Nanoparticle Characterization What To Measure

15214095, 2019, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.201901556 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/10/2022].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
REVIEW
Nanoparticle Characterization www.advmat.de
Nanoparticle Characterization: What to Measure?

Mario M. Modena,* Bastian Rühle, Thomas P. Burg, and Stefan Wuttke*
“Measure that which is measurable and make measurable that which is not”—Galileo Galilei
1. Introduction
What to measure? is a key question in nanoscience, and it is not
straightforward to address as different physicochemical properties Nanomaterials are defined as materials
that consist of nanoparticles of which
define a nanoparticle sample. Most prominent among these properties
at least 50% have one or more external
are size, shape, surface charge, and porosity. Today researchers have an dimensions between 1 and 100 nm.[1] Their
unprecedented variety of measurement techniques at their disposal to small dimensions do not only allow more
assign precise numerical values to those parameters. However, methods surface functionality in a given volume,
based on different physical principles probe different aspects, not only but also lead to physical properties that
of the particles themselves, but also of their preparation history and often differ from their bulk counterparts
in many aspects, including electronic,
their environment at the time of measurement. Understanding these
optical, and magnetic features.[2–7] Nano-
connections can be of great value for interpreting characterization results particles possess a much higher surface-
and ultimately controlling the nanoparticle structure–function relationship. to-mass ratio than bulk materials and,
Here, the current techniques that enable the precise measurement of these therefore, surface atoms and surface
fundamental nanoparticle properties are presented and their practical energy strongly contribute to the material
properties, e.g., leading to reduced lattice
advantages and disadvantages are discussed. Some recommendations
constants and lower melting points.[8–10]
of how the physicochemical parameters of nanoparticles should be Moreover, the high number of surface
investigated and how to fully characterize these properties in different atoms and the high surface energy of
environments according to the intended nanoparticle use are proposed. nanoparticles can have a strong impact on
The intention is to improve comparability of nanoparticle properties and the catalytic performance. Thus, catalyti-
performance to ensure the successful transfer of scientific knowledge to cally inactive bulk materials can become
very active catalysts when produced as
industrial real-world applications.
nanoparticles with high surface areas.[11]
If fewer atoms comprise a solid, a lower
number of orbitals contribute to the band
Dr. M. M. Modena formation. This effect leads to changes in the band structure,
ETH Zurich such as band gap variations in semiconductors, which depend
Department of Biosystems Science and Engineering
Mattenstrasse 26, 4058 Basel, BS, Switzerland on the nanomaterial dimensions.[12,13] These unique properties
E-mail: [email protected] render nanoparticles extremely attractive for a large range of
Dr. B. Rühle applications, including catalysis, gas and energy storage, photo-
Federal Institute for Materials Research and Testing (BAM) voltaic, electrical and optical devices, and biological and medical
Richard-Willstätter - Str 11, 12489 Berlin, Germany technologies.[14–23] For this reason, nanoparticles are not only a
Prof. T. P. Burg growing topic of interest in research settings, but they are also
Max Planck Institute for Biophysical Chemistry
already widely used in consumer products.[24]
Am Fassberg 11, 37077 Göttingen, Germany
Currently, a key issue hindering the utility of nanoparticles
Prof. T. P. Burg
Department of Electrical Engineering and Information Technology in industry is reproducibility. This problem is, however, partially
Technische Universität Darmstadt intrinsic, as the product of synthesis is always prone to yield a
Merckstrasse 25, 64283 Darmstadt, Germany polydispersion of nanoparticles, sometimes with a broad dis-
Prof. S. Wuttke tribution of sizes, shapes, and defects. Nanoparticle characteri-
Department of Chemistry zation is therefore a crucial step required to fully comprehend
Center for NanoScience (CeNS)
University of Munich (LMU) the origin of nanoparticle behavior, and subsequently translate
81377 Munich, Germany their performance benefits from laboratories into specific real-
E-mail: [email protected] word applications.
Prof. S. Wuttke Determining the physicochemical properties of nanoparti-
BCMaterials cles and exploring their structure–function relationships is a
Basque Center for Materials
critical challenge for scientists today. This endeavor is limited
UPV/EHU Science Park, 48940 Leioa, Spain
by our ability to fully investigate the nanoscale realm: Different
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adma.201901556. characterization techniques are based on different physical
properties, therefore only providing a partial picture of the
DOI: 10.1002/adma.201901556 nanoparticle characteristics. Making matters more challenging
Adv. Mater. 2019, 31, 1901556 1901556 (1 of 26) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
15214095, 2019, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.201901556 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de
yet, the characterization methods themselves can directly affect

the measured quantities.[2,25–31] Mario M. Modena is a senior
In this critical review we aim at providing a set of guidelines postdoc in the group of
to investigate and characterize the key parameters defining a Prof. Hierlemann, ETH Zurich
nanoparticle sample, namely size, shape, surface charge, and (Switzerland). He obtained a
porosity. We will first define these physicochemical terms and Ph.D. in physics with
their implication in affecting nanoparticle properties. We will Prof. Burg at the Max Planck
then provide a critical overview of established and specialized Institute for Biophysical
techniques currently used for evaluation of nanoparticles, and Chemistry, Goettingen
discuss their practical advantages and disadvantages. Finally, (Germany), where he
we will propose some reasonable recommendations of how the worked on the development
physicochemical parameters of nanoparticles should be investi- of microfluidic-based and
gated, and how to characterize these key properties in different label-free methods for the
environments according to the intended nanoparticle use. characterization of biomolecular complexes and nanopar-
ticles in solution. Currently, his research focus is centered
on the design of microfluidic platforms for the realization
of advanced microphysiological models for application in
2. Nanoparticle Parameters
biology and biomedical engineering.
Nanoparticles exist in various chemical compositions ranging
from micelles to metal(oxide)s, from synthetic polymers to
large biomolecules. Each of these materials features a com- Stefan Wuttke is leader of
pletely different chemistry, which can be analyzed by a variety of the research group “wutt-
methods including optical spectroscopy, X-ray fluorescence and kegroup for science” since
absorbance, Raman spectroscopy, and solid-state NMR.[32] How- 2011, hosted at the Institute
ever, often the behavior of nanoparticles is largely governed by of Physical Chemistry at
their nanometer dimensions. As such, throughout nanoparticle the University of Munich
characterization, the investigation of size, shape, surface charge (LMU), at the Center for
and porosity is a fundamental step for fully understanding and Nanoscience (CeNS) at
predicting their behavior. These essential parameters are the the LMU (Germany) and
focus of our review. at BCMaterials (Spain).
The ability to be dispersed into discrete entities, each Currently, he is an Ikerbasque
characterized by a size, shape, surface charge and porosity Research Professor at
discriminate nanoparticles from the wider class of nanoma- the Basque Centre for Materials, Applications and
terials encompassing nanostructured objects that might pre- Nanostructures (BCMaterials) in Bilbao. His principal
sent dimensions in the micro and millimeter regime, such as focus is on the design, synthesis, and functionalization of
nanostructured films or nanotubes.[33–35] Size and shape affect metal-organic frameworks and their nanometric counter-
the nanoparticle functionalization capacity, fluid drag and dif- parts to target biomedical applications. At the same time,
fusion, optical properties, and uptake into cells.[36] Surface he intends to establish the basic understanding of the
charge, besides controlling the stability of a colloidal suspen- chemical and physical elementary processes involved in
sion and its tendency toward aggregation,[37] also plays a major the synthesis, functionalization, and application of these
role in shaping the interactions between nanoparticles and the hybrid materials.
environment.[38,39] Finally, owing to their increased surface-to-
volume ratio, nanoparticles possess a large external surface
area that can be functionalized for different applications.[17–19] are needed to fully define the actual extension of an object
In addition, porous or hollow nanoparticles also exhibit a vast in space (Figure 1A). While measuring size might appear
internal surface area, which can be further functionalized to trivial for objects in the macroscopic scale (>1 mm) where
impart additional functionalities, such as the design of smart size is measured as the distance between different ends of
nanoparticle drug delivery systems.[18,40–44] an object, in the nanoregime size assumes different mean-
In the following, we provide a description of these physico- ings according to the technique employed to measure it.
chemical terms in the context of nanoparticle technology, and For a nanoparticle, size can refer to i) its overall physical
the different physical and experimental means in which these dimension(s) defined by the atomic structure; ii) an effective
parameters can be defined. size of the particle in a certain matrix according to its diffu-
sion/sedimentation behavior, possibly including adsorption
of matrix constituents to the nanoparticle surface, agglomera-
2.1. Size and Size Distribution tion or aggregation of the particles in the matrix; iii) an effec-
tive size of the nanoparticle, weighted by its mass/electron
Size refers to the spatial extent of an object. For a spherical distribution. This variety of size definitions reflects the wide
object, size can be unambiguously described by one dimen- spectrum of physical approaches that can be used for nano-
sion. However, for nonspherical objects, several dimensions particle characterization.
sedimentation-based methods are commonly employed for the

routine analysis of colloidal suspensions. Usually, no direct
information on nanoparticle shape can be obtained from these
approaches, and an equivalent diameter, corresponding to that
of a sphere behaving the same way as the sample under exami-
nation (Figure 1B) is usually returned as a characteristic size.
To translate this information into actual nanoparticle dimen-
sions, knowledge of nanoparticle shape is required.[29] Finally,
static scattering methods, either via light or X-ray illumination,
provide information on the mass/electron distribution of the
nanoparticles, and hence, indirectly, on the nanoparticle shape.
Although the aim of nanoparticle synthesis is to obtain a
monodisperse population of nanoparticles, a real-world sample
always displays a certain degree of variation. The nanoparticle
size distribution is, therefore, an intrinsic measure of the con-
trol and quality of the synthesis procedure, while the estimated
size value only refers to an averaged quantity derived from this
distribution. Different techniques present different sensitivity to
particles of varying dimensions, and care has to be taken when
comparing results obtained from different characterization
methods.[26,45] Techniques with single-particle resolution, such
as high-resolution microscopy and nanoparticle tracking anal-
ysis, are able to provide a number-based average of particle size,
where each particle is assigned an equal weight. Conversely,
scattering intensities scale with particle volume, and size esti-
mation using these methods is based on a volume-weighted
(static light scattering) or intensity-weighted (dynamic light
scattering) average, which results in a bias toward larger popu-
lation components (Figure 1C). Although conversions between
the types of distributions can be calculated if the weighing fac-
tors are known, the conversion might not have a unique solu-
tion if the shape of the population distribution is unknown.
Therefore, size estimation is the result of the detection
method and of the underlying weighing factors, and of the
average quantity that is chosen as representative of a given nan-
Figure 1. Nanoparticle size characterization. a) Spherical particles are oparticle population. Different averages can be selected, such
described by a single size parameter. However, for nonspherical nano- as arithmetic mean, median value or mode of the distribution.
particles, several dimensions are needed to fully report their dimensions; Such a choice should always be stated alongside the reported
b) the calculation of the effective radius is based on nanoparticle behavior size estimation. The median, for example, is usually defined as
or on the method of detection. The definition of this quantity might differ D50, which indicates the nanoparticle diameter at which 50%
considerably from the physical nanoparticle dimensions; c) different
mean sizes can be calculated for a nanoparticle population according to
of the population lies below. One single quantity, however, is
the weighting factors assigned to the population components. not sufficient to describe a whole population because similar
average values might correspond to largely different distribu-
tions. To provide more information, different metrics are com-
Nanoparticle size and morphology can be measured at the monly used, with the most common ones being the D10 and
single-particle level at sub-nanometer resolution using high- D90, corresponding to the value of diameter which encompasses
resolution microscopy techniques, such as electron micro 10% and 90% of the population, respectively, or the standard
scopy or scanning probe microscopy, providing extremely deviation, when the population follows a normal distribution.
detailed information on the shape of the nanoparticles under Derived quantities can then be calculated from these metrics,
examination. These methods of characterization are based such as the coefficient of variation (COV), defined as the ratio
on the interaction between the atomic structure and the of standard deviation and arithmetic mean, used for static
impinging electron beam, or with the used scanning probe. light scattering measurements and defined by the ISO 13320
However, these techniques usually have limited throughput, standard.[46]
are not ensemble techniques (which raises questions such as
how representative and statistically relevant the obtained data
are), and are carried out under high-vacuum conditions or 2.2. Shape
by placing the sample on a hard substrate, and can therefore
neglect important phenomena that may occur in suspension, Although often assumed spherical, nanoparticles feature a
such as swelling or aggregation. Light-scattering, diffusion- and large variety of geometric and irregular shapes.[47] Particles
Figure 3. Surface charge and zeta potential of nanoparticles in suspen-

