Determination of Critical Elements
in Foods in Accordance with US FDA
EAM 4.7 ICP-MS Method
Extending the scope of routine food analysis using
IntelliQuant data analysis
Authors
Jenny Nelson
Elaine Hasty2
Leanne Anderson
Macy Harris
1
Agilent Technologies, Inc.
2
CEM Corporation, USA
1
2
2
Introduction
Consumers expect that the food they buy will be safe to eat, so manufacturers take
steps to ensure the levels of harmful chemicals and pathogens are strictly controlled.
In addition, governments and regulatory bodies in many countries have a legal
obligation to protect public health in relation to food. The chemicals that are controlled
in foodstuffs include organic contaminants such as pesticide residues, and inorganic
contaminants such as heavy metals. In the United States (US), the Food and Drug
Administration (FDA) regulates a wide range of foods. The US FDA also publishes
details of the analytical methods that laboratories should use to help ensure food
safety. For example, FDA Elemental Analysis Manual (EAM) 4.7 is a comprehensive
method that describes how to determine 12 elements in food by ICP-MS following
microwave assisted acid decomposition. EAM 4.7 also outlines a series of quality
control (QC) tests to ensure instrument performance and data accuracy (1).
It is now easier than ever for food-testing laboratories using
the EAM 4.7 method to carry out the analysis with Agilent
ICP-MS instruments. Agilent single quadrupole ICP-MS use
an Octopole Reaction System (ORS4) cell operating in helium
(He) collision cell mode with Kinetic Energy Discrimination
(KED). This combination provides the optimum configuration
to control common polyatomic interferences, leading to
more accurate results. To improve detection of analytes
with intense background overlaps, such as Se-78, P-31, and
Si-28, the ORS4 cell can use an enhanced, high energy He
mode. Enhanced He mode provides low detection limits for
these interfered analytes, avoiding the need for a reactive cell
gas such as H2, O2, or NH3. Avoiding reactive gases ensures
that no new molecular interferences are formed in the cell,
improving the quality of the data and streamlining the
method (2). Agilent ICP-MS instruments have a wide (10 or
11 orders) linear dynamic range, so major and trace analytes
in food samples can be measured in a single run. The wide
dynamic range simplifies method setup by removing the need
for custom tuning conditions for major elements, while also
ensuring that fewer reruns occur due to over-range results.
Agilent High Matrix Introduction (HMI) technology further
improves the already exceptional plasma robustness,
enabling the ICP-MS to handle samples with total dissolved
solids (TDS) levels up to 3% (and up to 25% with Ultra (U)
HMI). HMI (and UHMI) uses aerosol dilution to handle
high-matrix samples, reducing sample preparation time
and minimizing the risk of introducing sampling errors or
contamination from conventional liquid dilution (3, 4). A
further benefit of HMI/UHMI is that it practically eliminates
matrix suppression, so varied, high matrix sample digests
can be run against a simple, synthetic calibration, without
requiring matrix-matching.
Agilent ICP-MS MassHunter IntelliQuant Quick Scan function
simplifies data review by collecting full mass spectrum data
for every sample in only a few seconds. The IntelliQuant
results provide semiquantitative concentrations for up to
78 elements, together with identification and confirmation
of unexpected elements by comparison with isotopic
abundance templates. The periodic table “heat map”
view of the results provides a quick and simple overview
of the concentration of all elements within the sample.
IntelliQuant can also calculate and display the total matrix
solids (TMS) content of each sample. TMS data provides a
useful indication of the total matrix load and any variation in
dissolved solids level through the batch (5).
This study describes the use of the Agilent 7800 ICP-MS and
Agilent SPS 4 autosampler for the analysis of 20 elements
in different food samples using a single He cell gas method.
The list of elements included the 12 elements that are
specified in EAM 4.7: arsenic, cadmium, chromium, copper,
lead, manganese, mercury, molybdenum, nickel, selenium,
thallium, and zinc. The data quality obtained for these
elements was assessed through the measurement of three
food standard reference materials (SRMs), a fortified method
blank (FMB), and two fortified analytical portions (FAPs). FAP
refers to samples that are spiked before sample preparation.
