental impact. The size of the specimens, mainly the wood
blocks, leads to accelerated leaching and does not adequately reflect an in-situ release of copper (Forest
Products Laboratory, 2000).
MCA is relatively new and studies of copper release into the environment near structures built with it are
not available. Studies on ACA have been performed, allowing for the ACA results in this study to provide
a point of comparison. Since the MCA wood has significantly less copper and released much less copper
during testing than ACA, the release of copper into the environment can be assumed to show a similar
trend.
A study by the Forest Service of a boardwalk created with ACA assessed the impact of the release of
copper on the flora in the localized area around the boardwalk. The ACA used was rated for ground
contact and had a very high concentration of copper (7.0-8.2 kg m~3 (0.44-0.51 Ib ft"3)), significantly
higher than the ACA (1.77 kg nr3 (0.11 Ib fir3)) and MCA (0.8 kg nr3 (0.05 Ib ft'3)) used in the current
study. Despite the high concentration, the study found no significant impact on the local environment.
High concentrations of copper (373 ppm compared to a baseline of 30 ppm) were found immediately next
to the wood and it was shown to be migrating into the soil after 6-12 months, but no drop in the soil
microbial flora was seen (Forest Products Laboratory, 2000). This suggests that the MCA, when factoring
in the leaching results and comparing to the ACA, would release less copper.
The above discussion focuses on the total ionic copper released and does not consider the impact of the
micronized copper on the environment. Currently, data is not available in the literature on the impact of
micronized copper on the environment. The results indicate the potential for a small fraction of
particulate copper to be released, though the significance of the amount is unknown. Given the current
25
-------
data set, it is not possible to determine that exact form of copper carbonate leached from the wood: bound
to cellulose, isolated particles/aggregates in solution, or a combination of the two. The importance of the
question is related to the potential transport of the materials and exposure dose. If the copper carbonate is
bound to the cellulose, than any potential transport will be limited by the size of the cellulose fragment.
However, if copper carbonate exists as a free particle/aggregate in solution then transport and exposure
will be governed by the size of the particle.
26
-------
5 Wipe Experiment
5.1 Copper Release
The wipe experiment was conducted in three parts. First, a set of boards that did not undergo any surface
alteration or exposure were samples to establish a baseline for comparisons with the two treatments to
determine the net effect of weathering on the release of copper. Second, a set of boards was left outdoors
to undergo natural weathering and finally, a third set of boards was artificially weathered through freeze
and thaw cycles. Results from the ICP analysis were determined as a concentration, which was then
converted to total mass of copper released, and then normalized to the copper released per square meter.
The results from the Outdoor and Freeze/Thaw tests are shown in Figures 5-1 and 5-2.
Figure 5-1: Copper Released by Wiping - Outdoor Weathering.
40
30 -
20 -
E
D) 6 -
13
o
4 -
2 -
MCA-1
MCA-2
ACA A
100
200
300
400
500
Time, d
The Outdoor experiment results indicate a large initial release of copper that stabilizes over time. The initial
quantity of copper released was variable within each type of wood, particularly for MCA-1, and is likely
due to manufacturing techniques and pressure-treating formulations as well as natural variations in the
wood material. There was very little difference between the release of copper from the MCA-2 and ACA.
MCA-1 was significantly higher than the others initially, but stabilized at a similar concentration after the
Day 14 sampling event. After the day 34 sampling event (the 3rd sampling), the average copper released for
each wipe event (Table 5-1) was ~1.5 mg m~2.
The Freeze/Thaw experiment results show a similar trend. As stated above, the boards were not sampled at
0 cycles. The amount of copper released decreased after the first two sampling events and stabilized below
1 mg m"2 (Table 5-1). Again, the MCA-1 wood had a much higher initial concentration, which eventually
fell to a concentration similar to the other two wood types. The MCA-2 and ACA woods showed very little
difference in copper release.
The cycles of freezing and thawing were anticipated to have a destructive effect on the wood due to
expansion and contraction. However, no visible effect could be seen on the wood and the copper release
results suggest very little deterioration occurred. By comparison, the Outdoor boards were expected to have
very little copper on the surface, most having been washed away by the weathering. Visually, the boards
were heavily cracked and faded and the initial green hue associated with the material was absent. Despite
the visual and weathering differences, the release of copper was very similar between the two. To
investigate, the third test was conducted on fresh boards with no weathering. The goal was to determine if
27
-------
the copper transfer was related to the number of times the boards were wiped or weathering conditions. The
results, Figure 5-3, are consistent with the Outdoor and Freeze/Thaw experiments, decreasing copper
concentration with successive sampling events and the concentration of copper released from the wood
stabilized after 3-4 sampling events for all three tests.
Figure 5-2: Copper Released by Wiping - Freeze/Thaw Weathering.
12
Cycles
While the trend is the same, the amount of copper released is slightly different between the wood types.
Figure 5-3 and Table 5-1 shows that the MCA-1 wood released approximately the same amount in all three
tests. The MCA-2 and ACA released more copper in the Outdoor experiment than the other two, possibly
indicating the additional weathering destabilizes copper in the wood products.