sion. Charges on the nanoparticle surface are screened by the free ions in
solution, giving rise to two ion layers: a first layer of adsorbed ions on the
Figure 2. Nanoparticle shape characterization. TEM provides nanometer nanoparticle surface, the so-called Stern layer; a second layer of stationary
resolution on nanoparticle morphology. However, the measured shape is a but diffusing ions that move with the particle. The zeta potential is defined
2D projection of the nanoparticle morphology, and therefore depends on the as the potential difference between the slipping plane, i.e., the plane that
relative orientation of the nanoparticle and the electron beam. Scattering- “separates” the cloud of stationary ions around the particles from freely
based techniques provide qualitative information on the nanoparticle shape. diffusing ions in solution, and the potential of the bulk solution.
with equal composition and similar dimensions might present charged molecules contribute to the appearance of a net charge
drastically different behaviors as a consequence of their shape, on the nanoparticle surface. This charge has a primary effect
such as surface-binding capability, cellular uptake and release, on the behavior of nanoparticles in different environments, in
optical and plasmonic effects,[47–51] to name some of the major particular on controlling their tendency toward aggregation,
properties affected by particle morphology. as electrostatic repulsion between particles is a key factor pro-
Shape is commonly characterized by use of high-resolu- moting the stability of colloidal solutions.[55] In particular, in
tion microscopy techniques, such as electron and scanning an electrolyte solution, mobile charges in solution are attracted
probe microscopy, which enable the detection of particle mor- by the static charges on the nanoparticle surface, effectively
phology with sub-nanometer resolution. These techniques leading to a screening of the electric potential, which can
are routinely employed during nanoparticle synthesis for vali- ultimately result in particle aggregation. A typical measure
dation and characterization of synthesis outcome. However, of surface charge and colloidal stability is given by the zeta
electron microscopy typically provides a 2D projection of the potential ζ, which is defined as the electric-potential differ-
particle shape onto a plane, which, for specific cases of highly ence between the stationary layer of charges surrounding the
anisotropic particles, might lead to erroneous estimates of particles and the solution potential (Figure 3).[56] Suspensions
the particle morphology (Figure 2). To circumvent this limita- that feature |ζ| ≥ 15 mV are usually considered to be colloidally
tion, the characterization of particles presenting distinct 3D stable. Several parameters affect the zeta potential of particles
anisotropy can be carried out by acquiring the projections in solution, namely the ionic strength of the solvent, the pres-
of a large number of randomly oriented identical particles ence of charged or uncharged molecules that can adsorb on
to reconstruct their spatial arrangement,[52] or by electron the particle surface, and the pH of the solution. In particular,
tomography.[52,53] when dealing with particles that have (de)protonatable groups
Information on shape and anisotropy of particles in solu- on their surface, one parameter of interest is represented by
tion can also be obtained by using scattering-based techniques, the pH at the point of zero charge, pHpzc, i.e., the pH value at
which can more readily be applied in solution, such as by which the particles present zero charge on their surface,[57] and
combining static and dynamic light scattering characteriza- hence show a major tendency toward forming large aggregates.
tion.[54] However, the underlying ensemble-based analysis of Finally, surface charge also has major implications on con-
the particles in solution only enable to infer, quantitatively, an trolling the interactions between nanoparticles and biological
anisotropy factor of the particles, and a detailed study of particle fluids and samples. The formation of a protein corona on the
morphology remains limited to high-resolution microscopy. surface of the nanoparticles and the probability of nanoparticle
Nevertheless, the qualitative shape information obtained by uptake by different types of human cells and tissues largely
scattering-based characterization methods are often necessary depend on the nanoparticle surface charge.[38,58] For these rea-
to confirm microscopy results, as sample preparation and elec- sons, the investigation of surface charge is a fundamental step
tron-microscopy evaluation might affect the sample agglomera- in the formulation of nanoparticle-based therapeutics.[59]
tion state or induce damages to the particle framework.
2.4. Porosity
2.3. Surface Charge
The possibility of synthesizing nanoparticles featuring porous
The boundary between the solid and the fluid phase is a frameworks has greatly expanded the range of application of
dynamic environment, and multiple phenomena, such as the nanomaterials.[60,61] Porosity provides the nanoparticles with
presence of dangling bonds, or the adsorption or grafting of a drastic increase in their surface-to-volume ratio, which can
exceed that of solid particles with equal dimensions by sev- the crystal lattice can give rise to porosity, and when the crystal
eral orders of magnitude. Over the last two decades, porous structure is known, the pore sizes and pore openings can be
nanoparticles have attracted considerable attention in the calculated. Moreover, the porosity of crystalline and noncrystal-
pharmaceutical and medical fields for delivery and targeting line nanomaterials can also be measured with different tech-
of therapeutics[62–65] and for disease diagnostics.[21,66] These niques such as gas sorption. Although these approaches can
functionalized nanoparticles have to fulfill diverse tasks during provide fundamental insights into the pore structure of dry par-
the delivery process, e.g., drug loading, nanoparticle sealing, ticles, they fail to predict possible variations that could occur in
targeting and cell uptake, endosomal escape, and controlled solution, such as selective permeation of solvent components
and triggered release of the drug.[20] For enabling such a wide within the porous framework. In our discussion, we will also
spectrum of functions, a chemistry has to be developed for the present some specialized techniques that can overcome such
integration of sophisticated inner and outer surface function- limitations in characterization and offer new information on
alizations:[18] inner pore functionalization for controlled host- porous nanoparticles in solution.
guest interactions,[67] and peripheral (outer) functionalization
for targeting purposes and drug release at the intended target
site.[43,68] The “orthogonality” of the inner and outer surface 3. Characterization Methods
chemistry is key in providing such capability to nanocarriers, as
it allows to differentiate between the targeting moieties or gate- Several characterization methods have been devised to inves-
keepers grafted on the nanoparticle surface and the functional tigate size, distribution, shape, surface charge and porosity of
groups attached to the inner surface. Besides “orthogonal” nanoparticles in different environments. Here, we discuss the
functional groups or derivatization reactions, steric hindrance main techniques for the characterization of these key parameters
or electrostatic repulsion can also be used to control the site both in the dry state and in solution. We also introduce some
selectivity of the functionalization and the type of molecules specialized techniques for nanoparticle characterization, which
that can enter the pores, and hence provide controlled reaction enable to expand the accessible range of information to gain
environments. deeper insights into specific nanoparticle properties. Given the
To enable the development and characterization of porous large number of existing approaches and techniques, including
nanoparticles, porosity needs to be investigated at different the combination of different methods in “hyphenated”
levels (Figure 4), namely, i) the size of the pore opening; ii) the techniques, different variations of the same techniques, and
dimensions and volume of the porous cavity; iii) the intercon- different approaches to data analysis for a same technique, this
nection of the porous structure, i.e., whether only superficial article cannot provide an exhaustive list of all available methods
pores are accessible from the outside or the whole internal for nanoparticle characterization. We rather provide a selection
porous network; iv) specific surface area (sum of inner and of methods that in our opinion are best suited to characterize a
external surface); v) surface-to-volume ratio; vi) the inner and broad range of nanomaterials, which are commonly used and
outer surface functionalization. Despite the large interest in this well established.
class of nanomaterials, there is a general lack of standardized
methods to investigate at depth all these key features of porous
nanoparticles. In the case of a crystalline material, the periodic 3.1. Characterization in Dry State
arrangement of individual building blocks of a nanomaterial in
The methods discussed in this section are summarized in
Table 1.
3.1.1. Transmission Electron Microscopy (TEM)
Transmission electron microscopy is undoubtedly one of the

most important nanoparticle characterization techniques.
TEM employs a focused electron beam on a thin (typically less
than 200 nm) sample to produce micrographs of nanoscale
materials with high lateral spatial resolution[69,70] (Figure 5a).
Current electron microscopes can achieve resolutions down to
0.05–0.1 nm by reducing image distortion by aberration correc-
tors, hence providing high-resolution images with atomic reso-
lution.[71,72] TEM also enables studying the crystalline structure
of selected microscopic regions of crystalline materials by spa-
tially confining and focusing the impinging beam and detecting
the resulting electron diffraction pattern.[73] Thanks to this high
spatial resolution and selectivity, TEM enables the investiga-
Figure 4. Porous nanoparticle. Porous nanoparticles feature different tion of size, shape, and crystal structure at the single-particle
characteristics of interest which need to be characterized, namely, pore level. Once a representative group of images of the nanopar-
aperture, pore volume, and external and internal surface functionalization. ticle sample is acquired, the individual size of ≈1000 randomly
Table 1. Summary of characterization methods for dry nanoparticles.
Technique NPa) state Parameters Advantages Limitations Reviews

TEM High vacuum Size (<1 nm to <1 µm) Particle morphology at sub-nm resolution, High energy beams, [78,80]
Size distribution information on internal structure very expensive
(number-based) of the particles
Shape
(2D projection)
SEM High vacuum or Size Single-particle resolution, Limited penetration depth [84,87,88]
low pressure (≈2–3 nm to >10 µm) lower energy beams than TEM,
Size distribution user friendly
(number-based)
Shape
(2D projection)
AFM Dry and in liquid Size High compatibility with different Samples need to be deposited [89]
(≈10–20 nm lateral res., samples and measurement on hard surface,
<1 nm vertical res.) environments limited throughput
Size distribution
(number-based),
Shape
(3D imaging)
XRD Dry Crystallite size Rapid, provides information No information on particle size [99,100]
(≈1 to ≈100 nm) on crystal structure
SAXS Dry and in suspension Size: radius of gyration High sensitivity, compatible Previous knowledge of particle [102]
(≈1 nm to ≈1 µm) for both dry particles morphology is required
Size distribution and in suspension for fitting the data
(intensity-based)
Shape
(through modeling)
MS Dry and in suspension Size Information on elemental Sample ionization [105,106]
(≈0.1 to ≈50 nm) composition (might affect particle stability)
Size distribution
(number-based)
Composition
Gas sorption Dry Surface area Compatible with polydisperse Requires sample degassing, [121,122]
Pore volume and aggregated samples no information on particle
Pore size distribution morphology
(≈1 to ≈50 nm)
a)NP stands for nanoparticle.
selected nanoparticles should be measured to obtain mean- 3.1.2. Scanning Electron Microscopy (SEM)
ingful statistics for size distribution determination. This pro-
cedure can be done either manually (inherently affected by Scanning electron microscope enables imaging the sample
human errors, bias, and subjectivity) or using automated par- surface by detecting secondary electrons emitted from the
ticle analysis methods.[74–77] sample upon interaction with the impinging electron beam[83]
Although TEM enables visual inspection of single particles (Figure 6a). In SEM, lower beam energies are utilized for
with nanometer resolution, the whole workflow of sample sample imaging as compared to TEM characterization, which
preparation, measurement, and analysis can be extremely labor results in a limited penetration depth of the beam and, hence,
intensive.[78,79] In addition, the nanoparticles have to be elec- in being sensitive solely to the specimen surface. However,
tron transparent and able to withstand the high vacuum and this superficial interaction also implies that SEM characteriza-
beam energy employed during characterization. Especially, due tion can be used for the analysis of the morphology of “thick”
to the high-energy electron beam, sample damage is a known (>100 nm) samples, which is not possible with TEM.[84–86] The
problem for organic, polymer and hybrid nanoparticles[26,80,81] moderate electron energies employed for SEM analysis limit
(Figure 5). This challenge can be addressed by using Cryo- the resolution to typically >2–3 nm, however at the same time
EM (see advanced methods) or by reducing the acceleration drastically decrease the possibility of beam-induced sample
voltage, however at the expense of increasing the complexity of damage compared to TEM. In addition, SEM is by far more
the measurement procedure or reducing the attainable resolu- user-friendly and enables faster measurements, and features
tion.[80,82] Lastly, TEM is a highly costly technique (acquisition lower acquisition and maintenance costs than TEM. SEM
and maintenance costs), and requires highly trained personnel. instruments typically also enable investigating the composition
SEM typically requires conductive substrates for high-resolu-

tion imaging and nonconductive samples can be coated with a
thin (5–10 nanometer) metallic film before being analyzed. This
modification of size and surface structure of nonconductive
nanoparticles due to sample preparation has to be accounted
for when interpreting SEM micrographs. Environmental SEM,
i.e., SEM imaging performed at low pressure instead of high
vacuum, enables imaging nonconductive samples, as surface
ionization can be reduced by interaction with the low-pressure
gas in the measurement chamber.[87,88] However, this method
of detection delivers lower spatial resolution than standard
SEM imaging. Finally, when comparing SEM and TEM images,
it should be kept in mind that SEM only yields information
on the sample surface structure, while TEM interacts with the
whole sample volume, hence providing information on sample
structure (e.g., it can provide information on the layer thick-
nesses of core/shell nanoparticles) (Figure 6b,c).
3.1.3. Atomic Force Microscopy (AFM)
Atomic force microscopy is a scanning probe microscopy tech-