Experimental
Calibration standards
The calibration standards were prepared in 3% nitric acid
(HNO3) and 0.5% hydrochloric acid (HCl). HCl is routinely
added to samples for analysis using Agilent ICP-MS systems,
as it ensures that chemically unstable elements such as Hg
are retained in solution. Any Cl-based polyatomic overlaps
formed are easily controlled using the standard He cell mode.
Calibration standards were prepared from Agilent standard
solutions including environmental calibration standard, p/n
5183-4688, multi-element calibration standard-1, p/n
8500-6944, and 1000 µg/mL single calibration standard for
Hg, p/n 5190-8485. Most elements were calibrated from
0.1 to 25 ppb. Cu, Zn, and Mn were calibrated up to 250 ppb.
Hg was calibrated from 0.01 to 2.5 ppb. Continuing calibration
verification (CCV) standards were prepared at 1 ppb (2 ppb
for Hg), and/or 10 ppb.
An Agilent internal standard (ISTD) solution (part number
5188-6525), containing 2 ppm Sc, Ge, Rh, In, Tb, Lu, and Bi,
was prepared in 1% HNO3, 0.5% HCl, and 10% isopropanol
(IPA). Per the 4.7 method; IPA was added to the ISTD to
help equalize As and Se sensitivities due to residual carbon
post microwave digestion. The ISTD solution was added
automatically online at a flow rate approximately 16 times
lower than the sample flow.
Reference materials and samples
Three varied food matrix SRMs from National Institute of
Standards and Technology (NIST, Gaithersburg, US) were
used to validate the method. The SRMs used were NIST
1546a Meat Homogenate, NIST 1549 Non-Fat Milk Powder,
and NIST 2385 Slurried Spinach.
In addition to the SRMs, a diverse set of food samples with
different % composition of fats, proteins, and carbohydrates
were analyzed in this study. The following products were
bought in a supermarket in North Carolina, USA: beef jerky,
fortified nutritional shake, gouda cheese, gummy bears,
powdered donuts, and dark chocolate. Other samples
consisting of pepperoni, rice noodles, frozen dinner, and
frozen pizza were also digested and analyzed in the same
batch and the results are reported elsewhere (6).
2
Standard and sample preparation
All the food samples included in this study were digested
“as received”, except for the donut, which was crushed in
a bag before sampling. The samples were then prepared
for analysis according to the digestion procedure outlined
in the EAM 4.7 method using a MARS 6 closed-vessel
microwave digestion system located at CEM Corporation,
USA. Approximately 0.5 g of each food sample or SRM was
accurately weighed into a 75 mL PFA-lined MARS Xpress
vessel. 8 mL of HNO3 and 1 mL of H2O2 was added to each
vessel. Duplicates of the samples, SRMs, and spiked samples
(FAPs) were prepared and digested in a single batch, using
the heating program shown in Table 1. Each digestion batch
can accommodate up to 40 vessels containing a variety of
food sample matrices, with a single program being used for
all sample types. Finally, 0.5 mL concentrated HCl was added
to the digests, followed by de-ionized water to a final weight
of 100 g.
Table 1. Microwave digestion parameters.
ParameterSetting
Power (W)1800
Ramp Time (min)25
Hold Time (min)15
Temperature (°C)200
The analytical sequence of calibration standards, samples,
and QC solutions is shown in Figure 1. The sample block was
analyzed repeatedly with automatic insertion of the periodic
QC block after every 10 samples.
Instrumentation
An Agilent 7800 ICP-MS, which includes the ORS4 collision cell
and HMI aerosol dilution system, was used for the analysis.
The 7800 has been superseded by the Agilent 7850 ICP-MS,
but the configuration and analytical settings reported here
apply to both models. Sampling was performed using an
Agilent SPS 4 autosampler. The ICP-MS was configured with
the standard sample introduction system, consisting of a
MicroMist glass concentric nebulizer, temperature-controlled
quartz spray chamber, and quartz torch with 2.5 mm id
injector. The interface consisted of a nickel-plated copper
sampling cone and a nickel skimmer cone.
