Food Additives and Contaminants
Vol. 29, No. 8, August 2012, 1183–1193
Measuring nanoparticles size distribution in food and consumer products: a review
L. Calzolai*, D. Gilliland and F. Rossi
European Commission – DG Joint Research Centre, Institute for Health and Consumer Protection, Ispra (VA), Italy
(Received 14 February 2012; final version received 26 April 2012)
Nanoparticles are already used in several consumer products including food, food packaging and cosmetics, and
their detection and measurement in food represent a particularly difficult challenge. In order to fill the void in the
official definition of what constitutes a nanomaterial, the European Commission published in October 2011 its
recommendation on the definition of ‘nanomaterial’. This will have an impact in many different areas of
legislation, such as the European Cosmetic Products Regulation, where the current definitions of nanomaterial
will come under discussion regarding how they should be adapted in light of this new definition. This new
definition calls for the measurement of the number-based particle size distribution in the 1–100 nm size range of
all the primary particles present in the sample independently of whether they are in a free, unbound state or as
part of an aggregate/agglomerate. This definition does present great technical challenges for those who must
develop valid and compatible measuring methods. This review will give an overview of the current state of the art,
focusing particularly on the suitability of the most used techniques for the size measurement of nanoparticles
when addressing this new definition of nanomaterials. The problems to be overcome in measuring nanoparticles
in food and consumer products will be illustrated with some practical examples. Finally, a possible way forward
(based on the combination of different measuring techniques) for solving this challenging analytical problem is
illustrated.
Keywords: method validation; metals; packaging; water; beverages
Introduction
Nanotechnology is having a large socio-economic
impact in many fields of industrial activity: a large
number of consumer products containing nanomaterials
are already on the market and there are great
expectations for the role that nanoparticles-based
materials could play in fields as disparate as medical
diagnostics and photovoltaic cells. At the moment
nanomaterials are also used in a wide range of different
applications in the area of consumer products, i.e.
potentially coming into direct contact with the general
public. The use of nanotechnology-based ingredients
additives and contact materials is expected to grow in
the near future (Tiede et al. 2008) and in some
countries nanomaterials are already used in alimentary
supplements and food packaging (Chaudhry et al.
2008).
These very fast developments have spurred scientific
and regulatory activities aimed at maximising the
benefits of products containing nanomaterials while
minimising their potential toxic effects. For example
the Organisation for Economic Co-operation and
Development (OECD) has set up an international
collaboration through the Working Party on
Manufactured Nanomaterials (http://www.oecd.org/
sti/nano) to advise upon emerging policy issues related
to the responsible development of nanotechnology.
The testing activities that have been performed by
different research groups and governmental organisations
up to now have been developed using a variety of
definitions of ‘nanomaterial’, since an official definition
did not exist. Until now nanomaterials have been
regarded as materials (Borm et al. 2006) with at least
one dimension below 100 nm (Loevestam et al. 2011).
In this context, the European Commission (2011)
published its recommendation on the definition of
nanomaterials. The recommended definition states
that:
Nanomaterial means a natural, incidental or manufactured
material containing particles, in an unbound
state or as an aggregate or as an agglomerate and
where, for 50% or more of the particles in the number
size distribution, one or more external dimensions is in
the size range 1 nm–100 nm.
Thus, to detect (and measure) the presence of
nanomaterials in food according to this definition
requires analytical methods able to determine the
number size distribution of particles at least in the
1–100 nm size range.
*Corresponding author. Email: luigi.calzolai@jrc.ec.europa.eu
ISSN 1944–0049 print/ISSN 1944–0057 online
ß 2012 Institute for Health and Consumer Protection, Italy
http://dx.doi.org/10.1080/19440049.2012.689777
http://www.tandfonline.com
This paper will address the known (or proposed)
uses of nanomaterials in the food sector and give an
overview of the available analytical techniques that
could be used to measure the number size distribution
of nanoparticles in food. Finally it will critically review
the pros and cons of the various techniques and
propose a framework for developing an integrated
approach to the development of a robust analytical
platform for the detection of nanoparticles in food and
consumer products. This activity is of particular
importance and urgent in view of the recent
European Cosmetic Products Regulation that requires
labelling nanomaterials in the list of ingredients
(European Parliament and European Council 2009),
and the regulation on food information to consumers
(Regulation EU No. 1169/2011).
Overview on nanoparticles in food, food packaging
and cosmetics
Information on the use of nanomaterials in consumer
products including food and food-related products at
the moment is somewhat sketchy and hypothetical as,
up to now, there are no legal requirements for
reporting such information to consumers and/or regulatory
authorities. The year 2013 will see entering into
force the new European regulation EC 1223/2009 on
cosmetic products that requires the labelling of ingredients
present as nanomaterial. The application of
nanotechnologies in the food industry, and its implication
in terms of risks and regulation, has been
recently reviewed by Cushen et al. (2011). They
concluded that progress in the field may be stifled by
lack of governance and potential risk.
The use of nanomaterials in the food industry can
be divided in three broad categories: food, food
additives and food packaging with various products
for each category already on the market (Blasco and
Pico` 2011). For example, a recent market research
report on ‘Nano-enabled packaging for the food and
beverage industry’ (Innovative Research and Products
Inc. 2009) estimated that the total nano-enabled food
and beverage packaging market to have been US$4.2
billion in 2009 and predicted an annual growth rate of
around 11% for the next years. Another very recent
review on the detection of organic nanoparticles in
food by Peters et al. (2011) points out the use of
nanotechnology for food-processing applications, such
as the ceramic NanoGlaze in cookware to prevent food
sticking. Recognising the developments in the field and
the possible applications of nanotechnology to the
food sector the European Food Safety Authority
(EFSA) published in 2011 the first practical guidance
for assessing nano applications in food and feed.