nique that can be used to probe and visualize the surface (and
several other force-related quantities) of nanometer-sized or
even atomic-sized objects.[89–92] A sharp tip at the end of a can-
tilever is rastered across the surface of a sample, and the forces
the cantilever experiences during the measurement as a result
of the interaction of the tip with the sample are recorded with
the help of a laser beam reflected off the tip of the cantilever
onto a photodiode array. Depending on the measurement mode,
this can either be a vertical or lateral deflection of the cantilever,
Figure 5. Electron beam–induced sample modifications observed in TEM. or a change in amplitude, frequency, or phase of an oscillating
The temporal sequence of high-resolution TEM images shows the coa- cantilever. In general, three different modes are used: contact
lescence of two gold nanoparticles caused by the electron beam during mode, noncontact mode, and tapping mode. In contact mode,
long-term acquisition. Reproduced with permission.[70] Copyright 2012, the tip is always in direct contact with the surface of the sample.
American Chemical Society. The repulsion of the tip and the surface atoms of the sample
result in a vertical deflection of the cantilever, depending on the
of the sample surface by measuring the amount of elastically underlying topography. The lateral deflection of the cantilever
backscattered electrons, which depends on the interaction can be used in lateral force microscopy (LFM) and measures
between the focused electrons and the sample material, or by changes in friction which often occur at the boundary of two
detecting x-ray emission due to e-beam ionization.[84,85] different materials. This mode is generally quite robust, but it
Figure 6. TEM versus SEM. a) Simplified schematic of TEM and SEM instruments; b) TEM micrograph of MIL-101(Cr). The crystalline structure of
the particles is clearly visible; c) SEM micrographs of the same particles as in (b), after coating with a thin layer of carbon to avoid sample ionization.
b,c) Reproduced under the terms of the Creative Commons Attribution-NonCommercial license.[217] Copyright 2018, The Authors. Published by Wiley-VCH.
is not noninvasive and can result in a displacement of nano- equation.[94] This equation considers the ideal condition of a
objects on a surface, and wear or even damage of the sample perfectly parallel, infinitely narrow and monochromatic X-ray
or the tip. In noncontact mode, a piezocrystal is used to drive beam incident on a monodisperse powder of cube-shaped crys-
oscillations of the cantilever at or close to its resonance fre- tallites.[95] The equation is
quency. These oscillations occur slightly above the surface
of the sample, and the tip is never in direct contact with the Kλ
Dhkl = (1)
sample surface. Yet, a change in the force between the tip and Bhkl cosθ
the sample surface results in a shift of the amplitude, reso-
nance frequency, and phase of the cantilever oscillations, which where Dhkl is the crystallite size in the direction perpendicular
can be used to draw conclusions about surface properties such to the lattice planes, hkl are the Miller indices of the planes
as surface topography. This mode is usually used in (ultra) high belonging to the peak that is being analyzed, K is a numerical
vacuum, and can achieve very high resolutions, down to the factor commonly referred to as the crystallite-shape factor,[96] λ
atomic level.[93] Tapping mode (or intermittent contact mode) is is the wavelength of the X-rays, Bhkl is the FWHM of the dif-
similar to noncontact mode. An oscillating cantilever is used as fraction peak in radians, and θ is the Bragg angle. Given the
well, but instead of oscillating strictly above the sample surface, strong assumptions regarding the instrumental ideality, the
the tip “taps” on the surface during the oscillations. This mode extracted mean size value always needs to be corrected for
is much less invasive than contact mode and is most commonly instrumental broadening effects. Standard phases with known
used under atmospheric conditions. It can even be used in liq- crystal size and narrow size distribution are commonly used to
uids. Besides the attractive or repulsive forces between the tip calculate the instrumental broadening, or FWHMinstr, whose
and the atoms on the sample surface that change depending on contribution needs to be subtracted from the FWHM used in
the distance between the cantilever and the surface atoms (and the Scherrer equation.
hence with surface topography), various other parameters can Although this equation constitutes a valuable approach to
be investigated with AFM, such as magnetic forces in MFM, estimating the mean crystallite size of a polycrystalline sample,
chemical forces in CFM, or the surface potential in KPM. If the the user must be aware of some important issues. First, when
nano-objects are directly on the surface of an (ideally smooth reflections related to different crystallographic directions are
and flat) substrate, no specific sample preparation is necessary used for extracting a unique value of crystallite size, any infor-
prior to the measurement. Particles in suspension have to be mation regarding the crystal shape is lost as the final size value
deposited on a smooth surface (such as mica or a silicon wafer) is averaged over all the crystallographic directions. Second,
first. Unlike in scanning tunneling microscopy, electric conduc- other structural features of the sample contribute to the peak
tivity is not necessary for a standard topography scan in AFM. broadening besides crystal size, such as lattice strain, defects
The scanned image shows directly the size and morphology of or the nanoparticle size itself,[97–100] resulting in overestimated
the nano-objects. However, it is worth noting that the size in FWHM values and consequent underestimated crystal size.
the z-direction is usually very precise, while the lateral size (and Lastly, but most importantly, the information available by XRD
shape) is a convolution of the nano-objects and the size/shape experiments arises from the structural properties of crystal
of the probe tip. Typically, the scanned area is limited to tens or domains and not necessarily from the entire particle.[101] For
a few hundreds of micrometers, and depending on the meas- this reason, when analyzing particle sizes, the use of different
urement mode and the size of the scan area, image acquisition analysis methods, such as TEM or SEM, is necessary to obtain
can take several minutes. reliable information.
3.1.4. X-Ray Diffraction (XRD) 3.1.5. Small-Angle X-Ray Scattering (SAXS)
X-Ray diffraction is a versatile technique used to investigate Small-angle X-ray scattering is a very versatile method for the
a wide range of structural aspects in crystalline samples. The characterization of nanomaterials.[102–104] The sample is illumi-
attainable information ranges from microscopic features, such nated with X-rays and the scattered irradiation is registered by a
as the arrangement of the crystal components, to macroscopic detector at small angles, usually between 0.1° and 5°. Based on
information, such as the mean shape and size of crystals. This the intensity distribution of the scattered X-ray photons that are
information can be obtained by analyzing the full width at half- passing through the sample, information about particle size,
maximum (FWHM) of the Bragg reflections. Since every Bragg size distribution, morphology, crystallinity, molecular weight,
peak is associated with a unique crystallographic direction, the and agglomeration can be obtained. If the absolute intensity
FWHM is influenced by the number of atoms contributing to that is monitored by the detector is properly calibrated, infor-
the scattering events and this number is directly connected to mation about particle concentration or porosity is also available.
the size of the crystal planes generating the specific reflection. The samples can be solids, powders, composites, or disper-
Different sizes for different crystallographic directions are asso- sions of nanoparticles in a liquid medium. Being an ensemble
ciated to a specific shape (i.e., morphology) of the crystallites, method, SAXS probes a very large number of nano-objects
which can therefore be refined against the XRD pattern. simultaneously, and the measured data gives a statistically rel-
For a defined crystallographic direction, the mean value of evant average over a large portion of the sample. Moreover, only
the crystal size can be estimated by using the Scherrer equation minimal sample preparation is required, the measurement is
(Equation (1)), often erroneously referred to as Debye–Scherrer usually nondestructive, and can often be used under in situ
or in operando conditions to study colloidal nanosystems in bodies. By recording the adsorption and desorption isotherms,
their native state. The recorded SAXS pattern consists of two fundamental data including the surface area, pore size distri-
components, namely the form factor, which contains informa- bution and accessible pore volume of a given material can be
tion about mean structural properties of the nanoparticle (i.e., obtained.[121,122] These different properties are determined by
morphology and size), and the structure factor, which can yield the evaluation of sorption isotherms using different data anal-
information about the positional correlation of the nanoparti- ysis techniques such as the Brunauer–Emmett–Teller (BET)
cles (e.g., if the particles interact with each other and assemble method. Sorption isotherms are subdivided into eight different
into higher order structures). However, for nonidentical and types according to the IUPAC definition.[123] The experimental
nonspherical particles, approximations such as the decoupling data allow the classification of the investigated material as
approximation (assuming that particle size and shape are inde- microporous (pore size below 2 nm), mesoporous (between
pendent of their spatial position) or the local monodisperse 2 and 50 nm), or macroporous (above 50 nm).
approximation (assuming that at short ranges all particles are For nonporous materials, the volume-specific surface area
identical) have to be used in order to separate the recorded (VSSA) can be used to classify it as a nanomaterial.[124–127]
intensity values into a product of form factor and structure VSSA is defined according to Equation (2)
factor. Also, due to the polydispersity and nonuniformity of any
synthesized nanomaterial, some prior knowledge of the sample S
VSSA = = SSA × ρ (2)
is often required to impose certain constraints on the model V
used to describe the data, since the solution might be ambig-
uous if all possible sizes and shapes are allowed.[102] where S is the external surface of the sample, V is the solid
volume, SSA is the specific surface area (surface per mass),
and ρ is the material density. According to the European Com-
3.1.6. Mass Spectrometry (MS) mission (EC) recommendation, a material with a VSSA larger
than 60 m2 cm−3 can be considered a nanomaterial.[128,129]
Mass spectrometry was used originally for the characterization The external surface can be readily measured for nonporous
of nanoparticle composition by revealing the stoichiometry of nanomaterials via gas adsorption measurements and by ana-
their building blocks after digestion and dissolution. With the lyzing the results with the BET method. However, for porous
introduction of soft ionization techniques, such as electrospray nanoparticles the BET surface area is a sum of the external and
ionization (ESI) and matrix-assisted laser desorption ioniza- internal surface. This implies that the external surface has to
tion (MALDI), and methods of separation and detection able to be extracted by using a modified t-plot approach.[125] The basic
analyze samples in the Megadalton range, such as ion-mobility principle of this method is to compare the measured gas adsorp-
spectrometry (IMS), time-of-flight (TOF) analysis, and single- tion isotherm with a reduced form of standard isotherms (the
particle inductively coupled plasma-mass spectrometry (single- t-curves) measured for nonporous solids. For this comparison,
particle ICP-MS), the range of application of MS has been the adsorbed nitrogen volume is plotted against the thickness t
extended to the analysis of intact nanoparticles ranging from of the adsorbed layer as obtained from the standard isotherms
few nanometers to hundreds of nanometers in diameter.[105–108] for the material in question (e.g., silica, alumina, carbon). In
Thanks to the versatility of the analysis methods, MS has been an intermediate partial pressure range (p/p0 = 0.25–0.55),
used to investigate a variety of nanoparticle properties besides this comparison usually results in a linear relationship with
elemental composition.[107–109] Integration of nanoparticle the slope relating to the BET value. If a second linear part can
size separation and MS detectors made it possible to study be observed at higher partial pressure values (p/p0 > 0.6), the
the number concentration and size distribution of nanoparti- slope of this second part can be related to the external surface.
cles of a few tens of nanometers with single particle detection Likewise, if a linear relationship is also measured at lower par-
capability.[110–113] MS is also unique in providing information tial pressure range (p/p0 < 0.3), the slope can be used to extract
on the functionalization layer grafted on the surface of nano- the internal surface area.
particles.[114,115] Furthermore, MS has often been applied in
the investigation of interactions between nanoparticles and
biomolecules after exposure to biological fluids to identify and 3.2. Characterization in Suspension
quantify the composition of complex protein coronas on the
nanoparticle surface.[116–118] Finally, the ability of MS to ionize The methods discussed in this section are summarized in
and investigate soft materials, such as tissues, has rendered Table 2.
this technique invaluable in studying the biodistribution and
uptake of nanoparticles within different organs upon admin-
istration of nanoparticle solutions for toxicology studies.[119,120] 3.2.1. Dynamic Light Scattering (DLS)
Dynamic light scattering estimates the particle size from the

3.1.7. Gas Sorption Brownian diffusion of the particles in solution. The low sample
amount required for analysis, the short acquisition and anal-
Analyzing the adsorption and desorption of gas molecules ysis times (in the range of few minutes in total per sample),
(e.g., nitrogen) under isothermal conditions on solid surfaces the dynamic range spanning from nanometer to ≈10 µm, and
is a widespread technique for the characterization of porous the compatibility of the technique with a large variety of solvent
Table 2. Summary of characterization methods for nanoparticles in suspension.