The Rare Earth Elements (REEs) have relatively low second
ionization potentials, so readily form doubly charged ions
(M2+) in the plasma. If REEs such as Nd, Sm, Gd, and Dy are
present in a sample at a high enough concentration, M2+
interferences can affect the accuracy of the measurement of
As and Se. Therefore, the EAM 4.7 method recommends that
146
147
analysts monitor the following isotopes:
163
Dy to assess the potential for M2+ overlaps. The Agilent
Nd,
Sm,
155
Gd,
ICP-MS MassHunter software includes an easy-to-use,
automated function to correct for the contribution that REE
2+
ions make to the signals measured for As and Se.
Unknown samples can also be easily screened for REEs (and
the entire periodic table) using the IntelliQuant function in the
ICP-MS MassHunter software (version 4.6 and later). IntelliQuant
works by performing a full mass-spectrum scan with only two
seconds measurement time. The IntelliQuant results provide
valuable information about the food samples, including:
• The elemental composition of each sample, including
REEs. The results can be displayed in a table or as a
periodic table heat map.
• Confirmation of an unexpected element using isotope
templates.
• An estimation of the Total Matrix Solids (TMS) level for
each sample or based on the analysis of a typical sample.
3
The TMS function uses the IntelliQuant semiquantitative
data to calculate the approximate solids levels of a sample.
The calculation excludes gas elements, such as Ar, O, and
N, together with C, P, S, and the halides, ensuring a more
accurate result. To simplify method setup and ensure the
best possible accuracy for uncalibrated elements, the
semiquant data (and TMS calculation) uses a mass response
profile generated from the quantitative calibration standards.
The TMS function is a great diagnostic tool to identify
possible causes of ISTD suppression. It is especially useful
when dealing with unknown and potentially difficult food
sample matrices and deciding if a sample needs to be diluted
or a higher HMI setting is needed. The measured TMS levels
for the samples analyzed in this study are shown in Table 2.
Table 2. Total matrix solids data for six food samples obtained by the TMS
function of ICP-MS MassHunter. Each sample was prepared in duplicate and
each preparation was measured twice. The data is corrected for dilution.
Concentration units: ppm.
Beef
Jerky
TMS36433 387215136768663717292
Fortified
Nutritional
Shake
Gouda
Cheese
Gummy
Bears
Powdered
Donuts
Dark
Chocolate
Main operating conditions of the Agilent 7800 – such
as plasma settings – are typically loaded from a preset
method, selected as appropriate for the sample types being
measured. In this case, a plasma preset of HMI-4 (aerosol
dilution factor of four) was used, based on the typical matrix
levels for the food types being measured (Table 2). When HMI
is selected, all related settings are autotuned as appropriate
for the matrix levels of the target sample types. Other
instrument operating settings were optimized automatically
using the ICP-MS MassHunter autotune function. All analytes
were acquired in He mode (enhanced He mode for Se).
Instrument operating conditions are listed in Table 3.
Table 3. ICP-MS operating conditions*.
ICP-MS ParameterSetting
RF Power (W)1600
Sampling Depth (mm)10
Nebulizer Gas Flow (L/min)0.6
Dilution (HMI) Gas Flow (L/min)0.35
Lens TuneAutotune
Helium Cell Gas Flow (mL/min)4.3 (10**)
Energy Discrimination (V)5 (7**)
* Shaded parameters are defined in the method and HMI- 4 plasma presets; all parameters
were automatically optim ized during star t-up and autotuning. * * Enhanced He mode settings
used for S e.