Some metal and metal oxides compounds approved
as food additives are also produced in nano forms and
are currently used in different consumer products.
Table 1 reports some of these compounds with their Enumber
classification, properties and possible use in
the food sector. Titanium dioxide, silicon oxide and
metallic silver are probably the most widely used
nanoparticles currently on the market. Although there
is little reported evidence of the deliberate use and
presence of nanoparticles in commercially available
food and food packaging, there are several food-grade
nanoparticle products on the market (such as foodgrade
silica) and thus their presence in some alimentary
products can be considered as being likely. However,
they are also used in cosmetics, where titanium dioxide
particles smaller than 100 nm in size are used as
transparent sunscreen blockers in contrast to ‘normal’
white titanium dioxide.
Techniques to measure nanoparticles size
The characterisation of nanomaterials and nanoparticles
is not a trivial task, especially in the case of
materials in complex matrices (such as food) and below
the 100 nm size limit. At the moment, to the best of our
knowledge, there is no single technique able satisfactorily
and routinely to measure the number particle size
distribution of objects in the 1–100 nm size range. In
practice the only method that is technically capable of
addressing the need to count and size particles in both
free and agglomerated states is transmission electron
microscopy. Unfortunately this method has a major
disadvantage in terms of cost and complexity and in
practice it is unlikely to be suitable for undertaking
numerous routine measurements.
Table 1. Metal and metal oxide compounds in food and food-contact materials.
Material Classification (E number) Uses
Titanium dioxide E171 Colouring agent (white)
Iron oxides E172 Colouring agent (various colours)
Metallic silver E174 Labelling colour for food packaging
Metallic gold E175 Colour
Silicon dioxide E551 Anticaking agent
1184 L. Calzolai et al.
Based on a recent literature survey – searching
peer-reviewed articles on Science Direct (http://
www.sciencedirect.com) combining nanoparticle and
sizing technique – the most cited techniques used for
measuring the size of nanoparticles are: electron
microscopy (cited in around 85% of the cases), laser
light-scattering (around 10% of citations), and field
flow fractionation, centrifugation techniques, particle
tracing analysis (combined all together in around 4%
of cases). These techniques, together with an emerging
technique called single-particle ICP-MS, show the
potential to partially address the measurement of the
number size distribution of nanoparticles and will be
discussed in the following sections.
Electron microscopy
Performing a search on the techniques used for the size
characterisation of nanomaterials using peer-reviewed
literature (SciFinder database at http://www.scifinder.
cas.org) indicates that the most cited techniques are
electron microscopy based: either scanning electron
microscopy (SEM) or transmission electron microscopy
(TEM). These techniques, while very accurate,
require some sample preparation: samples have to be
adsorbed on a grid surface and any volatile solvent has
to be removed to permit their introduction into the
high-vacuum environment common to the majority of
SEM and TEM instruments (Dudkiewicz et al. 2011).
The problem of sample drying can potentially be
reduced by the use of so-called ‘environmental’ SEM/
TEM systems which are more tolerant of water or
solvent containing materials (Lorenz et al. 2010).
These solutions, although feasible, use complex and
expensive equipment that also requires highly specialised
technical support. On the positive side, electron
microscopy techniques can handle samples containing
mixtures of nanoparticles not only of different sizes,
but also of different shapes – a characteristic which is
not measurable by most other methods. Furthermore,
electron microscopy provides true particle size measurements
(see example in Figure 1), while optical,
chromatographic and centrifugal methods determine
the apparent hydrodynamic size. This difference,
between true and hydrodynamic diameter, must be
recognised and taken into consideration as it will lead
to an overestimate of particle size with respect to true
particle size for spherical particles. In the case of nonspherical
particles, the non-imaging methods cannot
provide any reliable measure of size without detailed
prior knowledge of particle shape data which can only
be determined by electron microscopy.
For this situation there are already available
specialised image analysis software which can be
adapted partially to automate the labour-intensive
process of measuring the size of the relatively large
number of particles needed for meaningful statistical
analysis of the results (Tiede, Tear, et al. 2009). Time,
complexity and cost may prohibit the routine use of
electron microscopy for determining nanoparticle size
distributions, but the technique may instead provide
the means to produce well-characterised standards
which could then assist in calibrating the more
accessible optical or chromatographic method.
Laser light-scattering techniques
Of the techniques that can directly work with liquid
samples, the most widespread methods are based on
laser light-scattering techniques: the most used are
either known as photon correlation spectroscopy
(PCS) or dynamic light-scattering (DLS) (Brar and
Verma 2011). These techniques are quite simple to use,
fast, relatively low cost, but while performing well
when dealing with samples of nanoparticles of a single
size (monodispersed) they can give misleading results
when used to analyse samples containing particles of
different sizes. For example, the authors have recently
shown that using dynamic light scattering to measure a
mixture of gold nanoparticles of 5, 15 and 45 nm (with
a 350:15:1 number of particles ratio) practically shows
only the presence of gold particles of 45 nm (Calzolai
et al. 2011). These data highlight that while this
technique can be applied to both diluted and concentrated
solutions (Kaszuba et al. 2007), it has a limited
ability to resolve mixtures of particles of different sizes.