Thomas et al. (2004) found a similar result when performing wipe tests for arsenic on CCA wood. They
examined the amount of arsenic transferred with each wipe cycle (one back and forth motion) and found it
to decrease with each additional cycle, approaching a plateau after 3-4 cycles. They concluded that a hand
would become saturated, preventing further transfer, or the amount of available arsenic was depleted from
the surface. They concluded that 10 cycles was optimum to reach equilibrium between the surface and the
wipe, which is the basis for the method used in this study. Our results support the depletion of the available
material on a larger scale, over multiple, successive wipe events.
28
-------
Figure 5-3: Copper Released by Wiping - Weathering Comparison.
O)
E
d
o
O)
E
d
30
20
10'
6 -
4 -
2 -
0
12 -
10 -
E 8 H
O)
6 -
4 -
2 -
0
12 -
10 -
8 -
6 -
MCA-1
No Weathering
Outdoor
#
Freeze/Thaw
MCA-2
ACA
123456789
Rub Events
Table 5-1: Average Release of Copper During Wiping Experiments.
10 11
12 13
14
Weathering
Experiment
Outdoor
Freeze/Thaw
No Weathering
Copper Released by PTL Type (mg m~2)
MCA-1
1.4 ±0.8
0.8 ±0.6
1.6 ±0.5
MCA-2
1.4 ±0.6
0.3 ±0.1
0.8 ±0.1
ACA
1.4 ±0.7
0.2 ±0.1
0.3 ±0.1
Period*
Days 34 - 399
Cycles 12 - 24
Samplings 3-12
* Excludes the first two sampling events from each experiment (Day 0 and 14, 4 and 8 cycles, and
samples 1 and 2 for the Outdoor, Freeze/Thaw, and No Weathering experiments, respectively).
29
-------
The difference in the total quantity of copper released between the MCA-1, and MCA-2 and ACA is not
immediately apparent. The speciation data of the as-purchased materials does not provide any specific
insight into the differences in copper released, as the species of copper in the ACA and MCA-2 materials
share similarities with MCA-1. It is difficult to attribute the increased initial released of copper to the
presence of copper carbonate nanoparticles or copper organic complexes since MCA-2 is enriched in
malachite and ACA is enriched in copper organic complexes. Copper form and formulation could play a
role in the amount of copper released. Additional differences in the formulations, such as co-biocides or
stabilization chemicals, could also affect how the copper migrates through the wood. Finally, the initial
release may be a result of excess or residual pressure treatment solution remaining on the wood surface.
Often residual materials are visible on the surface of the pressure-treated wood products in the market place.
After the day 399 sampling event for the Outdoor experiment, the boards were sanded to simulate surface
wear and to determine if there would be a spike in the copper released. Figure 5-4 shows the day 399
sampling event followed by three other sampling events on day 400, 406, and 407. The sanding occurred
on day 400, just before the samples were collected. Sanding resulted in almost no change in concentration,
despite a visible change in the board where material had been removed. With the second and third sampling
events there is a visual decrease in the total quantity of copper released in Figure 5-4. Two possible
conclusions may be drawn from this exercise: 1) the copper is uniform enough in the depth of the board
that sanding/surface wear does not impact copper release or 2) the sanding method was not sufficient to
remove the top layer of weathered wood material. Further work is required to determine and clarify this
result, most likely involving complete sanding of the board surface to restore it to an original, pristine
appearance.
Figure 5-4: Copper Released by Wiping - Outdoor Weathering - After Sanding.
MCA-1
MCA-2
ACA
Day of Sanding
399
400
402
403 404
Time, d
405
406
407
408
5.2 Copper Speciation
5.2.1 Microscopy
No particulates or materials associated with the lumber were identified on the 10 kDa filter papers.
Particulates were identified during analysis, however the particle size was >5 jam indicating the materials
were a result of contamination of the filter papers prior to analysis. SEM images of the 0.45 jam filter
membranes from the one month wipe sample showed a wide variety of materials present for all three
pressure-treated wood types (MCA-1, MCA-2, and ACA) (Figure 5-5).
30
-------
Figure 5-5: SEM micrographs of materials collected on 0.45 (J,m filter papers. MCA-1 (A), MCA-2
(B), and ACA (C).
Particles enriched in silica aluminum, iron, and calcium were identified from EDS spectra. In Figure 5-5 A
and B the rhombohedral shapes were identified as copper carbonate, the active chemical compound in
micronized copper pressure-treated lumber, based on crystallite shape and the EDS spectra (Figures 5-6
and 5-7). In the two micronized copper samples, copper carbonate was the predominant particle identified.
Copper carbonate crystallites were not identified in the ACA sample. The cubic and non-spherical/irregular
particles present in the ACA sample (Figure 5C) were identified as calcite and silica, respectively, based
on crystallite shape and EDS spectra. For both MCA-1 and MCA-2, the copper carbonate present was
aggregated in isolated locations (Figures 5-5A and B, 5-6, and 5-7). The crystallites were always associated
with larger particles. This is especially evident for the MCA-1 micrograph (Figure 5-5A). As previously
discussed, the Whatman® Nuclepore® Polycarbonate Track Etched 0.45 jam membranes have a smooth
surface and are designed for sample collection and subsequent microscopy analysis. The larger particles
present in the micrographs do not resemble the SEM micrographs provided by the manufacturer, indicating
the particles are bound/associated with cellulose dislodged from the wood surface during sample collection.