Results and discussion
Typical 7800 ICP-MS instrument detection limits (DLs) and
background equivalent concentrations (BECs) calculated from
the ICP-MS MassHunter calibrations are shown in Table 4. The
EAM method detection and quantification limits – also shown in
Table 4 – were calculated based on method blanks measured at
the end of the run, n=10 (7). Data was acquired for 20 elements,
including the 12 elements required by EAM 4.7, using only He
as a cell gas. The 7800 ICP-MS analytical limits are better
than the nominal limits provided in EAM 4.7 for all elements.
Table 4. 7800 ICP-MS detection limits and EAM 4.7 nominal analytical limits.
ICP-MS
MassHunter
ElementDL
(µg/kg)
27 Al 0.1480.4211.1679.103
51 V 0.0650.3110.9967.770
52 Cr 0.0320.1760.2752.1445.3948.9
55 Mn 0.0160.1251.2249.5462.3321.2
56 Fe 0.0120.6710.8756.827
59 Co 0.0040.0060.0290.225
60 Ni 0.0110.0730.3162.4636.3858.0
63 Cu 0.0080.5500.19639.326.0254.7
66 Zn 0.0490.2861.86514.55537.4340
75 As 0.0050.0590.2171.6931.2711.6
78 Se0.0160.0460.4043.1497.2866.1
95 Mo 0.0100.0141.1969.3335.1847.1
111 Cd 0.0020.0030.0370.2870.4083.71
121 Sb 0.0050.0120.0750.582
137 Ba 0.0180.0090.1511.176
201 Hg0.0030.0100.6815.3130.8617.82
205 Tl 0.0020.0010.1741.355*0.281*2.10
208 Pb 0.0030.0130.0340.2661.2010.9
232 Th 0.0060.0040.0520.409
238 U 0.0010.0020.0260.202
All elements were acquired in He mode (enhanced He fo r Se). The Nominal Analytical Limits are
given in E AM 4.7 and are base d on method blanks measure d during the single lab validation over
one year ; n = 143. *Based on a si ngle lab validati on (n = 27).
BEC
(µg/kg)
Verification of instrument calibration and sample
digestion process
As part of the method quality control procedure specified in
EAM 4.7, and to ensure the ongoing validity of the calibration,
a CCV standard was analyzed five times during the analytical
sequence. As shown in Figure 2, all elements in the five CCVs
and the initial ICV were recovered within the EAM acceptance
criteria of ±10% of the actual concentration.
Calculated based on
EAM 4.7 Analytical Limits
LOD
(µg/kg)
LOQ
(µg/kg)
EAM 4.7 Nominal
Analytical Limits
LOD
(µg/kg)
LOQ
(µg/kg)
4
Table 5. Mean measured concentrations in three food SRMs corrected for dilution. Mean calculated from two separate digests, each measured twice in triplicate.
* FDA Elemental Analysis Manual (Section 3.4 Spec ial Calculations) 3.4 Equation 20. * * Recoveries not calcu lated as the certified values are close to or b elow the method <LOQ. R = non- certified
reference value.
Conc
(mg/kg)
Measured
Conc
(mg/kg)
R
0.019119
R
0.079103
Recovery
(%)*
QC Criteria Certified
Conc
(mg/kg)
Measured
Conc
(mg/kg)
Recovery
(%)*
QC Criteria Certified
Conc
(mg/kg)
R
Measured
Conc
(mg/kg)
0.81891
Recovery
(%)*
QC Criteria
A spike recovery (FAP) test was carried out to check the recovery
of the digestion and the accuracy of the 7800 ICP-MS method for
food sample analysis. Two food samples (beef jerky and gummy
bears) were selected at random. Both samples were spiked at
two levels – 1 and 50 ppb – for all analytes and measured using
ICP-MS. For samples that had naturally occurring elemental
concentrations below 5 ppb, the 1 ppb spike recovery is reported.
For samples with higher naturally occurring concentrations,
the 50 ppb spike results are reported. The recoveries for all
elements in the fortified food samples were within the EAM
4.7 method QC criteria of ±20%, as shown in Table 6.
Figure 2. CCV recoveries over the course of the 12-hour sequence, including
at the end of the analytical sequence.