This inability to detect the presence of smaller particles
among bigger ones is due to the fact that the scattering
intensity depends by the sixth power of the particle
radius (Berne and Pecora 2000; Gun’ko et al. 2003)
Figure 1. Scanning electron microscopy (SEM) analysis of
two samples of citrate-stabilised gold nanoparticles of
different sizes: (A) SEM micrograph of 3 nm gold nanoparticles;
(B) SEM micrograph of 12 nm gold nanoparticles; and
(C) particle size distribution obtained by semiautomatic
image analysis of sample A (left) and B (right).
Food Additives and Contaminants 1185
and thus larger particles tend to cover the signal
coming from smaller ones. In practice, this limits the
size resolution to the point where reliable separation of
particles can be achieved only when there is a factor of
3–4 difference in size.
A laser diffraction-related technique, particle tracking
analysis (PTA), exploits two important physical
phenomena of (nano)particle behaviour when suspended
in a liquid: the ability of individual particles to
scatter light and the characteristic movement of
particles produced by the effect of Brownian motion.
A specially designed optical cell containing a dilute
solution of particles is illuminated with a laser light
source and, using an optical microscope, the pin-points
of light scattered by the rapidly moving particles or
aggregates are observed and recorded using a highly
sensitive video camera. The video image obtained can
be analysed so as to allow particles to be located,
individually identified, and their movements and trajectories
followed on a frame-by-frame basis. Since the
diffusion of each individual particle depends only on
liquid viscosity, temperature and particle hydrodynamic
diameter, it is possible, through the Stokes–
Einstein equation, to calculate the particle size. When
the movements of a statistically relevant number of
particles in a population are evaluated over a suitable
lapse of time, reliable statistics for particle number–size
distributions can be obtained. Since this technique
calculates particle size on a particle-by-particle basis, it
is effective in overcoming the inherent weaknesses of
the DLS and SLS methods when confronted with
mixtures of relatively similarly sized particles. This
method has a number of important advantages including
relatively low instrument cost and high sensitivity
which can detect nanoparticles at concentrations as
low as low as 106
particle / cm3
(Gallego-Urrea et al.
2011).
These advantages are unfortunately contrasted by a
number of important disadvantages when considering
its application to the detection of nanoparticles as
specified in the recommended definition. The first
disadvantage is that the process of image analysis
requires a significant input from the operator, and as
such may be subject to accidental biasing towards
larger or smaller particles. The second, and probably
most important, disadvantage of this technique is that
it has a fundamental limitation on the lowest particle
size that can be detected. The scattering of light by a
particle in solution depends on a number of factors
including the wavelength of the scattering light, the
refractive indices of the particle and liquid and, most
importantly, the size of the particles. In practice the
effective lower limit is dependent on a combination of
these factors, but detecting particles below 20 nm
becomes problematic for many materials other than
those with high refractive indexes such as gold or TiO2.
It should be noted that this limitation has been reduced
by recent advances in laser diode technology which
have resulted in the availability of powerful, stable
near-UV light sources which can be used as an
alternative to the red light sources more commonly
used in this technology. The use of shorter wavelength
light permits more efficient detection of small particles
but reaching the 1 nm limit in the definition remains
beyond the capabilities of this technology.
Centrifugation-based techniques
Other interesting techniques, even if not very widely
cited in the peer-reviewed literature, are centrifugal
particle sedimentation (CPS) and analytical ultracentrifugation
(AUC). The major feature of these techniques
compared with the previously mentioned
optical methods is that they are more effective in
dealing with particle size mixtures. In the case of CPS
and AUC the principle of operation is dependent on
the fact that particles generally have a density which is
different (usually higher) than that of the liquid in
which they are suspended and consequently gravity or
centrifugal forces will tend to make them sink or float
in the liquid (Kamack 1951; Coelfen 2004). In CPS and
AUC, particles suspended in a liquid are subjected to a
centrifugal force induced by the acceleration of a rotor
that tends to move the particles through a disc or
column of liquid in the direction of the force. Since this
force is proportional to particle mass (and consequently
also to density), the larger particles sediment
faster than smaller ones (provided the mass of
the particle is greater than the displaced mass of the
solvent). The buoyant force (governed by the
Archimedes’ principle) and frictional force act in
opposite direction to the centrifugal force, impeding
sedimentation. The frictional force is generated by
movement of the solute through the solvent according
to the hydrodynamic treatment of viscous drag and is
proportional to a frictional coefficient and the solute
terminal velocity. The three forces come into balance
very quickly (within approximately 10À6
s) and the
particle achieves terminal velocity that is related to
particle size. This process of sedimentation can be
modelled by a theoretical treatment of thermodynamic
and hydrodynamic principles which, when combined
with experimentally determined particle sedimentation
times, can be used to determine particle size.
These two techniques offer two great advantages
when compared with the DLS optical methods. The
first advantage is the much better size resolution such
that particle size can be determined with an accuracy
better than 2–3 nm in the range of interest for the
definition. The second advantage is that as the particles
sediment, they are effectively separated in size before
passing though a (usually optical) detector which can
potentially be calibrated to give a quantified measure
1186 L. Calzolai et al.
of the amount of material of any given size passing
through the field of the detector (Carney et al. 2011).
