Background on XRF Testing, Detailed Methodology
and Methodological Limitations
HealthyStuff.org product tests are performed with a portable X-ray
Fluorescence (XRF) analyzer. XRFs are widely used by both product
manufacturers and government regulators to screen consumer products
for hazardous chemicals. XRFs, like all test methods, have limitations.
Researchers selected products based on our research interests and
consumer interest. The sampling was not random or necessarily designed
to be representative of all products on the market. This page provides
an overview of the product testing methodology used by HealthyStuff.org,
XRF Testing Methodology
Researchers tested products using a handheld X-ray fluorescence (XRF)
device manufactured by Innov-X Systems (download XRF
Factsheet). The XRF analyzer uses a technology known as x-ray fluorescence
spectrometry to detect chemical elements, such as lead, cadmium,
chlorine, arsenic, mercury, tin, and antimony.
The elemental composition of the materials reveals the presence of potentially
hazardous chemicals, such as metals, and also allows researchers to infer the
possible presence of toxic chemicals or materials, including brominated flame
retardants (BFRs), polyvinyl chloride (PVC) and possibly phthalate plasticizers.
We have translated the research results into a HealthyStuff.org product rating
system to allow users to easily compare the chemical levels of a variety of
consumer products. There are a number of chemicals of concern that cannot be
detected by this technology.
As several different organizations contributed data to this project,
researchers conducted tests in several locations across the country. The
exact XRF device sampling protocol is described below.
XRF Calibration - XRF machines were calibrated at the beginning of
each round of sampling using the stainless steel clip, as well as the EC681
standard. Both calibrations were done once with a test length of 60 seconds.
Results were compared to the known levels in the sample. During testing, a
standardization was taken after every 200 readings, using both the stainless
steel clip and the EC681 standard. In addition, one time standardization was
performed with Wako Chemical USA ABS discs with cadmium, chromium, lead in
high & low concentrations
(NMIJ CRM 8106-a & NMIJ CRM 8105-a) and PVC disk standards (low
and high concentrations). This can lead to an averaging of the chemicals of concern when near-surface materials are non-homogenous. This may lead to an underestimation of the level of chemicals of concern in a product.
Testing - Standard testing time was 30 seconds. The
XRF machine was preset to test for the full 30 seconds. Only one test
was taken per component/material of a product. An attempt was made to
sample major components/materials of each product. The actual number
of samples per product varied. Because the X-ray penetrates from just
a few microns (metal) to 1/4 inch (plastics and other softer substrates),
the measurement should be considered a surface or near surface measurement.
Background sampling - Prior to the start of testing,
the background material was sampled with the XRF. The same background was used
for all samples, whenever possible. When backgrounds were used, they were free of all chemicals of concern.
Machine positioning - When sampling, the XRF machine
was held in a position perpendicular to the product being tested, without
compressing the material. Tests were conducted on the flattest part of
the product, and the entire analyzer window was covered by the material
being tested, when possible. Products were not dismantled to test interiors.
X-ray Flourescence (XRF) Backgound Material
Levels of Detection
- Comparison of testing of plastics for lead by x-ray fluorescence and traditional nitric acid digestion/ GFAA after muffle furnace combustion, November 8, 2008. Danielle Cappellini, B.Sc., MHA and Woodhall Stopford, MD, MSPH, Duke University School of Medicine.
From the Study:"Originally billed as a "screening" technique, these results suggest that in the range of concern, x-ray fluorescence can be used to determine accurately the presence of excessive levels of lead in plastic materials."
- Study of the Effectiveness, Precision, and Reliability of X-ray Fluorescence Spectrometry and Other Alternative Methods for Measuring Lead in Paint, August, 2009. U.S. Consumer Product Safety Commission
From the Study: "With appropriate test methods and SRMs[Standards Reference Materials], XRF spectrometry is suitable in many cases for the the determination of lead in polymers. CPSC staff has conducted comparison tesing of pplastic samples and SRMs by XRF and by using current wet chemical methods and found generally good agreement. XRF produced good results on homogeneous plastic SRMs with certified concentrations as low as 13.6 parts per million (ppm)."
- Linking PBDEs in House Dust to Consumer Products using X-ray Fluorescence, Allen, Joeseph, et. al., Environmental Science and Technology April 30, 2008.
From the Study: "The portable XRF analyzer appears to be a promising tool for characterizing the bromine content of products in homes and improves our understanding of the extent to which such products may act as sources of PBDEs in the indoor environment."
- Common Research Uses for XRF Technology
- Summary or Detailed Review
with Abstracts A summary of over 80 peer review research papers, from
dozens of research areas, which utlized XRF testing as a core analytical
method. XRF analyzers are usd by US Customs, FDA, EPA, DOE, & Consumer Agencies.
Read more about the use of XRFs for compliance screening.
- LODs for highlighted elements
are as follows:
|Observed Detection Limits
(30 second sample)(2)
||1% - 5%
(1) InnovX Model ABC XRF Detector detection estimates
based on 1-2 minute test times and detection confidence of 3-sigma, or
99.7% confidence. Detection limits are a function of testing time, sample matrix and presence of interfering
(2) Observed detection limits varied by type of material
being tested. Detection limits presented here are the lowest observed
from all testing.
Quality Assurance/Product Variation - In order to evaluate
the variation per product to assess and verify the accuracy of our testing,
some repeat samples were taken. This process took place once every 200
samples, and was done for at least one product in every product category.
Repeat samples are taken in three different ways:
- three readings taken from the same sampling location of one product
- three readings taken from three different sampling locations (consisting
of the same material, color, etc.) on the same product
- three readings taken from the same sampling location on three different
but identical products
All repeat sample data was recorded and submitted for review, but is
NOT included in the product database.
Data Interpretation - We interpreted the results using
the concentrations and deviations reported by the analyzer, together with
visual examination of the spectra generated by the instrument. The analyzer
reports concentrations of elements by analyzing the spectra using reference
data for the elements it reports, and measuring the area under the curve
in the spectrum. We visually examined the spectra to confirm the presence
of elements with known interferences (lead, bromine, and arsenic), and have not reported
them where we could not confirm presence.
The levels of lead, cadmium, chlorine, and other elements shown in this
website are those reported by the XRF analyzer manufactured by Innov-X
Systems, Inc. Our testing methodology uses standards with known levels
of certain elements to check the accuracy of the analyzer in one type
of matrix material. However, the products we tested are
made of many different types of materials, in some cases even within
the same product. The prescence of materials may interfere with the analyzer's
ability to quantify the elements accurately. When the materials
in a single product are not homogeneous, the test results may vary depending
on the orientation between the object under test and the testing device.
Where the testing is not able to isolate a single material, the reported
levels may represent an averaging of the levels in the different materials.
Interferences can occur between elements as well, such as with lead and
arsenic, resulting in poorer precision. Test results are reviewed for
Therefore, the levels we report provide a general indication of the levels
in the products in order to guide consumers on product choices. More exhaustive
testing with the XRF, as well as laboratory testing, could provide more
detailed findings on the levels of elements and associated compounds.
HealthyStuff.org ratings do not provide a measure of health risk or chemical exposure associated with any individual product, or any individual element or related chemical. HealthyStuff.org ratings provide only a relative measure of high, medium, and low levels of concern for several hazardous chemicals or chemical elements in an individual product in comparison to criteria established in the site methodology.