Quantitative results for food samples
Quantitative results are given in Table 7 for six food products.
In addition to the 12 elements specified in EAM 4.7, data
is provided for Al, V, Fe, Sb, Ba, Th, and U. According to
To verify the sample digestion process, each of the three
NIST SRMs was prepared in duplicate and each preparation
was analyzed twice (with three replicates per analysis)
using the 7800 ICP-MS. As shown in Table 5, the mean
concentrations were in good agreement with the certified
concentrations, meeting the QC criteria requirements of
the FDA EAM method of 80–120%. Since not all SRMs are
certified for all analytes, blank cells indicate the absence of a
certified or reference value.
Matrix effects and spike recoveries
To test for non-spectral interferences (matrix effects), two
fortified method blanks (FMB) were prepared by spiking
the blank at 1 and 50 ppb. The higher-level spike was used
for the elements Al, Fe, Cu, and Zn, while the low-level
spike was used for the remaining trace elements. The FMB
was analyzed periodically throughout the entire method
procedure. All recoveries were within the EAM 4.7 method
the IntelliQuant data for each sample, M2+ interferences
150Nd2+
from
156Dy2+
on 78Se+ were unlikely to be a problem given the low
and
150Sm2+
on 75As+, and from
156Gd2+
and
concentrations of the REEs.
The measured concentrations of Cd and Pb in dark chocolate
are high compared to the other foods. Chocolate is known
to contain Cd, which is present in the soil of some cocoa
plantations (8). The US doesn’t have limits for Cd in chocolate,
but the European Union (EU) introduced maximum levels on
January 1, 2019 (9). The limit for dark chocolate with total dry
cocoa solids above 50% is 0.80 mg/kg (800 ppb). The measured
value of 205 ppb is well below the EU maximum level.
The FDA recommends a maximum Pb level of 0.1 ppm
(100 ppb) in candy that is likely to be consumed frequently
by small children (10). The measured concentrations for Pb
in gummy bears and chocolate are well below the maximum
guideline level.
acceptable % recovery range of 90-110%, as shown in Table 6.
5
Table 6. Mean recovery results based on the analysis of food sample digests. Mean calculated from two separate digests, each measured twice in triplicate.
Figure 3. Stability of ISTD measurements over 12 hours. The ISTD recoveries have been normalized to the calibration blank for all samples.
ISTD recovery (%)
The analytical sequence outlined in Figure 1 was analyzed
repeatedly over 12 hours. All the ISTD recovery plots were
within ±20%, with no internal standard failures throughout the
run, meeting the criteria specified in EAM 4.7 (Figure 3). The
results demonstrate the robustness of the plasma and high
matrix tolerance of the 7800 ICP-MS with HMI over long runs.
IntelliQuant data
When an analyst develops a quantitative method using
an ICP-MS MassHunter preset method, an IntelliQuant
Quick Scan acquisition is predefined in the He mode tune
step. No special setup or separate calibration is needed
for IntelliQuant, simplifying the analysis. IntelliQuant
automatically acquires full mass-spectrum data in every
sample with only 2 s additional measurement time, allowing
the analyst to quickly see which elements are present in the
samples. Because IntelliQuant data is acquired in He collision
cell mode, analytes are free from common polyatomic ion
overlaps, ensuring the quality of the data.
In this study, IntelliQuant data was acquired for each food
sample with the 7800 ICP-MS operating in He mode. The
data can be displayed in a periodic table heat map view, as
shown in Figures 4 and 5. The color intensity heat map shows
the approximate concentration of up to 78 elements in each
sample, with a darker color indicating a higher concentration
of that element.
The IntelliQuant data provides a complete picture of the
elements present in the sample, as data can be reported
for elements not included in the calibration standards. This
benefit is demonstrated in the heat map display for donuts
(Figure 4), which shows a relatively high concentration
of Ti (labeled as food additive “TiO2” on the packaging).