Unfortunately these methods also have some important
limitations that must be considered. While both
methods operate on basically similar principles, the
equipment used for AUC permits access to much
higher centrifugal forces (4100,000 g) which allows the
more rapid sedimentation of small particles. This
difference becomes critical when considering small,
low-density particles, which in the case of CPS may
sediment very slowly or in extreme cases may not
sediment at all. In practice, this may limit CPS to a
minimum size of 3–5 nm for dense particles such as
gold ( ¼ 19.3 g /cm3
) and as much as 20 nm for
amorphous silica ( ¼ 1.8–2.2 g /cm3
). This limitation
may be overcome in AUC by the higher centrifugal
forces, but the accuracy of this method may be
compromised by another complication that is
common to both methods: as previously described,
the sedimentation process depends on the mass of the
particles, therefore on their size and density. Thus,
without a reasonable knowledge of particle density, it
is not possible to calculate accurately the size from
sedimentation data. This problem becomes particularly
relevant in any sample where there is the formation of
aggregates since even in the case of primary particles
with known density but with a tendency to aggregate,
it will be difficult to know the exact density of those
aggregates. This problem is further exasperated when it
is necessary to consider the potential presence of
agglomerates, which are also relevant to the definition.
Single-particle ICP-MS
Ion-coupled plasma mass spectrometry (ICP-MS) is an
ideal method for measuring traces of inorganic
elements. Operated in the ‘single particle’ mode (SPICP-MS),
the technique allows the detection of individual
nanoparticles without prior digestion. SP-ICPMS
is particularly interesting and is the object of
intensive development. It consists in measuring the
signal emitted by the vaporisation of the nanoparticles
in the inductive plasma with a time resolution of
typically 10–20 ms. Instead of an average background
proportional to the concentration of the element
present in the sample, sharp peaks are analysed,
whose intensity depends on the size of the nanoparticle
in the sample. A size distribution can thus in theory be
reconstructed. The technique has been developed and
modelled and it was found that the signal depends
mostly on the fraction of NP reaching the plasma zone,
a parameter difficult to access and setup. The technique
has been recently applied successfully to the
analysis of silver nanoparticles in wastewater (Mitrano
et al. 2012), but there are still problems to be resolved
and it is not clear at this stage how general the use of
SP-ICP-MS could become.
Field flow fractionation
The final method that will be considered is field flow
fractionation (FFF), which is a one-phase chromatography
technique (schematically shown in Figure 2).
The flow and sample are confined within a channel
consisting of two plates separated by a spacer foil. The
upper channel plate is impermeable, while the bottom
channel plate is permeable and made of a porous frit
material covered with an ultra-filtration membrane
that retains the sample in the channel. A syringe pump
is connected to the lower channel plate and when
operated it can pull liquid through the dialysis membrane
and out of the flow channel.
Figure 2. Schematic view of a flow field flow separation channel showing smaller particles moving closer to the centre of the
channel in a higher velocity zone and thus existing before larger particles.
Food Additives and Contaminants 1187
A laminar flow of the liquid within the flow
channel produces a parabolic flow profile in which
the stream moves slower closer to the boundary edges
than it does at the centre of the channel flow. When the
syringe pump creates a perpendicular force field (crossflow)
across the flowing, laminar stream, the analytes
are driven towards the boundary layer or ‘accumulation
wall’ of the channel.
Diffusion associated with Brownian motion, in
turn, tends to counteract this motion. Particles with a
smaller mean hydrodynamic diameter, which have
higher diffusion rates, tend to reach an equilibrium
position closer to the centre of the channel, where the
longitudinal flow is faster. Thus, the velocity gradient
flowing inside the channel separates the different sizes
of particles, with smaller particles eluting before the
larger ones (Giddings et al. 1976). The flow of liquid
eluted from the separation channel can then be
introduced into any one of a variety of different
instrument for the detection and characterisation of
the size-separated particles.
Since the FFF technique is a size separation
method that can be coupled to a wide variety of
possible particle detectors, the detection limit depends
mainly on two factors: (1) the detector chosen and (2)
any loss of materials that can occur by irreversible
absorption of particles onto the dialysis membrane.
Common detection methods are UV absorption,
refractive index, multi-angle light-scattering (MALS)
as well as fluorescence. In the case of inorganic
nanoparticles, the method may be combined with online
ICP-MS, which can provide direct quantification
of particle quantity as function of size. In this case, the
combination of FFF with ICP-MS would seem to be
one of the most promising routes for high sensitivity
detection and quantification of nanoparticles according
to the recommended definition. The FFF is a
highly promising technique (von der Kammer et al.
2011): it can separate and measure complex mixtures
containing NP of different sizes, it can separate
particles down to 1 nm size, it has an excellent dynamic
range, and the various components can be recovered
for further analysis. This possibility of recovering the
size-separated components offers an important opportunity
for further characterisation of the nanoparticles,
such as for example of the NP–protein complexes as in
the case of biological systems (Laera et al. 2011).
FFF suffers from a limited precision in measuring
the absolute size of NP if the measurement is
determined on the basis of the particle elution time
from the separation channel. Fortunately, this limitation
of FFF can be largely overcome if the technique is
used in combination with other sizing techniques such
as light scattering. In this case, the light-scattering
methods can be more reliably used for particle sizing
since the FFF ensures that they are presented with
solutions of near monodispersed particles rather than
complex mixture of sizes.
One of the main limitations of FFF is that the
separation process does not distinguish between nanoparticles
and aggregates of the same size nor can it
distinguish particles with different shapes which may
have the same mean hydrodynamic diameter. These
considerations suggest the need to use different techniques
sensitive to different NM properties (for
instance FFF sensitive to the hydrodynamic diameter,
and CPS sensitive to size and density). In addition, the
interaction of nanoparticles with the dialysis membrane
could cause problems, and care should be taken
into optimising the experimental conditions, such as
the material of the membrane and eventual addition of
surfactants to the cross-flow buffer.
Comparison of methods
A qualitative comparison of the different techniques is
shown in Table 2, which reports the suitability of the
different techniques with respect to several parameters
that the authors consider critical for measuring the
number size distribution as dictated by the new
definition of nanomaterials. Table 2 indicates that
Table 2. Qualitative evaluation of the relative advantages and disadvantages for different techniques to measure the size of
nanoparticles in the 1–100 nm size range.