DLS In suspension Size: hydrodynamic radius Rapid, provides information Highly biased toward larger particles [134]
(≈5 nm to ≈10 µm) on nanoparticle behavior in suspension, no information
Size distribution in solution on particle shape
(intensity-based)
SLS In suspension Size: radius of gyration Molecular weight and radius Highly biased toward larger [29,136]
(≈5 nm to ≈1 µm) of gyration of particles in particles in suspension
Size distribution solution
(intensity-based)
NTA In suspension Size: hydrodynamic radius Single-particle resolution, Requires sample dilution [138]
(≈30 nm to ≈1 µm) suitable for highly and highly scattering particles
Size distribution polydisperse samples
(number-based)
Concentration
Electrophoretic light In suspension Zeta potential Rapid, typically combined Indirect estimation of zeta potential from [139]
scattering with DLS electrophoretic mobility, ensemble-based
AUC In suspension Size: hydrodynamic radius High-sensitivity, compatible High-cost equipment [143]
(≈1 nm to <1 µm) with multimodal population Highly trained users
Distribution (population-based)
Mass and density
SEC In suspension, Size (≈1 to ≈200 nm) Provides highly monodisperse Absolute size quantification might [147]
porous structure Size distribution sample fractions, compatible be challenging due to particle-solid
(population-based) with industrial settings phase interaction
FFF In suspension Size (≈1 nm to ≈50 µm) Highly tunable (different accu- Sample recovery and choice [155,156]
Size distribution mulation forces can be used) of experimental parameters
(population-based) Provides monodisperse can be challenging
sample fractions
FCS/FCCS In suspension Size Selectivity provided by the Need of fluorescent labels [159]
(≈1 nm to ≈1 µm) fluorescence detection (if sample is not fluorescent)
Size distribution
(according to fluorescence labeling)
TRPS In suspension Size Tunable detection range, Requires (highly) conductive solutions, [163,164]
(conductive solution) (≈50 nm to 10 µm) single-particle resolution, requires careful calibration
Size distribution provides information on
(number-based) surface charge
Shape
Concentration
Zeta potential
compositions render DLS extremely suitable for routine and kBT

rapid characterization measurements. D= (3)
6 πηR h
DLS size estimation is based on the determination of the
free diffusion coefficient of suspended particles.[130] A laser where D is the diffusion coefficient, kB is the Boltzmann
is transmitted through a measurement cell containing the constant, T indicates the temperature at which the analysis
particle suspension, and the random thermal motion of the is carried out, and η is the kinematic viscosity of the solvent.
particles causes time-dependent fluctuations of the intensity Rh indicates the hydrodynamic radius of the particles, i.e., the
of the scattered light (Figure 7a). The time correlation of these radius of a spherical particle whose diffusivity equals that of
fluctuations is directly dependent on the diffusion time of the the sample under examination. For nonspherical particles, Rh
scattering objects, i.e., the particles, through the laser spot. An can differ significantly from the actual particle dimensions, and
autocorrelation analysis of the time-domain signal is then used previous knowledge of particle shape is required to translate
to estimate the diffusion coefficients of the particles, which are this parameter into more representative size estimates.
proportional to the scattering correlation time, and to investi- DLS is an ensemble-based technique, and care needs to be
gate the population size distribution, which affects the shape of taken when interpreting measurements of particle suspensions
the autocorrelation curve. with a high level of polydispersity (Figure 7b,c). Different algo-
Particle size is then inferred from the diffusion coefficient rithms have been developed to interpret the correlation func-
through the Stokes–Einstein equation (Equation (3))[130] tion resulting from polydisperse populations of particles.[131–133]
Figure 7. DLS and NTA. a) Schematic illustration of DLS (top) and NTA (bottom) instruments. DLS measures the scattered-light fluctuations caused
by the Brownian motion of particles in solution. NTA tracks particle diffusion by video acquisition; b) comparison of size characterization of monodis-
perse nanoparticle samples by DLS and NTA; c) comparison of size characterization of bimodal dispersions of polystyrene nanoparticles of 100 and
400 nm nominal size. The ratios on the top of the graphs indicate the concentration ratio of the 100 and 400 nm particles in solution. (b,c) Adapted
with permission.[138] Copyright 2010, The Authors. Published by Springer Nature.
However, due to the high dependence of scattering intensity theoretically valid only for zero-angle scattering. However, the
on particle size, size estimation is typically biased toward the scattering intensity cannot be measured at this angle due to
larger population components.[134] Therefore, sample-size frac- the transmitted light. For particle below ≈40 nm, scattering can
tionation prior to data acquisition is the most effective route to be measured at a single angle and used to estimate the zero-
ensure correct estimation of size parameters for highly polydis- angle scattering intensity.[29] In contrast, for larger particles, the
perse sample populations.[135] measurement at multiple scattering angles is necessary due to
Finally, DLS can be carried out on any optically clear particle the different forward and backward scattering intensities. SLS
suspension, provided that the kinematic viscosity of the solvent is an ensemble-based measurement technique and, as for DLS,
is known or is characterized by using calibrated monodisperse scattering intensity scales dramatically with particle size. For
particle suspensions. polydisperse particle populations, the use of sample fractiona-
tion is advisable to properly estimate the radius and molecular
weight of the particles in suspension.[29]
3.2.2. Static Light Scattering (SLS)
Static light scattering is used to estimate the radius of gyration 3.2.3. Nanoparticle Tracking Analysis (NTA)
and the molecular weight of proteins and nanoparticles in sus-
pensions. The radius of gyration rg (Equation (4)) represents Similar to DLS, nanoparticle tracking analysis provides size
the mass-weighted radius of the particle and it is defined as characterization based on the free diffusion behavior of parti-
cles in solution. NTA measures particle diffusion by tracking
rg2 =
∑ mr
i
i i
2
(4)
the random motion of single particles in solution via high
temporal-resolution video acquisition and enhanced contrast
mi
microscopy[137] (Figure 7A). NTA measures the diffusion of
where mi and ri are the mass and distance from the center of the objects in suspension with single-particle resolution, and
mass, respectively, of the ith element composing the particle.[29] hence it is particularly suitable for the characterization of size
As an example, rg = 3 a 2 for a spherical particle, where a is the distribution for highly polydisperse nanoparticle populations
5
particle radius. As such, a priori knowledge of particle shape is (Figure 7b,c). NTA can be used for measuring particles with
necessary to properly relate the estimated radius of gyration to hydrodynamic radii ranging from ≈30 nm to ≈1 µm.[138] The
the particle dimensions.[29] ability to detect and track single particles in solution also makes
Molecular weight and radius of gyration are estimated by it possible to estimate the particle concentration during size
measuring the time-averaged intensity of light scattered by a characterization. Despite these key advantages with respect to
colloidal suspension at different concentrations.[54] A Debye scattering-based size characterization, the higher costs of the
plot is then generated by plotting the variation of scattering equipment, the need of sample concentration or dilution to
intensity as a function of particle concentration, and the slope typically 108 to 109 particles mL−1 (which might differ signifi-
and intercept at zero concentration of the obtained curve can cantly from the sample concentration required for the intended
then be used to calculate the molecular weight and the radius application), and the required high level of scattering of the
of gyration of the particles in solution.[54,136] This relation is particles limit the wide application of NTA. Therefore, DLS still
remains the main technique used for routine size characteriza- particle surface coverage.[4,144] Furthermore, the possibility of
tion of nanoparticles in liquid. integrating multiwavelength optical characterization during
sample sedimentation largely expands the range of applica-
tion of AUC in nanoparticle characterization, enabling, e.g., to
3.2.4. Electrophoretic Light Scattering investigate nanoparticle-protein interaction[145] or particle shape
and optical behavior of polydisperse sample populations in a
Electrophoretic light scattering is used to estimate the zeta single experiment.[146]
potential ζ of nanoparticles in suspension from their electro- However, data interpretation and translation of sedimenta-
v
phoretic mobility µe, defined as µe = E , where v is the particle tion parameters into size and shape information is not trivial,
velocity and E the externally applied electric field.[139] The zeta and can be affected by several factors, such as sample concen-
potential and the electrophoretic mobility are then related by tration or sample interactions. The high costs of the instrumen-
the Henry equation (Equation (5)) tation and the long analysis times required, however, strongly
limit the widespread application of this technique.
2ε rε 0 f (KA )
µe = ζ (5)
3η
3.2.6. Size Exclusion Chromatography (SEC)
where εr is the relative permittivity of the solution, ε0 is the
permittivity in vacuum, f(KA) the Henry function, and η the Typically used for the separation and analysis of macromol-
viscosity of the solution. Similar to DLS, a laser is transmitted ecules and viruses, size exclusion chromatography is increas-
through the measurement cuvette and an electric field is ingly used for the separation of polydisperse populations of
applied to the particle suspension. If the particles are charged, nanoparticles according to their hydrodynamic radius.[147] A
they experience a movement toward the electrode with opposite SEC separation column is packed with porous microparticles
charge sign with respect to their surface charge. The movement featuring different pore apertures ranging from a few nanom-
of the particles induces a shift in frequency, a so-called Doppler eters up to hundreds of nanometers, forming the so-called solid
shift, in the scattered light, which is proportional to the particle phase of the column. The sample solution, or liquid phase, is
velocity.[140] This Doppler shift and the movement direction then inserted into the column and pressure is applied to make
toward the positive or negative electrode are then used to esti- the nanoparticle suspension flow through the column. SEC
mate the electrophoretic mobility of the particles in suspension separation exploits the differential diffusion of nanoparticles
and, in turn, their zeta potential. However, ζ depends on the into the pore structure of the solid phase: Particles with smaller
particle environment, and the pH and ionic strength of the solu- hydrodynamic radius tend to diffuse more into the porous struc-
tion or the particle concentration strongly affect its value.[139] ture of the solid phase, while larger components of the popula-
As such, the zeta potential of a suspension of nanoparticles tion travel faster through the column due to shorter diffusion
should be characterized under conditions that closely mimic paths and less interaction with the solid phase. Effective sample
the final working environment of the nanoparticles. Different fractionation is obtained as the liquid phase is eluted from the
experimental arrangements have been developed to enable column. SEC is commonly used in both research and industrial
measurements in concentrated samples (>0.01–0.1% v/v), settings for producing highly monodisperse sample popula-
where multiple scattering effects and high solution turbidity tions both in polymer science and biotechnology, e.g., for the
would prevent electrophoretic characterization via light scat- separation and analysis of proteins and viruses.[148] Similarly,
tering detection.[141] Electrophoretic light scattering is often per- SEC separation has been widely used for the purification of
formed with the same instruments that are also used for DLS nanoparticles.[149–151] However, the application of this technique
characterization. for quantitative analysis and characterization of particle size
has numerous limitations. Although particles of known dimen-
sions can be used to calibrate the relationship between elution
3.2.5. Analytical Ultracentrifugation (AUC) time and particle size, absolute size quantification and frac-
tionation can be highly challenging.[152] Particles with different
Analytical ultracentrifugation provides the size, size distribu- surface charges and shape can interact differently with the solid
tion and density of a suspension of nanoparticles by detecting phase, which complicates the prediction and translation of elu-
the sedimentation properties of the nanoparticles when sub- tion times into size information. However, the compatibility of
jected to a centrifugal force.[142] The instrument consists of a this technique with industrial production and characterization
high-speed centrifuge equipped with a transparent cell and a is promoting efforts toward the investigation of SEC separation
detector to monitor the evolution of the dynamic concentration for quantitative nanoparticle size analysis.[147,152,153]
profile of the particles along the cell axis during centrifugation
(sedimentation velocity), and the final thermodynamic equi-
librium (sedimentation equilibrium).[143] During sedimenta- 3.2.7. Field-Flow Fractionation (FFF)
tion, particles of different size and molecular weight separate
and sediment at different rates, therefore multimodal sample Field-flow fractionation encompasses a class of techniques
distributions can be detected. The extremely high sensitivity that use differential transport of nanosized objects along a
of AUC to variations in density and mass of the nanoparticles microchannel for separation and characterization. The sample
renders this technique particularly suitable for investigating is introduced in a channel featuring a parabolic laminar-flow
profile and a force perpendicular to the flow direction is then chemical rate constants, molecular concentrations, or—espe-
applied, which induces sample accumulation against one cially in the case of FCCS—binding events in complex systems
channel wall. Due to sample diffusion, a concentration gradient that do not cause a significant change in mass and hence in
is generated across the cross section of the transport channel. diffusion properties.[162] Typically, the concentration should be
Particle diffusion is dependent on the hydrodynamic radius, rather low (on the order of 5 molecules at any given time in
therefore particles of different dimensions establish different the excitation volume), and the technique is especially useful
concentration gradients in the channel cross section. These for characterizing small, dynamic systems.
gradients of distributions ultimately result in different mean
transport velocities and elution times of particles with different
sizes.[154] The perpendicular accumulation force, flow velocity 3.2.9. Tunable Resistive Pulse Sensing (TRPS)
and channel dimensions can be tuned to change the resolving
power of the FFF separation. FFF can provide absolute size Tunable resistive pulse sensing is a technique for the charac-
quantification if the relationship between the elution time and terization of size, zeta potential and concentration of particles
the applied force is known or can be calibrated using stand- in suspension with single-particle resolution and in complex
ards. In addition, the technique can be used to separate the medium conditions.[163] TRPS is based on the Coulter prin-
sample population and generate monodisperse distributions ciple, which measures the variation of ionic current through a
of particles for downstream analysis. The low shear stresses pore due to partial blocking of the pore aperture caused by the
experienced by the particles and the tunable selectivity of the passage of a particle. As such, TRPS requires the use of con-
FFF analysis have rendered this technique a commonly used ductive solutions for the analysis. However, this requirement
analysis method for investigating nanoparticle populations.[155] renders this technique compatible with the characterization of
As different forces can be applied to induce sample nanoparticles under physiological buffer conditions for inves-
accumulation at the channel wall, FFF has been used to investigating the fate and interactions of nanoparticles in complex
tigate different physicochemical properties, such as magnetic media.[164] TRPS differs from standard resistive pulse meas-
susceptibility by use of magnetic gradients, dielectric properties urements, as the pore aperture can be tuned by stretching the
by applying external electric fields, or particle composition by pore-containing membrane. The tunability of the pore aperture
applying thermal gradients or sedimentation forces.[156] How- enables both to clear the pore in the case of clogging, as well
ever, FFF is typically used for size-based separation and inves- as to adapt the measurement sensitivity during acquisition.[165]
tigation of polydisperse samples. The most common FFF Equation (6) describes the flux of particles (particles m−2 s−1)
technique is flow field-flow fractionation (usually referred to through the pore[165]
as FIFFF), which exploits a perpendicular cross-flow to induce
Cε
a drag force for driving the particles toward the accumulation
wall. This perpendicular force is proportional to the cross-flow
J ≈ ( J ep + J eo ) + J pdf =
η
(ζ part − ζ pore ) E + CA Q (6)
velocity u, the fluid dynamic viscosity η and the particle hydro-
dynamic radius Rh, as defined by Stokes’ law F = 6πηRhu .[157] where Jep and Jeo represents the electrophoretic and electro-
Porous membranes at the channel walls allow fluid to exit osmotic fluxes generated by the electric field E, respectively,
perpendicular to the axial transport direction, while retaining and Jpdf indicates pressure driven flux with volume flow rate Q.
particles within the separation channel. According to the con- C indicates the particle concentration, and ε and η are the solu-
figuration of the cross flow and used porous membrane, FIFFF tion permittivity and viscosity. Finally, ζpart and ζpore are the zeta
is further divided in different subtechniques, with hollow-fiber potential of the particle and pore, respectively.
FIFFF (also known as HF5) and asymmetric-F1FFF (AF4) From Equation (6), when Jep and Jeo are negligible with
being the most commonly used techniques for routine analysis. respect to Jpdf, particle concentration can be measured by cal-
culating the rate of recorded events, if the pressure-driven flow
is known. The amplitude of the induced current variations
3.2.8. Fluorescence Correlation Spectroscopy (FCS) carries information on particle size and scales with the ratio of
particle size and pore aperture. Furthermore, investigation of
Fluorescence correlation spectroscopy and fluorescence cross- the induced-peak shape can also provide information on par-
correlation spectroscopy (FCCS)[158,159] rely on the detection of ticle shape and anisotropy.[166] Finally, TRPS can also be used to
emitted light rather than scattered light. As fluorophores dif- measure the surface charge of particles in solution. This prop-
fuse in and out of the confined volume of excitation light in a erty can be obtained either by varying the applied pressure and
confocal microscopy setup, they emit fluorescent light, and the measuring the electric field across the pore so that J = 0, or,
resulting intensity fluctuations of the emitted light are recorded if Jpdf is known from calibration, by estimation from the par-
by a detector. Similar to DLS, the intensity fluctuations can be ticle velocity, which is proportional to the width of the induced
used to generate an autocorrelation or cross-correlation func- peak.[165] This latter method, in particular, enables to obtain
tion which in turn allows for extracting translational diffusion single-particle surface-charge measurements.
coefficients, and the diffusion coefficients themselves depend Although it is generally assumed that the signal amplitude
on the hydrodynamic particle size.[160,161] At the same time, the scales linearly with the particle volume, this relationship deviates
amplitude of the signal depends on the number of fluorophores from linearity when the particle size approaches that of the pore
present in the excitation volume. Besides determining diffu- aperture or for highly stretched pores, which might feature a
sion coefficients, the techniques can also be used to investigate nonpredictable aperture shape and size.[163] As these effects
are difficult to model with precision, the use of calibrated par- as their interaction and assembly.[168,171] The power of LC TEM
ticles of different dimensions can be used to characterize the has been demonstrated by studying several aspects of synthesis
response of TRPS for samples of different sizes. and use of gold nanoparticles, namely their nucleation rates,[172]
and the dynamics and assembly between bare and function-
alized particles (Figure 8b).[173–176] Finally, atomic-resolution
3.3. Advanced Techniques transmission electron microscopy (AR-TEM) enables the study
of dynamic processes and provides real-time observation of
The advanced techniques presented here are of interest for the changes at the molecular level with atomic sensitivity.[177,178]
characterization of specific properties of nanoparticles and of The first study in 2007 reported AR-TEM movies of conforma-
their behavior, which cannot be accessed using standard char- tional changes of a single hydrocarbon molecule.[179] The use of
acterization methods. The methods discussed in this section this technique for nanoparticles is just at the beginning, but it
are summarized in Table 3. opens interesting avenues, such as the investigation of nano-
particle-protein interactions.[177]
3.3.1. In Situ Liquid-Cell (LC) and Atomic-Resolution (AR) TEM