The micrographs in Figure 5-5 of MCA-1 and 2 are very similar in appearance to those published by
Matsunga and Kiguchi (Matsunaga et al., 2009) who evaluated the distribution of copper carbonate in
pressure-treated lumber. The similarity in the images suggests that the malachite dislodged from the wood
surface remains associated with the dislodged cellulose and not as a separate entity or free particle. As
previously discussed, the difference in particle size between MCA-1 and 2 is immediately evident when
comparing the images in Figures 5-6 and 5-7. The average particle length, width, and aspect ratio were
determined for copper carbonate present in both micronized copper treated samples (Figures 5-6 and 5-7
and Table 5-2). The histograms in Figures 5-6 and 5-7 highlight the skewed size distribution of the copper
carbonate, similar to the plots for the as-purchased materials. In comparison with the as-purchased material,
the particle size and aspect ratio are very similar. The differences in the particle sizes are likely due to
different imaging methods, BSE SEM versus TEM. The similarity in the shape and size indicates that after
one month of exposure outdoors, very little transformation of the copper carbonate occurred.
31
-------
Figure 5-6: BSE SEM micrograph of MCA-1 (A), EDS map of Cu from the same sample (B), and histogram and cumulative abundance
plots for particle length, width, and aspect ratio (length/width) (C).
160-
120-
80-
40-
n.
Length . .
^
'
•
400
„ 300
I'00
100
Fl_
0
•
1
., L_L J_LU • ,
1 -
-
.
•
100
80
60
40
20
n
200 400 600
Length (nm)
800
c
o
200-
160-
80-
40-
n-
W
I
idl
.
h
•
1 125-
100
75
50
L
0
1
.
1
-
.
-
-
mo
80
60
40
20
n
50 100 150 200 250 300 350 400
Width (nm)
200-
150-
—
i 100-
o
0
en .
3V
0-
Aspect Ratio
. • •
• " 8
(
•
i
'
' 6
5
4
a
\-
.
2
'•'7
\Jr
.
.
•
1 -
1
100
80
60
40
20
n
0 1
3456
Aspect Ratio
7 8 9 10
32
-------
Figure 5-7 BSE SEM micrograph of MCA-2 (A), EDS map of Cu from the same sample (B), and histogram and cumulative abundance
plots for particle length, width, and aspect ratio (length/width) (C).
lum
60-
50-
40-
§ 30-
20-
Length
|j;
.
450
300
150
0
-
-
_
-
I :
.,,,.!,, :
0 200 400 600 800 1000 1200
Length (nm)
ru
100
80
60
20
0
60-
50-
g30H
40 O
20-I
10-
0
Width
-1
i
1
• * '"
' 250
200
150
100
k „
0
.
.
100
80
60 j2
40 J4°
100
80
60
20
20
0
0 100 200 300 400 500 600
Width (nm)
As pec
tR
"
atio
. " 5
»
•
4
• <
• •-
•
012345678
Aspect Ratio
100
80
60
40
20
0
33
-------
Table 5-2: Average particle size and aspect ratio of micronized copper particles dislodged from the
wood surface after wiping with a polyester cloth as determined by analysis of BSE SEM
micrographs. Previous data for the as-purchased pressure-treated lumber is presented for
comparison. Aspect ratio is equal to the ratio between the particle length and width.
MCA-1
MCA-2
Wood Type As-purchased 1 Month As-purchased 1 Month
Average/Standard Deviation
Length
Width
Aspect
Number
121/65
56/42
2.5/1.2
314
168/132
65/52
3.0/2.1
537
244/125
105/58
2.7/1.6
370
291/175
136/83
2.4/1.3
378
A second set of filters was analyzed after two months of exposure (Figure 5-8). As with the first set, there
were no detectable particles associated with the wood matrix present on the 10 kDa filter papers. Copper
particles were identified on the 0.45 jam filter papers. The relative abundance of particulates present was
significantly less compared to the one-month sample. This is not unexpected given the reduction in the total
copper that was dislodged from the surface (Figure 5-1). After the two-month time period the copper
particulates collected were weathered/degraded (Figure 5-8). The well-defined rhombahedral crystals were
degraded and it was no longer possible to identify or distinguish individual crystallites in the BSE images.
A portion of the particulates present from the micronized copper treated lumber had also under gone
changes in their chemical composition as evidenced in Figure 5-8C and D. The EDX maps of the chemical
composition of the particles indicated there was no co-localization of Cu and C. The EDX map of O
indicated co-localization of Cu and O ruling out the presence of a purely metallic Cu particle. Instead the
images indicate that a portion of the malachite was transformed to either cuprite or tenorite (Cu2O and CuO,
respectively). The particle in Figure 5-8B appears to be a larger aggregate of smaller particles. However, it
is not possible to rule out the possibility of changes in the surface morphology of a single crystal. It is
important to note that not all of the copper carbonate was transformed to copper oxide. An example of this
is evident by the arrow highlighting the particle in Figure 5-8C.