SEM TEM FFF CPS PTA AUC DLS SP-ICP-MS
Minimum size þþ þþþ þþþ þ þ þþ þþþ þ
Dynamic range þþþ þþ þþ þþþ þ þþ þþþ þþ
Accuracy of measure þþ þþ þ þ þ þþ þ þ
Suitable for mixtures þ þ þþ þþ þþ þþ À þþ
In-situ measure À À þ þ þþ þ þþ þþ
Ease of use À À þ þþ þ þ þþ þ
Cost À À þþ þþ þþ þ þþþ þ
Notes: Different properties are evaluated as: excellent (þþþ), good (þþ), fair (þ) and insufficient (À).
AUC, analytical ultracentrifugation; CPS, centrifugal particle sedimentation; DLS, dynamic light-scattering; FFF, field flow
fractionation; PTA, particle tracking analysis; SEM, scanning electron microscopy; SP-ICP-MS, single particle inductively
coupled plasma; TEM, transmission electron microscopy.
1188 L. Calzolai et al.
electron microscopy-based techniques (SEM/TEM) are
probably the most accurate and universally applicable
techniques but they suffer from the need of extensive
sample preparation that can potentially introduce bias
in measuring the particle size distribution relative to
the starting sample. They are also quite expensive to
acquire and run and are not amenable even to medium
levels of throughput. Centrifugal particle sedimentation
(and to a lesser extent analytical ultracentrifugation)
could represent a lower-cost alternative to
electron microscopy techniques. Unfortunately CPS,
in the case of low-density particles, may not be able to
detect particles much smaller than 20 nm and the
different components cannot be recovered after
separation.
FFF is a promising alternative, especially if used in
combination with sizing techniques such as lightscattering
methods.
It must be noted that most of the existing work has
been performed on pristine nanoparticles in ‘simple’
matrices, such as water and water plus buffer systems.
The challenge in the case of real systems is much more
complex due to the huge variety and composition of
food matrices in which nanoparticles could be embedded
and on the large range of concentrations that need
to be analysed.
Analysis of nanomaterials in food products and
cosmetics
Several reviews have covered both the available analytical
techniques for identifying and characterising NP
in food (Luykx et al. 2008) and in complex matrices in
general (Tiede et al. 2008). Very recent reviews have
also covered the first examples of applications to the
analysis of NP in food matrices both in general
(Blasco and Pico` 2011) or targeted towards the use of
specific techniques such as flow FFF (von der Kammer
et al. 2011), particle tracking analysis (Gallego-Urrea
et al. 2011) and electron microscopy (Dudkiewicz et al.
2011). These reviews provide an excellent overview on
the use of the different techniques to detect nanoparticles
in complex matrices and we refer to those
references for an in-depth analysis of each technique.
The general review of Blasco and Pico` (2011) reports in
their Table 4 some examples of applications for
detection and characterisation on nanomaterial in
food and food-like matrices. The reported examples
used a large variety of different experimental techniques,
and in most cases they were limited to simple
detection thus underlining the inherent difficulty of
measuring the size of NP in food matrices.
Among the different techniques, FFF (coupled to
various detectors for NP sizing and identification)
seems to emerge as one of the few techniques (probably
together with TEM) able to measure the size
distribution in the case of samples containing nanoparticles
of different sizes in the 1–100 nm range, even
if the limited number of examples suggests a cautious
approach. A detailed analysis of the use of FFF for the
characterisation of NPs in different matrices was
carried out by von der Kammer et al. (2011); they
concluded that while there are already several examples
of using FFF for NPs in environmental samples, the
literature dealing with NPs in food samples is relatively
scarce. There are examples of the use of FFF for the
analysis of milk suspensions (Saeseaw et al. 2005) or
liposomes (Hupfeld et al. 2009).
In one excellent work, Schmidt et al. (2011) coupled
FFF with light-scattering detection (DLS and MALS),
and with ICP-MS, to measure quantitatively the size
distribution of mixtures of gold and polystyrene
nanoparticles (with diameters in the 10–100 nm size
range) in both water and biological samples. In
particular, rats were exposed to mixtures of gold NP
of 10 and 60 nm, and their livers were subsequently
analysed to detect the presence of gold NP. Even with
extensive sample treatment (homogenised liver were
treated with tetramethyl ammonium hydroxide), the
separation of AuNPs by FFF was not possible due to
the association of AuNP with undissolved biological
tissues. This example just shows the challenges in the
sample treatment that will be encountered in the
development of methods for detecting NP in solid
food matrices.
The literature survey on the detection and characterisation
of NM in food shows that, up to now, a very
limited number experimental studies have actually
been reported and that these publications each adopt
different experimental approaches and techniques
applied to a large variety of materials in different
food matrices. Almost all the published literature
related to NP in food has been mostly concerned
with the detection and characterisation of the nanomaterial
and we did not try to address the much more
challenging task of determining the number size
distribution of the primary particles, including those
present in the form of agglomerates/aggregates. The
few examples that successfully tackled the measurement
of the particle size distribution did so using
simplified systems such as mixtures of gold nanoparticles
of different sizes in water by using FFF for
separation and light-scattering and electron microscopy
for size measurement (Calzolai et al. 2011;
Schmidt et al. 2011). Using a similar analytical
platform, it was possible to determine the partial
particle size distribution of titanium dioxide in sunscreen
lotions (Contado and Pagnoni 2008; Samontha
et al. 2011).