3.3.2. Electron Cryo-Microscopy (Cryo-EM)
As discussed previously when presenting the dry-state charac-
terization techniques, TEM is a unique and powerful charac- Electron cryo-microscopy fills an important gap, as it allows
terization method, however normally restricted to samples in direct imaging of biological nanoparticles that cannot be dried
dry state. In situ liquid-cell TEM has been developed to over- without damage.[180] Such nanoparticles are abundant in nature,
come this major limitation.[167] Here, a nanoparticle suspension e.g., protein/RNA complexes, organelles, and viruses.[181]
is sandwiched between two electron-transparent windows and However, also artificial nanoparticles can be challenging to
the system is hermetically sealed to protect the liquid from the dry without affecting their stability, as it is often the case for
vacuum environment of the TEM[168–170] (Figure 8a). This tech- liposomes and many types of drug formulations. The field of
nique makes it possible to image previously “unobserved” pro- cryo-EM has been transformed by a series of breakthroughs
cesses such as nucleation and growth of nanoparticles as well in hardware development and computational methods over
Table 3. Summary of advanced characterization methods.

LC- and AR-TEM In solution Size Single-particle and atomic Extremely low throughput, [170]
(atomic to ≈100 nm) resolution, imaging high costs, highly
Shape of dynamic processes trained users
(2D projection)
Dynamics
Cryo-EM Flash-frozen solution Size No drying or staining Very low signal-to-noise ratio, [82,181,183]
(<1 to ≈100 nm) required, compatible with high costs, highly trained users
Shape materials sensitive to e-beams
(2D projection and 3D
reconstruction)
Electron tomography Dry and in solution Size 3D imaging of individual Extremely low throughput, [53,187]
(<1 to ≈100 nm) nanoparticles high costs, highly trained users
Shape
(3D imaging)
NMR cryoporometry and In solution (freezing Pore aperture Investigation of pore volume Limited throughput, [194]
DSC thermoporosimetry and thawing) Pore volume and volume accessible ensemble-based
to solvents in solution
Super-resolution Dry and in liquid Surface coating Selectivity provided by Limited optical [199]
microscopy NP structure and behavior fluorescent tag, can be performed (>30 nm) and temporal
on dynamic systems (50–100 ms) resolution
Single molecule fluores- In liquid Pore structure Information on the internal Host material must have low autofluo- [204,205]
cence microscopy Pore functionalization porous structure of materials rescence, dye has to be able to diffuse
inside the porous structure
Nanomechanical Dry and in Mass Information on composition, Low throughput, not always possible to
resonators suspension Density size and interactions obtain single-particle characterization
Size (indirect) with the solvent
Figure 8. LC-TEM and AR-TEM. a) A nanoparticle suspension is sandwiched between two electron-transparent windows (e.g., made of thin silicon
nitride) to prevent the evaporation of the solution in the high vacuum environment during TEM imaging; b) formation of gold nanorod chains by
sequential attachment of nanorods in solution. Adapted with permission.[176] Copyright 2017, American Chemical Society; c) conformational changes
of a biotinylated molecule grafted to the tip of a tapered single-walled carbon nanotube. The number refers to the frame number extracted from the
TEM movies. Adapted with permission.[178] Copyright 2015, American Chemical Society.
the past decade. Although the attainable resolution is strongly can tolerate a higher electron dose, tilt-series can be acquired to
sample dependent, values down to 2 Å have been reported,[182] enable reconstruction of the 3D-structure of individual particles
while values better than 10 Å are achieved relatively routinely by cryo-electron tomography (cryo-ET).[82,181]
today.[183] Another valuable technique is cryogenic scanning elec-
Ice crystallization constitutes the biggest risk for the integ- tron microscopy (cryo-SEM). Because the signal in SEM
rity of the structure when samples are prepared for cryo-EM. To emerges mainly from the surface of the sample, different
prepare samples that are frozen amorphously, or vitrified, for preparation techniques than in cryo-TEM are required to access
TEM analysis, a few microliters of a suspension are applied to a nanoparticles embedded in vitreous ice. Vitrified samples for
TEM grid, blotted in a humidified chamber to leave a water film cryo-SEM are often made by high-pressure freezing, which
only ≈100 nm thick, and flash frozen by plunging the grid into is able to yield amorphous ice specimens up to several hun-
liquid ethane or propane at high speed (Figure 9a). dred micrometers thick. To expose the inner structures, one
TEM is then used to obtain 2D projection images of the approach is to fracture the vitrified blocks in a method known
nanoparticles embedded in vitreous ice. These images often as freeze fracture. Alternatively, focused ion beam (cryo-FIB)
have very low contrast, as the electron density in organic mate- milling or cryo-ultramicrotomy can be used to make serial sec-
rials differs only weakly from the solvent (Figure 9b).[181,184] tions. An important advantage of cryo-SEM is that the cost for
When the particles are identical, such as certain types of viruses the instruments is significantly lower than for TEM. The wide
or protein complexes, analysis algorithms are able to recon- field of view of the SEM is useful for observing the interaction
struct the 3D structure from a large set of images of randomly of nanoparticles with larger objects, such as whole cells. How-
oriented single particles. When the particles are not identical, ever, the attainable resolution rarely exceeds 1 nm and does not
the projection images still provide valuable information about provide the detailed insight into the molecular structure that
their inner structure, size, and size distribution. If the material modern cryo-TEM can achieve.[181]
reconstruction), and a high level of complexity (especially the

tomographic reconstruction of the 3D image), which limit this
technique to highly trained specialists. In addition, particle drift
and beam damage induced changes of the object during the
investigation have to be taken into account during the recon-
struction and rendering of the 3D image.
3.3.4. NMR Cryoporometry and DSC Thermoporosimetry
Liquids confined within nanometer-sized pores experience a