Copper particles were also identified in the ACA wood sample after two months of environmental exposure.
It is difficult to determine if the particle in Figure 5-8E is a single crystal or an aggregate of smaller
materials. As with the materials identified in MCA-2, there is evidence of a copper oxide particle. The
number of Cu particulates identified in the ACA wood was less in comparison to the two micronized copper
treated wood materials.
While not clearly visible in Figure 5-8, all of the copper particulates identified on the filter were associated
with larger pieces of cellulose; again indicating that the copper particles are not present as individual
particles, but associated with larger cellulose particles. As discussed in the leaching section there is evidence
to support that copper is associated/bound to cellulosic materials and that copper carbonate exists as a free
particle/aggregate. Identifying the species/form of copper is important when considering potential exposure
pathways. From an inhalation perspective, the association of copper carbonate with cellulose will be
integral in predicting whether penetration depth in the lung will be based on the particle size of the dislodged
cellulose or the micronized copper particle size.
34
-------
A third set of filters was collected after 3 months. However, however no copper particles were identified
by SEM analysis.
Figure 5-8: BSE SEM Micrographs and EDX chemical maps of particulates wiped from the surface
of three different pressure-treated lumber samples and deposited onto a 0.45 \im filter paper. MCA-
1 (A), MCA-2 BSE micrograph and EDX maps (B, C, and D), respectively. ACA BSE micrograph
and EDX maps (E, F and G), respectively.
D
The change in particle size, shape, and chemistry with time is another important factor for risk assessment
purposes. Based on the current data, the potential for exposure to micronized copper particulates from skin
contact with the wood is greatly reduced after three months of environmental exposure. While the
environmental conditions will vary, this data, in conjunction with the wipe results from the un-weathered
lumber, suggests that particles in close proximity to the surface are readily dislodged. The strong correlation
between copper particulates and cellulose would also indicate that migration of the malachite from within
the wood structure to the surface is unlikely. Therefore a long term risk associated with the continued
release of copper particles seems unlikely unless the surface is cleared of weathered materials exposing an
un-weathered surface.
5.2.2 X-ray Absorption Fine Structure Spectroscopy (XAFS)
Copper speciation of the materials retained by the 0.45 jam and 10 kDa filters was assessed by XAFS. The
concentration of copper retained by the 10 kDa filters was not sufficiently great enough to allow for a
quantitative or even qualitative analysis of the XANES spectra. A copper edge was present, but it was not
35
-------
possible to further investigate the oxidation state or speciation of copper due to the poor signal to noise
ratio. XANES spectra were obtained from the 0.45 jam filters for the 1 and 2-month wipe samples (Figure
5-9). Similar to the microscopy samples, after 2 months the total quantity of Cu retained on the 0.45 jam
filters was not large enough to collect useful XAFS data. The 1 and 2-month spectra are presented with the
XANES spectra from the as-purchased materials for comparison. The quality of the spectra for the two
wipes is reduced compared to the original material and is a function of the total copper present. Hence, the
reduced quality of the spectra between 1 and 2 months is due to a decreased quantity of copper
dislodged/wiped from the wood surface. The dashed lines in Figure 5-9 highlight two areas where the
transformation of the copper from the original composition is present. The first dashed line near 8980 eV
indicates the location of the Cu(I) derivative peak, and the second dashed line near 8985 eV indicates the
location of the malachite derivative peak. Immediately evident in all three wood types for both wipe events
is the presence of Cu(I). The species is not associated with the original material indicating in situ formation.
Previous research has demonstrated that in an organic rich matrix Cu(II) will reduce to Cu(I) (Leal and Van
Den Berg, 1998; Kogut and Voelker, 2001; Pham etal, 2012; Yuan etal, 2012). In previous research, the
mechanism occurred through the reduction of the Cu(II) ion, suggesting that the formation of the Cu(I) is
controlled by the ionic Cu present and not copper carbonate. Also evident is the shift to lower energies of
the derivative peak near 8985 eV. The shift in peak location is associated with the change in the Cu(II)
species present.
Figure 5-9: Cu Ka XANES spectra for the as-purchased pressure-treated lumber and the material
dislodged from the wood surface and retained on a 0.45 (j,m filter after 1 month (Wipe 1) and two
months (Wipe 2). MCA-1 (A), MCA-2 (B), ACA (C). Dashed lines indicate the energy positions
associated with Cu1+ and malachite.
K
u?
I
I
MCA-1 Wipe 2
- MCA-1 Wipe 1
-MCA-1
Cu,CO,
-------
Table 5-3: Linear combination fitting results for Cu Ka XANES first derivative spectra for materials
wiped off of the treated wood samples exposed to the environment after 1 and 2 months of exposure
retained by a 0.45 \am filter.