One recurring theme in the available literature is
the need for extensive, carefully executed and documented
sample preparation. In fact, irrespective of the
analytical technique used, a proper measurement of
Food Additives and Contaminants 1189
nanoparticle size usually requires a simplification of
the matrix into which nanoparticles are embedded.
To this end two approaches can be used: either
extracting the NPs from the embedding matrix or
removing (or at least simplifying) the complex matrix.
If any sample preparation treatment is used, it will be
necessary to control that the treatment method does
not modify the original particle size distribution.
Another challenge is posed by the need to distinguish
between engineered nanoparticles that have been
added on purpose to the food material and naturally
occurring nanoparticles. Different types of food naturally
contain particles below the 100 nm size, e.g. milk
contains lactoglobulin proteins, the properties of which
clearly fit the definition of nanomaterials. In most
cases the techniques used to measure the size of
particles cannot provide unambiguous data about
their chemical identity. In this case some other techniques
will be needed to discriminate between engineered
nanoparticles and naturally occurring
nanoparticles.
Measuring particle size distribution in complex
matrices
The available, peer-reviewed, experimental work
described above was mainly attempting to detect and
characterise NP in complex matrices. The task at hand
to fulfil the requirement of the European Union
definition of what constitutes a nanomaterial is much
more challenging than that: it effectively requires the
measurement of the size distribution in terms of the
number of primary particles, including those present in
agglomerates, across the whole particle size range from
1 nm upwards. Given these stringent requirements
there are few techniques apart from electron microscopy
that can directly size and count particles and none
which can effectively deal with counting primary
particles within agglomerates. If the simplified situation
of a sample free from agglomerates is considered,
then one possible technique for direct sizing/counting
could be single-particle detection using ICP-AES and
ICP-MS. These techniques have been used for direct
analysis of individual nanoparticles (Suzuki et al. 2011)
and cells. Very recently single-particle ICP-MS has
been used to measure the number concentration of
silver nanoparticles of different sizes (Laborda et al.
2011).
The overall picture that emerges from reviewing the
available techniques for measuring particle size distribution
is that, most probably, simple techniques will
not work.
Further challenges for full method development
It is likely that many sample types will not be suitable
for instrumental analysis without some degree of pretreatment.
While varying degrees of pre-treatment
could be envisioned, it is likely that as a minimum
most samples will be subjected to varying degrees of
ultrasonic treatment to homogenise the sample and to
ensure the breaking up of any loosely agglomerated
material. In principle this should be quite simple, but
experimental studies show that the re-dispersion of
even simple powders of nanomaterials can give very
different results depending on the treatment used. This
can be illustrated in the examples in Figure 3, which
shows the results obtained from re-dispersion of nanoZnO
powders using centrifugal particle sedimentation.
When nano-ZnO is added to pure water and treated
Figure 3. Size distribution obtained by CPS analysis of a ZnO sample dissolved in water and treated with ultrasonic probe for
varying time periods of 20 s (right curve) or 30 s (left curve).
1190 L. Calzolai et al.
with an ultrasonic probe for only 20 min, the CPS data
shows the sample as being effectively non-nano, while
a relatively minor increase in ultrasonic treatment to
30 min shows a clear shift of the size distribution to
smaller values with easily detectable sub-100 nm
material which, on a number basis, may well require
this material to be reclassified as being nano.
In another example shown in Figure 4, CeO2
nanopowder was re-dispersed in water by firstly
vortexing and then using ultrasonic treatment for
increasing amounts of time. In this case it can be seen
that while simple vortexing can produce a nanodispersion,
it is also clear that brief short and
medium (10 min and 10 þ 20 min) treatments can
further reduce the mean particle size. On the contrary
longer ultrasonic treatment (10 þ 20 þ 40 min) actually
begins to produce deleterious results with a clear
increase in the apparent mean particle size, larger even
that after simple vortexing. This behaviour is likely due
to the onset of particle fusion caused by an excessively
high total input of ultrasonic energy. While this effect
is not unknown, it serves in this case to illustrate that
even an apparently simple single-step process, ultrasonic
homogenisation, must be carefully evaluated and
optimised to ensure that it is assisting the measurement
process and not altering the results to the point of
producing a false-negative result.
These two examples, although very simple experimentally,
serve to illustrate that in addition to the
challenges of developing the instrumental measuring
methods, much effort will be required to define
material pre-treatment procedures without artificially
distorting the original size distribution of the particles.
Outlook
The ideal fit-for-purpose analytical method should
have the following characteristics:
. Be sensitive enough to measure a low concentration
and be able to analyse the size distribution
and properties of NP in addition to
concentration.
. Minimise sample disturbance to ensure that
laboratory analyses reflect the unperturbed
environmental state.
. Be non-destructive.
. Be able to analyse samples of diverse elemental
compositions and samples containing more
than one type of nanoparticle.
. Be able to provide a wide size-separation
range.
A possible solution towards an analytical platform
able to measure the number particle size distribution in
complex mixtures or matrices is the use of a multi-step
approach (Tiede, Hassello¨ v, et al. 2009) where different
building blocks are used in a sequential order. These
building blocks would be: size separation, followed by
size measurement, and particle quantification and
identification (Figure 5). For the size separation step
Figure 4. CPS analysis of a CeO2 sample re-dissolved using simple vortexing (blue line) or a combination of vortexing and
ultrasonic treatment for 10 s (yellow line), 30 s (pink line) and 70 s (cyan line).
Figure 5. Building blocks for an analytical platform to measure the number size distribution of nanoparticles.