decrease in freezing and melting temperature with respect
to their bulk counterparts as a consequence of the extremely
high surface-to-volume ratios of the confined crystals and the
imposed surface curvature.[189,190] Measuring the variations of
the temperature of phase changes provides direct information
on the volume and the dimensions of the pores accessible to
the liquid[191] (Figure 10a). Although both freezing and melting
phase changes can potentially be used for such a study, detec-
tion of melting temperatures is usually preferred to avoid
uncertainties arising from supercooling effects. These transi-
tion temperatures can be measured with high precision using
either differential scanning calorimetry (DSC) thermoporo-
Figure 9. Cryo-EM and electron tomography. a) Main steps required for simetry[192] or nuclear magnetic resonance (NMR) cryoporo-
sample preparation for cryo-EM; b) spherical nanoparticles formed by metry.[193,194] With respect to commonly used gas sorption
a poly(styrene) core with grafted chains of poly(styrene sulfonate acid). techniques to measure the pore windows and the pore sizes,
Cesium ions and bovine serum albumin were used to enhance the image NMR and DSC measurements are carried out in liquid and no
contrast of the brush layer. Reproduced with permission.[184] Copyright 2005,
drying of the sample that could potentially affect the nanopar-
American Chemical Society; c) to perform electron tomography, different
EM micrographs are obtained by varying the angle between the sample and ticle framework is required. This is particularly important when
the incident EM beam; d) Cryo-EM tomogram revealing the localization detecting the pore size and pore size distribution of delicate
of clusters of superparamagnetic iron oxide nanoparticles (purple) within structures envisioned to work in liquid environments, such as
the bilayer of liposomes (yellow/green). Scale bar 50 nm. Reproduced with hydrogel nanoparticles[195] or porous biopolymeric nanoparti-
permission.[188] Copyright 2014, American Chemical Society. cles.[196] Furthermore, the ability to probe nanoparticles directly
in solution enables studying dynamic variations in the porous
3.3.3. Electron Tomography structure of the nanoparticles when exposed to different solvent
conditions (Figure 10b).[197]
Conventional microscopy provides a 2D projection of a 3D
object. Therefore, information such as shape, thickness or
uniformity (e.g., spatial dimension) of nanoparticles can only 3.3.5. Super-Resolution Microscopy
be roughly estimated, and the interpretation of the projection
of complex nanoparticles is quite challenging and can lead to The ability of providing dynamic information with spatial reso-
incorrect conclusions. Hence, more information is needed to lution below the conventional diffraction limit of optical micros-
make reliable statements about the 3D structure of the investi- copy has rendered super-resolution microscopy, or nanoscopy,
gated samples. an essential tool in cell and tissue biology research.[198] Spatial
Tomography is the best approach to provide 3D information resolution below the diffraction limit is achieved by employing
of objects. This method is based on the acquisition of multiple acquisition strategies to improve the localization of single emit-
2D projections of the sample at different angles relative to the ting fluorophores (such as photoactivated localization micro
incident beam, and on the reconstruction of a 3D representa- scopy or stochastic optical reconstruction microscopy) or by
tion from these partial images[185] (Figure 9c). Tomography has reducing the size of the point-spread function of the emitted
been originally developed for X-ray analysis and is nowadays a light (such as stimulated emission depletion or structured illu-
widely used approach in medicine (computed tomography), but mination microscopy) (Figure 11a). Recently, super-resolution
it is not suitable for nanoparticle analysis. microscopy has also found applications in material science
Electron tomography is the technique of choice for studying for providing superior optical characterization of samples
the 3D structure of a nanoparticle (Figure 9d).[186–188] Recently, under investigation.[199] Although higher spatial resolution
this method could be expanded to study individual nanocrystals can usually be obtained by electron microscopy or scanning
in solution.[52] However, electron tomography has several impor- probe microscopy, fluorescence-based characterization can
tant limitations such as poor statistics, very low throughput (a provide higher selectivity, higher contrast and nonsuperfi-
single nanoparticle 3D visualization can require about one day cial characterization, depending on the fluorescent labeling
of data acquisition and multiple days of data processing and 3D strategy employed. Examples of such enhanced selectivity and
the application include the investigation and

mapping of the internal porous structure and
defects of porous nanomaterials by following
the diffusional movement of fluorescent dyes
inside the pores (Figure 11d),[206–208] or the
determination of the diffusion coefficients
of dyes and oligonucleotides inside a porous
nanomaterial in dependence of the function-
alization of the pore walls.[209–211] The latter
can give detailed information on how the
presence and density of functional groups
on the pore walls affect the diffusional move-
ment of the guest mole cules, for example
through electrostatic interactions.
3.3.7. Nanomechanical Resonators
Micromechanical resonators present a

characteristic resonance frequency which
depends on the oscillating mass of the reso-
nator. Thanks to the extremely low inertial
masses of nano- and micron-sized resona-
tors (attogram to nanogram inertial mass)
and the high quality factors, adsorption of
particles on the surface of a resonator causes
a detectable shift in resonance frequency,
proportional to the mass of the adsorbed par-
ticle. This nanomechanical-based mass spec-
trometry has enabled the characterization of
mass of single intact nanoparticles regard-
less of the ionization state of the sample,
allowing the mass measurement of neutral
particles in the Megadalton to Gigadalton
Figure 10. NMR cryoporometry. a) An ideal NMR cryoporometry curve reporting the main regime, which is not possible with standard
curve features; b) melting curve of porous polymer nanoparticles loaded with Carboplatin mass spectrometry approaches.[212,213]
drug after different incubation times in artificial cerebral spinal fluid. The NMR cryoporometry This method of detection can be extended
curves show an increase in small pores and total pore volume within the first 2 days, which to colloidal suspensions by embedding a
corresponds to a burst release of drug in solution. Reproduced under the terms of the CC-BY microfluidic channel into the resonator to
license.[197] Copyright 2014, the Authors. Published by Elsevier.
enable particles in solution to cross it, while
retaining its oscillation in a vacuum environ-
nonsuperficial characterization are the visualization of the ment.[214] These devices, called suspended microchannel reso-
hollow nature of lipid coated nanocapsules,[200] the charac- nators, have enabled the characterization of samples of gold
terization of the internal compartmentalization of core–shell nanoparticle mixtures ranging from 10 to 20 nm in diameter,
microgels,[201] the swelling and contraction of microgel nano- with single-particle detection capability.[215] By allowing mul-
particles (Figure 11b), [202] and the penetration and adsorption of tiple particles to flow simultaneously through the embedded
fluorescently labeled proteins with different molecular weights channel and by employing an autocorrelation analysis of the
into the framework of mesoporous silica nanoparticles with time-domain mass signal, the resolution of these devices could
different pore sizes (Figure 11c).[203] be extended to the characterization of nanometer-sized polymer
particles.[216] This technique, called Nanomechanical Mass Cor-
relation Spectroscopy, was used to investigate the complex par-
3.3.6. Single Molecule Fluorescence Microscopy titioning of binary solvent mixtures within the framework of
mesoporous metal-organic framework (MOF) nanoparticles
Single molecule fluorescence microscopy and tracking can be with different surface functionalization (Figure 12).[217] The
used to obtain valuable information about the porous system particles exhibited different effective densities in solution as
and the host-guest interactions of nanomaterials.[204,205] The a function of pore functionalization and solvent composition,
dynamics of guests inside porous nanoparticles, and the inter- indicating that the local microenvironment within the pore
actions of the guests with the walls of the nanoparticles are structure of porous nanoparticles might not reflect the bulk
sometimes the key for the targeted application. Examples of solvent composition.
Figure 11. Super-resolution microscopy and single molecule fluorescence microscopy. a) Different methods to achieve optical resolution below the
diffraction limit. In STED microscopy (left), a doughnut-shaped light pattern (red) is used to suppress the emission of excited fluorophores (green) in
an annulus via stimulated emission (“off” state). The emitting region is thereby confined to a subdiffraction-sized spot at the center of the doughnut.
In STORM microscopy (right), only a sparse collection of randomly distributed fluorophores is turned on at any time by a low-intensity wide-field
excitation Single molecules are localized with subdiffraction accuracy. This process is repeated for multiple images to reconstruct the complete object;
b) contraction of fluorescently labeled microgel nanoparticles upon addition of methanol in solution visualized by dSTORM. Reproduced with permis-
sion.[202] Copyright 2016, Elsevier; c) adsorption of different labeled proteins on the surface and within the porous structures of mesoporous silica nano-
particles featuring pores of different size. On the left, the sphere of best fit as calculated from the STORM images. On the right, the fluorescence data
are flattened and their distribution is plotted in 2D. Reproduced with permission.[203] Copyright 2017, American Chemical Society; d) left: “maximum
projection” overlay of the individual frames of a movie that shows single dye molecules diffusing inside the pores of a mesoporous silica thin film.
Right: The underlying porous structure was reconstructed by fitting each single molecule in each frame and linking them together to form trajectories.
The trajectories show the real pore structure, including pore alignment and defects, and can also be used to extract dynamic information such as the
diffusion coefficient of the dye inside the porous host. Reproduced with permission.[207] Copyright 2012, American Chemical Society.
4. Nanoparticle Characterization Overview Solids and powders often come as aerosols or as particles
deposited on a surface from a suspension. These samples
Nanoparticle characterization is generally motivated by the provide considerable freedom for studying size distribution
requirements of specific applications. Such requirements are and particle morphology by electron microscopy, atomic force
often given in the form of safety regulations or quality control microscopy, X-ray scattering, or mass spectrometry. Dry sam-
standards. If the application does not already prescribe a spe- ples can be characterized in air or vacuum where no solvent
cific set of characterization techniques, some general guide- molecules interfere with the measurement.
lines can be considered. Here we aim to provide some general In most situations, electron microscopy is the method of
recommendations to help with the selection of methods in a choice for dry samples because the technique is extremely
wide range of applications. As a starting point, the flowchart versatile, precise, relatively simple to use (especially SEM),
in Figure 13 provides a framework organizing the most impor- and widely available. Unfortunately, nonconducting samples
tant techniques described in this article into groups and key accumulate charge and often deform slowly in the electron
questions to help the reader decide which techniques are most microscope. Beam-sensitive samples need to be coated with a
suitable for the sample under investigation. Although it is not thin conducting layer (typically platinum, gold, or graphite) or
possible to provide a comprehensive overview of all techniques analyzed at elevated pressure by environmental SEM at a low
for all types of samples, we believe that this workflow can be dose. TEM analysis of nonconducting samples also requires
applied to many typical cases of nanomaterials. attention to the same issues of beam damage and charging.
Nanoparticle characterization techniques can be categorized When electron microscopy cannot be used, AFM can be con-
as targeting solid samples or powders, shown on the left side of sidered as an alternative. On the one hand, AFM analysis lacks
the flowchart, or suspensions, featured on the right. many of the material contrast and elemental analysis options
Figure 12. Nanomechanical mass correlation spectroscopy. a) The detection principle of nanomechanical mass correlation spectroscopy. Particles
flowing through the microfluidic channel embedded in a mechanical resonator induce transient fluctuations to the resonant resonance frequency,
which are proportional to the particle buoyant mass. An autocorrelation analysis of the time-domain signal is used to enhance the particle contribution
with respect to the uncorrelated noise background; b,c) MIL-101(Cr) porous nanoparticles were suspended in binary mixtures of different polarities
to investigate the dependence of the particle density to the composition of the solvent mixtures. (a–c) Reproduced under the terms of the Creative
Commons Attribution-NonCommercial license.[217] Copyright 2018, The Authors. Published by Wiley-VCH.
offered by SEM. AFM also has lower throughput and can be compounds, this technique now is able to access the range of
challenging to perform, as particles that are not stably attached large macromolecular complexes in the Megadalton range and
to a surface can be easily dislodged by the scanning AFM tip. even beyond. However, the measurement is based on the mass-
On the other hand, AFM provides sub-nanometer topographic to-charge ratio and provides limited information on the actual
resolution and can be used to analyze mechanical properties. size and morphology of the particles.
SAXS provides less detail than microscopy in terms of the Measuring the surface area of nanoparticles is straightfor-
size distribution and morphology. Scattering patterns are gener- ward by gas sorption analysis when particles come in the form
ated from large populations of particles. Although it is possible of a dry solid or a powder.
to extract information about heterogeneity and aggregation of Nanoparticle suspensions are more challenging to analyze
nanoparticles from the scattering patterns, the level of detail by microscopic methods than dry samples. In liquids, electron
to expect does not match the fine grained size distribution microscopy is only an option in a few cases where liquid cell
and morphology that can be extracted from analyzing electron TEM can be used. However, liquid cell TEM is still largely at
microscopy images or AFM images. a developmental stage. For most users today, AFM will be the
For the characterization of biological nanoparticles, the only routinely available method that is able to image nanoparti-
molecular weight is often more relevant than physical size or cles in liquids with nanometer resolution.
morphology. Mass spectrometry is becoming an excellent tool In choosing alternative characterization methods for sus-
even for large particles. Long reserved for low molecular weight pensions, it is first advisable to consider the expected degree
www.advmat.de
Figure 13. Flowchart. Flowchart depicting the main steps for the characterization of size, size distribution, morphology, surface charge, and porosity
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1901556 (20 of 26)
of nanoparticles, both in dry state and in solution. NP stands for nanoparticle.
www.advancedsciencenews.com
Adv. Mater. 2019, 31, 1901556