Sample
MCA-1 as-purchased
MCA-1 Wipe 1
MCA-1 Wipe 2
MCA-2 as-purchased
MCA-2 Wipe 1
MCA-2 Wipe 2
ACA as-purchased
ACA Wipe 1
ACA Wipe 2
Month
Cu-Organic
CuiO
Average/Standard
0
1
2
0
1
2
0
1
2
88/1
49/4
20/5
42/3
52/5
16/5
90/5
73/3
67/5
10/1
5/2
14/2
14/3
10/1
5/2
Cu2CO3(OH)2
Deviation %
12/1
41/4
76/8
58/2
34/5
70/5
10/2
CuCh (aq)
18/2
28/4
*The summation of Cu-Acetate and Cu-Oxalate
The Cu(I) in the sample is likely present as a solid species based on the microscopy and LCF results. The
microscopy data indicated that Cu particulates were not co-localized with carbon, ruling out malachite as a
potential species. The LCF results did not indicate that CuO was present in the samples; therefore, it stands
to reason that the Cu(I) species is Q^O (cuprite). Additional Cu(I) present as an adsorbed phase cannot be
ruled out, but there is evidence supporting the formation of cuprite. The in situ formation of cuprite in
pressure-treated lumber has not previously been reported in the literature, and current available data does
not allow for the formulation of a mechanism to describe Cu(II) reduction and subsequent precipitation.
Copper (I) is meta-stable under ambient conditions and will eventually oxidize to Cu(II) as seen in the
decrease in the relative abundance of Cu(I) from 1 to 2 months for MCA-1 and ACA.
Over the two month time period the relative abundance of the Cu-organic complexes decreased while the
relative abundance of malachite increased for the two micronized copper treated samples. The biocidal
mechanism for micronized pressure-treated lumber is based on the slow dissolution and subsequent release
of aqueous Cu. Therefore, with time the ratio of malachite to ionic copper should increase as Cu is released
from the wood surface. The change in speciation of Cu in the two micronized wood samples supports the
biocidal mechanism. The reason for the increase in Cu-Organic fraction and decrease in malachite fraction
after one month for MCA-2 is unclear. The initial distribution of Cu species at the wood surface may be
different than the bulk composition (enriched in Cu-organic complexes) or it could be attributed to rapid
dissolution of nano-sized malachite crystals. Comparing the percent distribution of particle sizes between
the as-purchased MCA-2 and the MCA-2 wipe 1 reveals that there is little difference between the relative
abundance of particles for a given length (Figure 5-10). For particle widths there is a 10-25% reduction in
the relative abundance of particles of a given width between 50 and 100 nm (Figure 5-10). This shift to a
lower percentage of particle less than 100 nm after 1 month would indicate that a portion of the nano-sized
material underwent rapid dissolution. It is impossible to fully attribute the subsequent decrease in malachite
and increase in Cu-organic species totally to dissolution of nano-phases. Instead, a change in the distribution
of Cu species at the wood surface in combination with particle dissolution is likely responsible.
37
-------
Figure 5-10: Comparison of the percent distribution of particle size between the as-purchased MCA-
2 and the MCA-2 Wipe 1 samples. Length (A) Width (B).
•HI
MCA-2
MCA: Act i
0 75 150 225 300 375
0 100 200 300 400 500 600 700
Length (nm)
For ACA, there was a marked decrease in the Cu-orgainic phase coupled with an increase in ionic Cu(II).
The increase in ionic Cu is not associated with the MCA samples. The specific reason is unclear, but may
be related to the formulation of the initial treatment solution.
5.3 Copper Toxicity Assessment
A 2003 Consumer Product Safety Commission report pertaining to CCA wood provides a detailed review
of the potential toxicity of copper to humans (Osterhout, 2003). Dermal exposure to copper is not known
to be dangerous outside of potential allergic reactions, but ingestion of copper has been shown to carry risk
(Osterhout, 2003). The report summarizes several case studies and lists the acute and chronic levels of
copper that could produce adverse effects in humans if ingested. Table 5-4 details the most relevant copper
levels as they relate to the above results. Also shown in Table 5-4 are results from other published studies
as well as the Institute of Medicine of the U.S. National Academy of Sciences' list of tolerable upper intake
levels (TUIL) for copper for children, young adults, and adults, based on age (IOM, 2001; Stern, 2010).
Above these tolerable levels, mild symptoms, such as nausea, vomiting, and abdominal pain, may be
observed. More significant effects of copper exposure are possible. For acute effects, intake amounts are
extremely high (MOO mg per dose), which causes more severe symptoms and potentially death. For chronic
effects, doses over 20 mg d"1 have resulted in liver failure, requiring a transplant (Stern, 2010). These levels
are very high and likely not obtainable from PTL outside of extreme cases (i.e., direct consumption of large
quantities of PTL). Based on the reported toxicity levels, children between the ages of 1 and 13 are likely
the most at risk to copper toxicity from contact with PTL.
38
-------
Table 5-4: Summary of Copper Toxicity Levels.