The IntelliQuant semiquantitative result for Ti in the donut
samples was ~90 ppm. Figure 5 shows the heat map results
for dark chocolate, indicating a relatively high concentration
of Ca, Cr, Ni, W, and Pb. Figure 6 shows the IntelliQuant
Quick Scan spectrum for the dark chocolate, illustrating a
good fit to the natural isotope template for W, confirming the
presence of this element in dark chocolate.
Figure 4. Periodic table heat map view of ICP-MS IntelliQuant data acquired
for powdered donuts.
7
Figure 5. Periodic table heat map view of ICP-MS IntelliQuant data acquired
for dark chocolate.
Figure 6. The unexpectedly high concentration of W reported in dark
chocolate was confirmed by the isotope template fit in the IntelliQuant
Quick Scan mass spectrum.
Conclusion
The Agilent 7800 ICP-MS was used to analyze multiple
elements in a range of everyday foods in accordance with
US FDA EAM method 4.7 for food and related products.
All samples were prepared in the same batch using a single
microwave digestion method.
In-built features of the ICP-MS and ICP-MS MassHunter
software were used to simplify method development,
improve instrument performance, and ensure high-quality
data.
–A preset method and autotune were used to predefine
and optimize instrument operating parameters,
accelerating instrument setup.
–The ORS4 cell was operated with a single gas (helium-
KED mode), which is the standard operating mode for
quantitative analysis of unknown samples. As indicated
by the spike recovery data, He mode effectively removed
polyatomic interferences, enabling high-quality multielement data to be obtained for all elements, including As
and Se.
–IntelliQuant was used to provide semiquantitative
concentrations for all measurable elements present
in each sample, including confirming the unexpected
presence of nontarget elements. IntelliQuant also
provided an estimate of the total matrix content of the
food digests, helping with method optimization and
scheduling of routine maintenance.
The accuracy of the method was evaluated by analyzing
three food-based SRMs and conducting a spike recovery test
for 12 elements in two food samples. Excellent recoveries
were achieved in all cases. The instrument also exceeded the
nominal detection limit requirements specified in the EAM
method and showed excellent stability over a 12-hour run.
The study showed that Agilent ICP-MS instruments are
suitable for the routine, multi-element screening of trace
level elements in foods, making them ideal for food safety
management programs.
8
References
1. Patrick J. Gray, William R. Mindak, John Cheng, US FDA
Elemental Analysis Manual, 4.7 Inductively Coupled
Plasma-Mass Spectrometric Determination of Arsenic,
Cadmium, Chromium, Lead, Mercury, and Other
Elements in Food Using Microwave Assisted Digestion,
Version 1.2 (February 2020), accessed November 2020,
https://www.fda.gov/media/87509/download
2. Enhanced Helium Mode Cell Performance for Improved
Interference Removal in ICP-MS, Agilent publication,
5990-7573EN
3. High Matrix Introduction, Agilent ICP-MS technology
brief, 5994-1170EN
4. Wim Proper, Ed McCurdy, Junichi Takahashi,
Performance of the Agilent 7900 ICP-MS with UHMI for
High Salt Matrix Analysis: Extending the matrix tolerance
of ICP-MS to percent levels of total dissolved solids,
Agilent publication, 5991-4257EN
5. Agilent ICP-MS IntelliQuant Software: For greater sample
insight and confidence in results, Agilent publication,
7. William C. Cunningham, William R. Mindak, Stephen G.
Capar, US FDA Elemental Analysis Manual For Food
and Related Products, 3.2 Terminology, 2014, accessed
November 2020, https://www.fda.gov/media/89337/
download
8. Oliver Nieburg, Killing at source: How to avoid cadmium
and lead in chocolate, Confectionery News, 16Sep-2016, accessed November 2020, https://www.
9. Cadmium in Food, European Union, accessed November
2020, https://ec.europa.eu/food/safety/chemical_safety/
contaminants/catalogue/cadmium_en
10. US FDA, Guidance for Industry: Lead in Candy Likely To
Be Consumed Frequently by Small Children, accessed
November 2020, https://www.fda.gov/regulatory-