of polydispersity. The answer to this should be followed by a rich detail contained in electron micrographs is well suited to
critical assessment whether the target application requires the inform subsequent decisions regarding polydispersity, drying
discrimination of all populations within the sample or not. If artifacts, and beam sensitivity. This information forms a basis
the sample is highly polydisperse, with diameters spanning for deciding if and how subsequent analyses should be per-
more than one order of magnitude, then techniques on the left formed, and we suggest that the flowchart in Figure 13 is used
branch of the flowchart for suspensions in Figure 5 are usu- as a guide in this process.
ally preferable. The reason for this is that otherwise there is a
risk of obtaining results that are skewed by outliers, which can
affect the measurement even if their population is very small. 4.1. Assessment of Multifunctional Efficiency of Nanoparticles
Methods that are well suited for the characterization of poly-
disperse suspensions operate either by detecting particles one- To assess the efficiency and functionality of nanoparticles, we
by-one or by first separating the mixture into subpopulations, have to consider the ratio of functional and nonfunctional com-
sometimes referred to as bands or subensembles, which are ponents, and the number of synthetic steps required to produce
themselves made up of large numbers of particles with more these multifunctional products. We recently suggested to eval-
homogeneous characteristics. uate the synthesis of functional nanoparticles not only based
Among the most established techniques that operate by sepa- on yield and product selectivity (e.g., atom economy) but also
ration are analytical ultracentrifugation, field-flow-fractionation, with regard to the specific tasks the nanoparticles can fulfill
and chromatography. There are two potential caveats when (number of functional units) and the simplicity of the produc-
using these methods: first, if the nanoparticles are in a dynamic tion process (number of process steps)[18]
equilibrium, then separation can perturb this equilibrium and
may lead to misleading results. Second, the detection sensitivity nFU
FR: = (7a)
may be insufficient for some of the less-numerous subpopu- m BU
lations in the ensemble. If this is the case, one can consider
collecting fractions of the separated sample and subjecting nFU
these fractions to light scattering, SAXS, or more specialized PE: = (7b)
rPRS
analyses.
Methods for single nanoparticle characterization in liquid
have become commercially available only relatively recently. 2
nFU
MFE: = FR ⋅ PE = (7c)
Their physical principles are diverse, but all exploit the m BU ⋅ rPRS
mobility or the difference in conductivity/density to the sur-
rounding liquid. Thus, the solvent environment is an integral The functionality ratio (FR; Equation (7a)) of a nanoparticle
part of the measurement. With this in mind, NTA, TRPS, is defined as the number of functional units (nFU) divided by
and nanomechanical resonators are powerful tools for the the total number of building units of the nanoparticle (mBU).
characterization of nanoparticle suspensions. The main limita- The process efficiency (PE, Equation (7b)) of a nanoparticle syn-
tion of all three techniques is their relatively limited throughput. thesis is defined as the number of functional units (nFU) divided
Which of the three techniques is most appropriate depends on by the number of process steps (rPRS). Finally, the multifunc-
the specific characteristics of the sample and should be decided tional efficiency (MFE, Equation (7c)) involves both parameters
case by case based on the application examples provided in the FR and PE and results in a quantity that values both a high
detailed overview presented before. degree of functionality and a facile manufacturing process. The
Other important characteristics of nanoparticles in suspen- key advantage of this concept is that a functional nanoparticle
sion besides the mean size, size distribution, and morphology (e.g., nanocarrier) can be evaluated from the economic and
are the surface charge and surface area, including the area synthetic viewpoint before the expensive and time-consuming
inside of pores. Characterization of surface charge is most com- biological assessment takes place.[18]
monly performed by electrophoretic methods, but it is impor-
tant to note that TRPS, by the very nature of the method, is
also sensitive to surface charge. Porosity is a more challenging 5. Conclusions
parameter to quantify in solution. Note, however, that this is
not only due to the limitations of measurement technology. The growth of nanoparticle material classes and their applica-
Especially for microporous and mesoporous materials (pore tion to tackle global problems of our time requires progresses
sizes <2 and 2–50 nm, respectively), the accessibility of pores in nanoparticle characterization. The four properties size, shape,
can differ for bulky solvent molecules or bulky solutes. There- surface charge, and porosity of nanoparticles are intimately con-
fore, it is important to consider carefully what is being meas- nected with their functionality and their effects on health and
ured by the techniques available. the environment. Measuring these properties is important for
In summary, the choice of methods is typically a compro- translating potential benefits of nanomaterials into specific
mise between the demands of the application and the restric- applications. Characterization is also the first step to ensure
tions of the different techniques. As a rule, first the simplest that synthesized compounds possess the desired properties and
techniques should be applied to obtain answers to some key that the properties of different batches are reproducible.
questions. However, the quantitative results at this initial stage To be able to correlate the physicochemical properties of
should also be treated with a healthy degree of skepticism. The the nanoparticles with their performance in a specific task,
characterization needs to be both accurate and precise. There- corona that gives a “biological identity” to the particles.[20,218] As
fore, standardized standard operating procedures should be the application of nanoparticles in the biological context is a key
developed to improve the comparability of results between technology, comprehensive knowledge of the physicochemical
different materials and laboratories. It is crucial that new data parameters under physiological conditions is of fundamental
can be compared and evaluated with previously published importance.
results in a meaningful way. This will require conclusive and Nanoparticle characterization is a rich and complex disci-
harmonized analytical protocols to be applied worldwide. pline. In these exciting times for nanoscience, measurement
We wish to give the following general recommendations for and standardization often lag behind the rapid development of
nanoparticle characterization: new materials and their applications. However, the demands
on reproducibility and quality control increase steadily as new
1. Consider the intended form the target application of the materials leave the discovery stage, and we expect that this
nanoparticles. The form (colloidal suspension or solid pull will also drive innovations in the important field of nano
powder) often dictates the need for the specific characteriza- particle-measurement science and technology.
tions that should be used.
2. Place reasonable bounds on the precision and statistical con-
fidence with which physicochemical properties need to be Acknowledgements
measured to satisfy the needs of the application. For example,
the statistical demands in the quality control for pharmaceu- All authors would like to acknowledge their families who have supported
tical formulations will typically be higher than for industrial them throughout the writing process.
coating materials.
3. Develop or adopt standard characterization and standard
operating procedures to be systematically and thoroughly Conflict of Interest
followed for every batch.
The authors declare no conflict of interest.
4. The protocol used to measure the physicochemical
parameters as well as the meta-data should be described in
detail. Besides a detailed description of the experimental pro-
tocol, the experimental parameters, and method-specific set- Keywords
tings that were used, this should include how the data were nanoparticle characterization, nanoparticles, porosity, shape, size,
analyzed (e.g., what model was used to obtain a certain value). surface charge
5. In the case of suspensions, the exact chemical composition
of the liquid matrix should be specified. Both pH and ionic Received: March 10, 2019
strength should be measured and reported for aqueous Revised: April 19, 2019
solutions. Published online: May 30, 2019
6. Measured parameters should be calibrated against a stand-
ardized reference. If no standardized references for a
quantitative comparison with the nanomaterial at hand are [1] Recommendation on the Definition of Nanomaterial, European
available, internal references can be used to, at least, ensure Commission, Brussels, Belgium 2011, https://eur-lex.europa.eu/
consistency between batches. For example, obtaining similar legal-content/EN/TXT/?uri=CELEX:32011H0696 (accessed: May
size distributions from different nanoparticle batches and/ 2019).
or a constant reference sample should ensure the reproduc- [2] D. V. Talapin, E. V. Shevchenko, Chem. Rev. 2016, 116, 10343.
ibility of results. [3] R. Kodama, J. Magn. Magn. Mater. 1999, 200, 359.
7. Whenever feasible, more than one technique should be used [4] H. Kim, R. P. Carney, J. Reguera, Q. K. Ong, X. Liu, F. Stellacci,
to characterize the same quantity. Adv. Mater. 2012, 24, 3857.
[5] A. Cirri, A. Silakov, B. J. Lear, Angew. Chem., Int. Ed. 2015, 54,
8. The obtained results should be compared with published
11750.
data whenever feasible.
[6] M. Reza Nejadnik, W. Jiskoot, J. Pharm. Sci. 2015, 104, 698.
[7] C. Gong, M. R. S. Dias, G. C. Wessler, J. A. Taillon, L. G. Salamanca-
The development of inexpensive analytical methods should Riba, M. S. Leite, Adv. Opt. Mater. 2017, 5, 1600568.
be an integral part of future research. At the same time, existing [8] S. L. Lai, J. Y. Guo, V. Petrova, G. Ramanath, L. H. Allen, Phys. Rev.
techniques should be improved and their current boundaries Lett. 1996, 77, 99.
and limitations expanded. In particular, novel techniques are [9] W. H. Qi, M. P. Wang, J. Nanopart. Res. 2005, 7, 51.
needed for cross-validation, and to measure physicochemical [10] S. Deshpande, S. Patil, S. V. Kuchibhatla, S. Seal, Appl. Phys. Lett.
properties that are currently difficult to test. Especially the iden- 2005, 87, 133113.
tification of the interaction of ions and/or molecules inside the [11] M. Stratakis, H. Garcia, Chem. Rev. 2012, 112, 4469.