Copper Dose
5.3mg
0.6 mg
1.2mg
O.Smg
l-2g
0.2 mg d'1
0.4 - 0.6 mg d'1
0.4-l.Omgd-1
S.Orngd'1
S.lSmgd'1
10.73 mg d-1
20 - 30 mg d'1
30 - 60 mg d-1
1 mg d'1
3 mg d-1
5 mg d-1
8 mg d'1
lOmgd'1
Effect
LOAELa
NOAELb
NOAELb
LOAEL3
Hemolytic Anemia
& Renal Damage
No effect
No effect
Cirrhosis
Cirrhosis
Abdominal Pain
and Vomiting
Abdominal Pain
and Vomiting
Liver Failure
&Transplant
Cirrhosis/Liver
Transplant
TUILC
TUILC
TUILC
TUILC
TUILC
Age / Age
Group
Adult
Infant
Adult
Adult
l-2y
Infant
Infant
Infant
3-4y
5-7y
Adult
Adult
Adult
l-3y
4-8y
9- 13 y
14-18y
19+ y
Dose Type
Acute
Acute
Acute
Acute
Acute
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Study
Wyllie 1957*
Olivares et al. 1998*
Arayaetal. 2001*
Olivares etal. 2001*
Osterhout 2003
Dessel de Vergara et al.
1999*
Olivares et al. 1998*
Mueller-Hoecker et al. 1988*
Trollman et al. 1999*
Spitalny 1984*
Spitalny 1984*
Stern 20 10
O'donohue et al. 1999*
IOM2001
IOM2001
IOM2001
IOM2001
IOM2001
a Lowest-observable-adverse-effect level for gastrointestinal symptoms
b No-observable-adverse-effect level for gastrointestinal symptoms
0 Tolerable Upper Intake Level
* As reported in Osterhout, 2003
To equate the exposure levels to the gathered data, Thomas et al. (2004) determined that the surface area a
child would come into contact with during a typical visit to a playground, the most likely place for a child
to encounter PTL, to be 1.29 m2. Using this area as an upper limit on typical contact with the treated
wood, an estimate can be made on how much copper could be wiped off and potentially ingested during a
playground visit. At 2 mg m"2, a value at the upper range found during testing, the amount of copper
released onto the hand would be 2.58 mg. The relative abundance for copper carbonate for the 2 MCA
samples after two months was -70% copper carbonate. Converting the total copper released to
micronized copper released onto the hand (70% of 2.56 mg) would equal 1.8 mg of particulate copper
carbonate. Comparing to the toxicity levels in Table 5-4, only children 3 years old and younger would be
effected, their TUIL being 1 mg d"1. Children ages 4-8 could be at risk if they consume more than 3 mg
of copper in their regular diet. At 2.56 mg of copper per visit, mild symptoms, such as nausea, vomiting,
and abdominal pain, could be observed.
Other factors may increase or decrease the toxicity. For instance, the area determined by Thomas et al.
(2004) assumes only three visits per week though the exposure and TUIL levels are daily limits.
39
-------
Averaged over a week, three visits to the playground would amount to 1.11 mg d"1, much closer to the
TUIL for the youngest age group. In order to arrive at a single value for the wiped area, they make
several assumptions about the average area of a child's hand and how much of it is contacted with the
wood as well as the number of contacts and duration of the playground trip, parameters that could
increase or decrease the exposure significantly on an individual basis. The levels also assume complete
ingestion of the transferred copper, which is unlikely. The TUIL values used for comparison are also
general recommendations and not known limits that guarantee symptoms will occur. Finally, the
micronized copper formulations release comparable amounts of copper to the aqueous formulation, which
has been on the market and in use for several years.
The above comparison focuses solely on the ingestion of total copper, regardless of form (ionic v.
particulate). A study conducted in parallel to this work examined the fate of micronized copper particles
after ingestion. They determined that the copper particles dissolved completely upon coming into contact
with stomach fluids and the copper was -90% bioavailable (Lenibel et al, 2015). This indicates that the
toxicity of the particles is not significantly different than the ionic copper. However, their study is not a
comprehensive investigation on the toxicity of micronized copper versus ionic copper.
40
-------
6 Conclusion
Representative micronized copper azole (MCA) and aqueous alkaline copper azole (ACA) pressure-treated
wood products were extensively characterized to determine the total copper and particle concentration, the
speciation of copper present in the wood products, and the particle size distribution. The three pressure-
treated wood types and an untreated wood sample were assessed for copper release during typical use:
transfer from hand contact and leaching from solution contact.
• The total concentration of copper present in the MCA treated lumber samples were similar and
significantly less than the ACA treated sample. Copper carbonate was identified as the solid phase
present in the MCA treated samples. Additionally, trace quantities of copper carbonate were
identified in the ACA samples by XRD and XAFS analysis. The copper speciation of the three
pressure-treated samples differ greatly. The MCA-2 sample was dominated by copper carbonate
(88%), the ACA sample by copper complexed with organics (90%), and a mixture of copper
carbonate and organically complexed copper in MCA-1.
• Microscopy data showed that the copper carbonate present in MCA-1 was approximately half the
size of the material present in MCA-2. Histograms, total count, and box plots of the particle width
and length indicated the data was not normally distributed but skewed towards smaller values.