Food Additives and Contaminants 1191
different techniques able to separate particles based on
size could be used, such as: size exclusion chromatography,
hydrodynamic chromatography, field flow fractionation,
electrophoresis and ion mobility techniques.
The measurement of size on the nanometer scale could
be performed by either dynamic light scattering or static
light scattering. Finally for the particle identification
some possible techniques that have been already used
for the identification of NP are: ion-coupled plasma
mass spectrometry, atomic emission spectroscopy and
electro-spray ionisation mass spectrometry.
The combination of some of these techniques has
already been tested and has given quite promising
results. For example, by combining FFF with DLS we
were able to separate and correctly quantify the
relative number of particles of a mixture of gold
nanoparticles of 5, 15 and 45 nm (Calzolai et al. 2011).
Directly coupling FFF with MALS detection and ICPMS
Schmidt et al. (2009) measured the migration of
nanoclay nanoparticles from a biopolymer composite
to the ethanol liquid phase used as food simulant.
An important factor in accelerating the development
of methods for detecting specific nanomaterials
in food (and other complex matrices) would definitely
be the availability of standard reference materials,
ideally matrix based. In addition, the whole process
would be helped by the availability of databases with
information about the chemical composition, main
characteristics and uses of nanomaterials present in
consumer products.
Conclusions
All the instrumental techniques providing size distributions
have advantages and drawbacks. At the
moment there is no single technique that can by itself
provide a robust analytical method, especially considering
the need to measure the number size distribution
of nanoparticles introduced by the definition of
nanomaterials. The most likely solution to this problem
will be to use combinations of instruments, each
with a different physical principle of operation, and in
this way complement the weaknesses of each instrument
with the strengths of another.
Finally, the association of separation and size
measurement, and particle identification analysis is
particularly promising, by alleviating the difficulties
inherent to size distribution measurement alone. In any
case, one should not underestimate the measurement of
nanoparticle size particularly when complex media are
involved and requiring appropriate sample preparation,
i.e. separation of nanomaterials from the matrix.
Acknowledgements
The authors thank Dr Cesar Pascual Garcia of the JRC for
providing Figure 1.
References
Berne BJ, Pecora R. 2000. Dynamic light-scattering: with
applications to chemistry, biology, and physics. Mineola,
NY: Dover.
Blasco C, Pico` Y. 2011. Determining nanomaterials in food.
TrAC – Trend Analyt Chem. 30(1):84–99.
Borm PJ, Robbins D, Haubold S, Kuhlbusch T, Fissan H,
Donaldson K, Schins R, Stone V, Kreyling W,
Lademann J, et al. 2006. The potential risks of nanomaterials:
a review carried out for ECETOC. Part Fibre
Toxicol. 3:11.
Brar SK, Verma M. 2011. Measurement of nanoparticles by
light-scattering techniques. TrAC – Trend Analyt Chem.
30(1):4–17.
Calzolai L, Gilliland D, Garcia CP, Rossi F. 2011.
Separation and characterization of gold nanoparticle
mixtures by flow-field-flow fractionation. J Chromatogr
A. 1218(27):4234–4239.
Carney RP, Kim JY, Qian H, Jin R, Mehenni H, Stellacci F,
Bakr OM. 2011. Determination of nanoparticle size
distribution together with density or molecular weight by
2D analytical ultracentrifugation. Nat Commun. 2:335.
Chaudhry Q, Scotter M, Blackburn J, Ross B, Boxall A,
Castle L, Aitken R, Watkins R. 2008. Applications and
implications of nanotechnologies for the food sector. Food
Addit Contam A. 25(3):241–258.
Coelfen H. 2004. Feature article: analytical ultracentrifugation
of nanoparticles. Polym News. 29(4):101–116.
Contado C, Pagnoni A. 2008. TiO2 in commercial sunscreen
lotion: flow field-flow fractionation and ICP-AES together
for size analysis. Analyt Chem. 80(19):7594–7608.
Cushen M, Kerry J, Morris M, Cruz-Romero M,
Cummins E. 2011. Nanotechnologies in the food industry.
Recent developments, risks and regulation. Trends Food
Sci Amp Tech. 24(1):30–60.
Dudkiewicz A, Tiede K, Lo¨ schner K, Jensen LHS, Jensen E,
Wierzbicki R, Boxall ABA, Mølhave K. 2011.
Characterization of nanomaterials in food by electron
microscopy. TrAC – Trend Analyt Chem. 30(1):28–43.
European Commission. 2011. Off J Eur Comm. L 275:38–40.
European Parliament and European Council. 2009.
Regulation (EC) No. 1223/2009 of the European
Parliament and of the Council of 30 November on the
provision of food information to consumers, amending
Regulations (EC) No 1924/2006 and (EC) No 1925/2006
of the European Parliament and of the Council, and
repealing Commission Directive 87/250/EEC, Council
Directive 90/496/EEC, Commission Directive 1999/10/
EC, Directive 2000/13/EC of the European Parliament
and of the Council, Commission Directives 2002/67/EC
and 2008/5/EC and Commission Regulation (EC) No 608/
2004. Off J Eur Union. L304:18–209.
European Parliament and European Council. 2011.
Regulation (EU) No. 1169/2011 of the European
Parliament and of the Council of 25 October 2011 on
cosmetic products. Off J Eur Union. L342:59–63.
Gallego-Urrea JA, Tuoriniemi J, Hassell M. 2011.
Applications of particle-tracking analysis to the determination
of size distributions and concentrations of nanoparticles
in environmental, biological and food samples.
TrAC – Trend Analyt Chem. 30(3):473–483.
1192 L. Calzolai et al.
Giddings JC, Yang FJ, Myers MN. 1976. Flow-field-flow
fractionation: a versatile new separation method. Science.