Previous research has demonstrated that micronized copper is distributed on the cell walls. Copper
present in the BSE SEM images was always associated with cellulosic material and not present as
isolated particles or aggregates. Further, the appearance of the copper carbonate and physical
distribution was similar to the distribution of copper in the wood micro structure.
• The total quantity of solution copper released, during leaching, from the two MCA samples was
similar and significantly less than the total copper leached from the ACA treated sample. Based on
the solution concentrations, greater than -95% of the copper leached from the blocks was in ionic
form. The total solution concentration of copper leached from the three lumber types was similar
for the three different solution chemistries examined.
• Solid phase data from the leaching studies showed that the majority of the copper that was captured
during filtration was retained on the 2.5 jam filter membrane with significantly less material retained
by the 0.45 jam and 10 kDa membranes. XAFS speciation data showed different trends for the two
MCA samples. For MCA-1, the relative abundance of copper carbonate increased with smaller size
fractions while the reverse was true for MCA-2. In all three pressure-treated wood samples a Cu1+
species was present in the material retained on the 2.5 and 0.45 jam membranes. SEM BSE images
of the materials retained on the 2.5 and 0.45 jam membranes showed a decrease in the particle size
from the 2.5 to 0.45 jam membrane. A single particle was detected on the MCA-2 10 kDa
membrane. The aggregate was several microns in size and suggested that the copper carbonate had
aggregated on the membrane.
OO O
• Comparing the MCA wood to the ACA wood, MCA released substantially less copper than the
ACA, leading to a lower potential impact on the environment. ACA has already been shown to
have little impact on soil environmental communities in close proximity to PTL surfaces, indicating
the MCA would have a negligible impact.
• MCA treated wood, as well as ACA treated wood, released a large quantity of copper during the
initial wipe sampling events, and decreased with each additional sampling event. During the
41
-------
initial period of high release, contact with the wood causes far more depletion of the copper than
any type of environmental exposure. During testing, MCA-1 released slightly more copper than
MCA-2, possibly due to the smaller size of the copper particles in the MCA-1 formulation or the
increased concentration of organically complexed copper. The total quantity of copper released
for all of the wipe experiments exhibited the same trend of decreasing copper concentration
followed by a steady state after 3 to 4 sampling events. Light sanding of the weathered surface
after a year of exposure to ambient environmental conditions did not result in an increase in the
amount of copper transferred during subsequent wipe sampling events.
• Copper speciation by XAFS analysis was only possible for the first two months of the Outdoor
experiment. After three months the total quantity of copper retained on the filter membranes was
too low for bulk XAFS analysis. Both MCA samples of materials retained on the 0.45 jam
membrane increased in the relative abundance of copper carbonate after two months. Indicating
that organically complexed copper was the primary species released from the wood samples.
• After one month of environmental exposure there was no change in the particle length of copper
carbonate, however, there was an increase in the average width of the particles indicating active
dissolution of the nano-size copper carbonate. After two months of exposure the particles were
heavily weathered, and the presence of a Cu(I) oxide phase was identified. After three months
copper carbonate particles were not identified by microscopy analysis.
• The risk of copper exposure is low for older children (>8 y) and adults, but younger children (<8
y), particularly toddlers (1-3 y), may encounter quantities above the recommended thresholds,
resulting in mild symptoms of nausea, vomiting, and abdominal pain. This would only be a
potential in the upper bound of possibilities where a child had significant contact with the wood
three times a week and then ingested most of what accumulated on the contacted skin. Severe
acute and chronic symptoms are very unlikely during normal use.
• In summary, the particulate copper released from micronized copper PTL constituted a small
fraction (< ~5%) of the total released. The total copper released was less than or comparable to the
current ACA formulation available to consumers.
42
-------
7 References
Bunker, G., 2010. A Practical Guide to X-ray Absorption Fine Structure Spectroscopy. Cambridge
University Press.
Cobb, D., 2003. Chromated Copper Arsenic (CCA) CCA-treated Wood Analysis. Memorandum from
David Cobb to Patricia Bittner. U.S Consumer Product Safety Commission. Washington, DC
Cookson, L.J., Creffield, J.W., McCarthy, K.J., Scown, O.K., 2010. Trials on the Efficacy of Micronized
Copper in Australia. Forest Products Journal 60, 6-12.
Cushman, T., 2009. Pressure-Treated Wood: The Next Generation. Journal of Light Construction.
http://www.jlconline.com/codes-and-standards/pressure-treated-wood~the-next-generation.aspx#
Evans, P.D., Limaye, A., Averdunk, H., Turner, M., Senden, T.J., Knackstedt, M.A., 2012. Use of X-Ray
Micro-Computed Tomography to Visualize Copper in Preservative Treated Wood. The International
Research Group on Wood Protection Annual Meeting. The International Research Group on Wood
Protection, Kuala Lumpur, Malaysia, pp. IRG/WP 12-20488.
Evans, P., Matsunaga, H., Kiguchi, M., 2008. Large-scale application of nanotechnology for wood
protection. Nature Nanotechnology 3, 577.
Forest Products Laboratory, 2000. Environmental impact of preservative treated wood in a wetland
boardwalk: Res. Pap. FPL-RP-582. Madison, WI: U.S. Department of Agriculture, Forest Service, Forest
Products Laboratory. 126 p.
Fulda, B., Voegelin, A., Maurer, F., Christl, I., Kretzschmar, R., 2013. Copper Redox Transformation and
Complexation by Reduced and Oxidized Soil Humic Acid. 1. X-ray Absorption Spectroscopy Study.
Environmental Science & Technology 47, 10903-10911.
Institute of Medicine, 2001. Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron,
Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc. The
National Academies Press, Washington, DC
Kartal, S.N., Green, III, F., Clausen, C.A., 2009. Do the unique properties of nanometals affect
leachability or efficacy against fungi and termites? International Biodeterioraton & Biodegradation, 63,
490-495
Kelly, S.D., Hesterberg, D., Ravel, B., 2008. Analysis of Soils and Minerals Using X-ray Absorption
Spectroscopy. Methods of Soil Analysis. Part 5. Mineralogical Methods. SSSA, Madison, WI.
Kogut, M.B., Voelker, B.M., 2001. Strong copper-binding behavior of terrestrial humic substances in
seawater. Environmental Science & Technology 35, 1149-1156.
Leal, M.F., Van Den Berg, C.G., 1998. Evidence for Strong Copper(I) Complexation by Organic Ligands
in Seawater. Aquatic Geochemistry 4, 49-75.
43
-------
Lebow, S., 2004. Alternatives to Chromated Copper Arsenate (CCA) for Residential Construction. In:
Proceedings of Environmental Impacts of Preservative-Treated Wood Conference, 1-12
Lenibel Santiago-Rodriguez, Jennifer L. Griggs, Karen D. Bradham, Clay Nelson, Todd Luxton, William
E. Flatten, III, Kim R. Rogers. 2015. Assessment of the Bioaccessibility of Micronized Copper Wood in
Simulated Stomach Fluid. Environmental Nanotechnology, Monitoring & Management (in review).
Matsunaga, H., Kiguchi, M., Evans, P.D., 2009. Microdistribution of copper-carbonate and iron oxide
nanoparticles in treated wood. Journal ofNanoparticle Research 11, 1087-1098.
Matsunaga, H., Kataoka, Y., Kiguchi, M., Evans, P., 2010. Copper nanoparticles in southern pine wood
treated with a micronized preservative: Can nanoparticles penetrate the cell walls of tracheids and ray
parenchyma? , The International Research Group on Wood Protection Annual Meeting. The International
Research Group on Wood Protection, Biarritz, France. IRG/WP 10-30547.
Osterhout, C.A., 2004. Toxicity Review for Copper. Memorandum from Cheryl Osterhoutto Patricia
Bittner. U.S Consumer Product Safety Commision. Washington, DC
Pham, A.N., Rose, A.L., Waite, T.D., 2012. Kinetics of Cu(II) Reduction by Natural Organic Matter.
Journal of Physical Chemistry A 116, 6590-6599.
Service, U.S.F., 1936. Southern Yellow Pine: Technical Note 214. In: Laboratory, U.S.F.S.F.P. (Ed.).
United States Forest Service, Madison, WI.
Stern, B.R., 2010. Essentially and Toxicity in Copper Health Risk Assessment: Overview, Update and
Regulatory Considerations, Journal of Toxicology and Environmental Health, Part A. 73, 114-127
Stirling, R., Drummond, J., Zhang, J., Ziobro, R.J., 2008. Micro-Distribution of Micronized Copper in
Southern Pine. The International Research Group on Wood Protection Annual Meeting. The International
Research Group on Wood Protection, Istanbul, Turkey. IRG/WP 08-30479.
Thomas, T.A., Levenson, M.S., Cobb, D.G., Midgett, J.D., Porter, W.K., Saltzman, L.E., Bittner, P.M.,
2004. The Development of a Standards Hand Method and Correlated Surrogate Method for Sampling
CCA (Pressure)-Treated Wood Surfaces for Chemical Residue, Journal of Children's Health. 2(3-4), 181-
196
U.S. EPA, 1986. Acid digestion of sediments, sludges, and soils. Section A. Part I. Chapter 3. Metallic
analytes. 3.2, Sample preparation methods. Method 3050b. USEPA, Washington, DC.
U.S. EPA, 2007. Acid digestion of aqueous samples and extracts. Section A. Part I. Chapter 3. Metallic
analytes. 3.2, Sample preparation methods. Method 3015a. USEPA, Washington, DC.
Yu, L., Cao, J., Gao, W., Su, H., 2011. Evaluation of ACQ-D treated Chinese fir and Mongolian Scots
pine with different post-treatments after 20 months of exposure. International Biodeterioration &
Biodegradation 65, 585-590.
44
-------
Yuan, X., Pham, A.N., Xing, G.W., Rose, A.L., Waite, T.D., 2012. Effects of pH, Chloride, and
Bicarbonate on Cu(I) Oxidation Kinetics at Circumneutral pH. Environmental Science & Technology 46,
1527-1535.
45
------- |