193(4259):1244–1245.
Gun’ko VM, Klyueva AV, Levchuk YN, Leboda R. 2003.
Photon correlation spectroscopy investigations of proteins.
Adv Colloid Interface Sci. 105:201–328.
Hupfeld S, Ausbacher D, Brandl M. 2009. Asymmetric flow
field-flow fractionation of liposomes: optimization of
fractionation variables. J Separat Sci. 32(9):1465–1470.
Innovative Research and Products Inc. 2009. Nano-enabled
packaging for the food and beverage industry – A global
technology, industry and market analysis. Available online
from http://www.innoresearch.net/ReportContent/?id=
68&pg=107&pd=7/1/2009.
Kamack H. 1951. Particle-size determination by centrifugal
pipet sedimentation. Analyt Chem. 23(6):844–850.
Kaszuba M, Connah MT, McNeil-Watson FK,
Nobbmann U. 2007. Resolving concentrated particle size
mixtures using dynamic light-scattering. Particle Particle
Systems Characterization. 243:159–162.
Laborda F, Jimenez-Lamana J, Bolea E, Castillo JR. 2011.
Selective identification, characterization and determination
of dissolved silver(i) and silver nanoparticles based on
single particle detection by inductively coupled plasma
mass spectrometry. J Analyt Atomic Spectrom.
26(7):1362–1371.
Laera S, Ceccone G, Rossi F, Gilliland D, Hussain R,
Siligardi G, Calzolai L. 2011. Measuring protein structure
and stability of protein–nanoparticle systems with
synchrotron radiation circular dichroism. Nano Lett.
11(10):4480–4484.
Loevestam G, Rauscher H, Roebben G, Sokull-Kluettgen B,
Gibson P, Putaud J-P, Stamm H. 2011. Considerations on
a definition of nanomaterial for regulatory purposes. JRC
Reference Report, EUR. 24403:202010-1.
Lorenz C, Tiede K, Tear S, Boxall A, von Goetz N,
Hungerbu¨ hler K. 2010. Imaging and characterization of
engineered nanoparticles in sunscreens by electron microscopy,
under wet and dry conditions. Int J Occup Environ
Hlth. 16(4):406–428.
Luykx DM, Peters RJ, van Ruth SM, Bouwmeester H. 2008.
A review of analytical methods for the identification and
characterization of nano delivery systems in food. J Agric
Food Chem. 56(18):8231–8247.
Mitrano DM, Lesher EK, Bednar A, Monserud J, Higgins
CP, Ranville JF. 2012. Detecting nanoparticulate silver
using single-particle inductively coupled plasma-mass
spectrometry. Environ Toxicol Chem. 31(1):115–121.
Peters R, ten Dam G, Bouwmeester H, Helsper H,
Allmaier G, vd Kammer F, Ramsch R, Solans C,
Tomaniova´ M, Hajslova J, et al. 2011. Identification
and characterization of organic nanoparticles in food.
TrAC – Trend Analyt Chem. 30(1):100–112.
Saeseaw S, Shiowatana J, Siripinyanond A. 2005.
Sedimentation field-flow fractionation: Size characterization
of food materials. Food Res Int. 38(7):777–786.
Samontha A, Shiowatana J, Siripinyanond A. 2011. Particle
size characterization of titanium dioxide in sunscreen
products using sedimentation field-flow fractionationinductively
coupled plasma-mass spectrometry. Analyt
Bioanalyt Chem. 399(2):973–978.
Schmidt B, Lo¨ schner K, Hadrup N, Mortensen A, Sloth JJ,
Bender Koch C, Larsen EH. 2011. Quantitative characterization
of gold nanoparticles by field-flow fractionation
coupled online with light-scattering detection and
inductively coupled plasma mass spectrometry. Analyt
Chem. 83(7):2461–2468.
Schmidt B, Petersen JH, Bender Koch C, Plackett D,
Johansen NR, Katiyar V, Larsen EH. 2009. Combining
asymmetrical flow field-flow fractionation with lightscattering
and inductively coupled plasma mass spectrometric
detection for characterization of nanoclay used in
biopolymer nanocomposites. Food Addit Contam A.
26(12):1619–1627.
Suzuki Y, Sato H, Hikida S, Nishiguchi K, Furuta N. 2011.
Real-time monitoring and determination of Pb in a single
airborne nanoparticle. J Analyt Atom Spectrom.
25(7):947–949.
Tiede K, Boxall ABA, Tear SP, Lewis J, David H,
Hassello¨ v M. 2008. Detection and characterization of
engineered nanoparticles in food and the environment.
Food Addit Contam A. 25(7):795–821.
Tiede K, Hassello¨ v M, Breitbarth E, Chaudhry Q,
Boxall AB. 2009. Considerations for environmental fate
and ecotoxicity testing to support environmental risk
assessments for engineered nanoparticles. J Chromatogr
A. 1216(3):503–509.
Tiede K, Tear SP, David H, Boxall AB. 2009. Imaging of
engineered nanoparticles and their aggregates under fully
liquid conditions in environmental matrices. Water Res.
43(13):3335–3343.
von der Kammer F, Legros S, Hofmann T, Larsen EH,
Loeschner K. 2011. Separation and characterization of
nanoparticles in complex food and environmental samples
by field-flow fractionation. TrAC – Trend Analyt Chem.
30(3):425–436.
Food Additives and Contaminants 1193
Copyright of Food Additives & Contaminants. Part A: Chemistry, Analysis, Control, Exposure & Risk
Assessment is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple
sites or posted to a listserv without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.