internal volume of porous nanoparticles in a liquid medium [12] F. Wang, H. Yu, S. Jeong, J. M. Pietryga, J. A. Hollingsworth,
P. C. Gibbons, W. E. Buhro, ACS Nano 2008, 2, 1903.
is so far a largely underexplored topic. Another important
[13] D. Segets, J. M. Lucas, R. N. Klupp Taylor, M. Scheele, H. Zheng,
aspect is the characterization of nanoparticles under the envi- A. P. Alivisatos, W. Peukert, ACS Nano 2012, 6, 9021.
sioned operating conditions of the nanomaterial. One impor- [14] W. Lin, Chem. Rev. 2015, 115, 10407.
tant lesson learned in the last decade is that in biological fluids [15] S. Kunjachan, J. Ehling, G. Storm, F. Kiessling, T. Lammers, Chem.
proteins assemble with the nanoparticles, leading to a protein Rev. 2015, 115, 10907.
[16] A. B. Chinen, C. M. Guan, J. R. Ferrer, S. N. Barnaby, T. J. Merkel, [39] E. Blanco, H. Shen, M. Ferrari, Nat. Biotechnol. 2015, 33, 941.
C. A. Mirkin, Chem. Rev. 2015, 115, 10530. [40] P. Horcajada, T. Chalati, C. Serre, B. Gillet, C. Sebrie, T. Baati,
[17] B. Pelaz, C. Alexiou, R. A. Alvarez-Puebla, F. Alves, A. M. Andrews, J. F. Eubank, D. Heurtaux, P. Clayette, C. Kreuz, J.-S. Chang,
S. Ashraf, L. P. Balogh, L. Ballerini, A. Bestetti, C. Brendel, S. Bosi, Y. K. Hwang, V. Marsaud, P.-N. Bories, L. Cynober, S. Gil, G. Férey,
M. Carril, W. C. W. Chan, C. Chen, X. Chen, X. Chen, Z. Cheng, P. Couvreur, R. Gref, Nat. Mater. 2010, 9, 172.
D. Cui, J. Du, C. Dullin, A. Escudero, N. Feliu, M. Gao, M. George, [41] S. Wuttke, S. Braig, T. Preiß, A. Zimpel, J. Sicklinger, C. Bellomo,
Y. Gogotsi, A. Grünweller, Z. Gu, N. J. Halas, N. Hampp, J. O. Rädler, A. M. Vollmar, T. Bein, Chem. Commun. 2015, 51,
R. K. Hartmann, M. C. Hersam, P. Hunziker, J. Jian, X. Jiang, 15752.
P. Jungebluth, P. Kadhiresan, K. Kataoka, A. Khademhosseini, [42] X. Liu, F. Zhang, X. Jing, M. Pan, P. Liu, W. Li, B. Zhu, J. Li,
J. Kopeček, N. A. Kotov, H. F. Krug, D. S. Lee, C.-M. Lehr, H. Chen, L. Wang, J. Lin, Y. Liu, D. Zhao, H. Yan, C. Fan, Nature
K. W. Leong, X.-J. Liang, M. Ling Lim, L. M. Liz-Marzán, 2018, 559, 593.
X. Ma, P. Macchiarini, H. Meng, H. Möhwald, P. Mulvaney, [43] B. Rühle, P. Saint-Cricq, J. I. Zink, ChemPhysChem 2016, 17,
A. E. Nel, S. Nie, P. Nordlander, T. Okano, J. Oliveira, T. H. Park, 1769.
R. M. Penner, M. Prato, V. Puntes, V. M. Rotello, A. Samarakoon, [44] D. Tarn, C. E. Ashley, M. Xue, E. C. Carnes, J. I. Zink, C. J. Brinker,
R. E. Schaak, Y. Shen, S. Sjöqvist, A. G. Skirtach, M. G. Soliman, Acc. Chem. Res. 2013, 46, 792.
M. M. Stevens, H.-W. Sung, B. Z. Tang, R. Tietze, B. N. Udugama, [45] R. A. Sperling, T. Liedl, S. Duhr, S. Kudera, M. Zanella, C.-A. J. Lin,
J. S. VanEpps, T. Weil, P. S. Weiss, I. Willner, Y. Wu, L. Yang, Z. Yue, W. H. Chang, D. Braun, W. J. Parak, J. Phys. Chem. C 2007, 111,
Q. Zhang, Q. Zhang, X.-E. Zhang, Y. Zhao, X. Zhou, W. J. Parak, 11552.
ACS Nano 2017, 11, 2313. [46] ISO 13320:2009(en), Particle Size Analysis—Laser Diffraction
[18] R. Freund, U. Lächelt, T. Gruber, B. Rühle, S. Wuttke, ACS Nano Methods, https://www.iso.org/obp/ui/#iso:std:iso:13320:ed-1:v2:en
2018, 12, 2094. (accessed: November 2018).
[19] S. Wuttke, M. Lismont, A. Escudero, B. Rungtaweevoranit, [47] N. Li, P. Zhao, D. Astruc, Angew. Chem., Int. Ed. 2014, 53, 1756.
W. J. Parak, Biomaterials 2017, 123, 172. [48] L. Talamini, M. B. Violatto, Q. Cai, M. P. Monopoli, K. Kantner,
[20] D. Rosenblum, N. Joshi, W. Tao, J. M. Karp, D. Peer, Nat. Commun. Ž. Krpetić, A. Perez-Potti, J. Cookman, D. Garry, C. P. Silveira,
2018, 9, 1410. L. Boselli, B. Pelaz, T. Serchi, S. Cambier, A. C. Gutleb, N. Feliu,
[21] M. Peller, K. Böll, A. Zimpel, S. Wuttke, Inorg. Chem. Front. 2018, Y. Yan, M. Salmona, W. J. Parak, K. A. Dawson, P. Bigini, ACS
5, 1760. Nano 2017, 11, 5519.
[22] F. Li, J. Lu, X. Kong, T. Hyeon, D. Ling, Adv. Mater. 2017, 29, [49] A. Banerjee, J. Qi, R. Gogoi, J. Wong, S. Mitragotri, J. Controlled
1605897. Release 2016, 238, 176.
[23] H. Goesmann, C. Feldmann, Angew. Chem., Int. Ed. 2010, 49, [50] C. Forestiere, G. Miano, S. V. Boriskina, L. Dal Negro, Opt. Express
1362. 2009, 17, 9648.
[24] W. J. Stark, P. R. Stoessel, W. Wohlleben, A. Hafner, Chem. Soc. [51] C. J. Orendorff, T. K. Sau, C. J. Murphy, Small 2006, 2, 636.
Rev. 2015, 44, 5793. [52] J. Park, H. Elmlund, P. Ercius, J. M. Yuk, D. T. Limmer, Q. Chen,
[25] M. Faria, M. Björnmalm, K. J. Thurecht, S. J. Kent, R. G. Parton, K. Kim, S. H. Han, D. A. Weitz, A. Zettl, A. P. Alivisatos, Science
M. Kavallaris, A. P. R. Johnston, J. J. Gooding, S. R. Corrie, 2015, 349, 290.
B. J. Boyd, P. Thordarson, A. K. Whittaker, M. M. Stevens, [53] P. Ercius, O. Alaidi, M. J. Rames, G. Ren, Adv. Mater. 2015, 27,
C. A. Prestidge, C. J. H. Porter, W. J. Parak, T. P. Davis, 5638.
E. J. Crampin, F. Caruso, Nat. Nanotechnol. 2018, 13, 777. [54] B. Chu, T. Liu, J. Nanopart. Res. 2000, 2, 29.
[26] P. Hirschle, T. Preiß, F. Auras, A. Pick, J. Völkner, D. Valdepérez, [55] Y. Ju-Nam, J. R. Lead, Sci. Total Environ. 2008, 400, 396.
G. Witte, W. J. Parak, J. O. Rädler, S. Wuttke, CrystEngComm 2016, [56] V. S. Kulkarni, Handbook of Non-Invasive Drug Delivery Systems:
18, 4359. Science and Technology, William Andrew, Oxford 2009.
[27] C. Y. Tay, M. I. Setyawati, J. Xie, W. J. Parak, D. T. Leong, Adv. Funct. [57] A. D. McNaught, A. Wilkinson, IUPAC Compendium of Chemical
Mater. 2014, 24, 5936. Terminology, 2nd ed., Blackwell Scientific Publications, Oxford
[28] B. Demeler, T.-L. Nguyen, G. E. Gorbet, V. Schirf, E. H. Brookes, 1997, https://doi.org/10.1351/goldbook.P04704.
P. Mulvaney, A. O. El-Ballouli, J. Pan, O. M. Bakr, A. K. Demeler, [58] U. Sakulkhu, M. Mahmoudi, L. Maurizi, G. Coullerez,
B. I. Hernandez Uribe, N. Bhattarai, R. L. Whetten, Anal. Chem. M. Hofmann-Amtenbrink, M. Vries, M. Motazacker, F. Rezaee,
2014, 86, 7688. H. Hofmann, Biomater. Sci. 2015, 3, 265.
[29] P. J. Wyatt, Anal. Chem. 2014, 86, 7171. [59] D. H. Jo, J. H. Kim, T. G. Lee, J. H. Kim, Nanomedicine 2015, 11,
[30] E. J. Cho, H. Holback, K. C. Liu, S. A. Abouelmagd, J. Park, Y. Yeo, 1603.
Mol. Pharmaceutics 2013, 10, 2093. [60] A. Baeza, D. Ruiz-Molina, M. Vallet-Regí, Expert Opin. Drug
[31] Q. Ong, Z. Luo, F. Stellacci, Acc. Chem. Res. 2017, 50, 1911. Delivery 2017, 14, 783.
[32] Y. Ma, in Toxicology of Nanomaterials, Wiley-VCH Verlag GmbH & [61] J. Liu, C. Wu, D. Xiao, P. Kopold, L. Gu, P. A. van Aken, J. Maier,
Co. KGaA (Eds: Y. Zhao, Z. Zhang, W. Feng), Weinheim, Germany Y. Yu, Small 2016, 12, 2354.
2016, pp. 1–21. [62] Y. Min, J. M. Caster, M. J. Eblan, A. Z. Wang, Chem. Rev. 2015, 115,
[33] C. Jiang, Y. Cao, G. Xiao, R. Zhu, Y. Lu, RSC Adv. 2017, 7, 7531. 11147.
[34] D. Lee, M. F. Rubner, R. E. Cohen, Nano Lett. 2006, 6, 2305. [63] H. Cabral, K. Miyata, K. Osada, K. Kataoka, Chem. Rev. 2018, 118,
[35] S. Kumar, R. Rani, N. Dilbaghi, K. Tankeshwar, K.-H. Kim, Chem. 6844.
Soc. Rev. 2017, 46, 158. [64] J. Shi, P. W. Kantoff, R. Wooster, O. C. Farokhzad, Nat. Rev. Cancer
[36] S. Zhang, J. Li, G. Lykotrafitis, G. Bao, S. Suresh, Adv. Mater. 2009, 2017, 17, 20.
21, 419. [65] A. Z. Wang, R. Langer, O. C. Farokhzad, Annu. Rev. Med. 2012, 63,
[37] H. H. Liu, S. Surawanvijit, R. Rallo, G. Orkoulas, Y. Cohen, Environ. 185.
Sci. Technol. 2011, 45, 9284. [66] N. Lee, D. Yoo, D. Ling, M. H. Cho, T. Hyeon, J. Cheon, Chem. Rev.
[38] S. G. Elci, Y. Jiang, B. Yan, S. T. Kim, K. Saha, D. F. Moyano, 2015, 115, 10637.
G. Yesilbag Tonga, L. C. Jackson, V. M. Rotello, R. W. Vachet, ACS [67] Z. Li, D. L. Clemens, B.-Y. Lee, B. J. Dillon, M. A. Horwitz, J. I. Zink,
Nano 2016, 10, 5536. ACS Nano 2015, 9, 10778.
[68] Z. Li, J. C. Barnes, A. Bosoy, J. F. Stoddart, J. I. Zink, Chem. Soc. [105] J. M. Costa-Fernández, M. Menéndez-Miranda, D. Bouzas-Ramos,
Rev. 2012, 41, 2590. J. R. Encinar, A. Sanz-Medel, TrAC Trends Anal. Chem. 2016, 84,
[69] D. B. Williams, C. B. Carter, Transmission Electron Microscopy, 139.
Springer US, Boston, MA 2009. [106] A. R. M. Bustos, J. R. Encinar, A. Sanz-Medel, Anal. Bioanal. Chem.
[70] A. Surrey, D. Pohl, L. Schultz, B. Rellinghaus, Nano Lett. 2012, 12, 2013, 405, 5637.
6071. [107] E. A. Kapellios, S. A. Pergantis, J. Anal. At. Spectrom. 2012, 27, 21.
[71] M. A. O’Keefe, Y. Shao-Horn, Microsc. Microanal. 2004, 10, 86. [108] S. Gschwind, L. Flamigni, J. Koch, O. Borovinskaya, S. Groh,
[72] U. Dahmen, R. Erni, V. Radmilovic, C. Ksielowski, M.-D. Rossell, K. Niemax, D. Günther, J. Anal. At. Spectrom. 2011, 26, 1166.
P. Denes, Philos. Trans. R. Soc., A 2009, 367, 3795. [109] A. Kasuya, R. Sivamohan, Y. A. Barnakov, I. M. Dmitruk,
[73] W. Zhou, H. F. Greer, Eur. J. Inorg. Chem. 2016, 2016, 941. T. Nirasawa, V. R. Romanyuk, V. Kumar, S. V. Mamykin,
[74] I. Srnová-Šloufová, F. Lednický, A. Gemperle, J. Gemperlová, Lang- K. Tohji, B. Jeyadevan, K. Shinoda, T. Kudo, O. Terasaki, Z. Liu,
muir 2000, 16, 9928. R. V. Belosludov, V. Sundararajan, Y. Kawazoe, Nat. Mater. 2004,
[75] M. T. Reetz, M. Maase, T. Schilling, B. Tesche, J. Phys. Chem. B 3, 99.
2000, 104, 8779. [110] S. Jeon, D. R. Oberreit, G. Van Schooneveld, C. J. Hogan, Analyst
[76] S. Mondini, A. M. Ferretti, A. Puglisi, A. Ponti, Nanoscale 2012, 4, 2016, 141, 1363.
5356. [111] S. Elzey, D.-H. Tsai, L. L. Yu, M. R. Winchester, M. E. Kelley,
[77] C. R. Murthy, B. Gao, A. R. Tao, G. Arya, Nanoscale 2015, 7, 9793. V. A. Hackley, Anal. Bioanal. Chem. 2013, 405, 2279.
[78] E. V. Orlova, H. R. Saibil, Chem. Rev. 2011, 111, 7710. [112] C. Degueldre, P.-Y. Favarger, Colloids Surf. A 2003, 217, 137.
[79] W. Anderson, D. Kozak, V. A. Coleman, Å. K. Jämting, M. Trau, [113] J. K. Navin, M. E. Grass, G. A. Somorjai, A. L. Marsh, Anal. Chem.
J. Colloid Interface Sci. 2013, 405, 322. 2009, 81, 6295.
[80] E. Nakamura, N. A. J. M. Sommerdijk, H. Zheng, Acc. Chem. Res. [114] S. Nicolardi, Y. E. M. van der Burgt, J. D. C. Codée, M. Wuhrer,
2017, 50, 1795. C. H. Hokke, F. Chiodo, ACS Nano 2017, 11, 8257.
[81] S. T. Skowron, T. W. Chamberlain, J. Biskupek, U. Kaiser, E. Besley, [115] H. Hinterwirth, S. Kappel, T. Waitz, T. Prohaska, W. Lindner,
A. N. Khlobystov, Acc. Chem. Res. 2017, 50, 1797. M. Lämmerhofer, ACS Nano 2013, 7, 1129.
[82] J. P. Patterson, Y. Xu, M.-A. Moradi, N. A. J. M. Sommerdijk, [116] N. Fernández-Iglesias, J. Bettmer, Nanoscale 2015, 7, 14324.
H. Friedrich, Acc. Chem. Res. 2017, 50, 1495. [117] F. Chen, G. Wang, J. I. Griffin, B. Brenneman, N. K. Banda,
[83] J. Goldstein, D. E. Newbury, J. R. Michael, N. W. M. Ritchie, V. M. Holers, D. S. Backos, L. Wu, S. M. Moghimi, D. Simberg,
J. H. J. Scott, D. C. Joy, Scanning Electron Microscopy and X-Ray Nat. Nanotechnol. 2017, 12, 387.
Microanalysis, Springer-Verlag, New York 2018. [118] M. Magro, M. Zaccarin, G. Miotto, L. Da Dalt, D. Baratella,
[84] S. Rades, V.-D. Hodoroaba, T. Salge, T. Wirth, M. P. Lobera, P. Fariselli, G. Gabai, F. Vianello, Anal. Bioanal. Chem. 2018, 410,
R. H. Labrador, K. Natte, T. Behnke, T. Gross, W. E. S. Unger, RSC 2949.
Adv. 2014, 4, 49577. [119] S. Chen, C. Xiong, H. Liu, Q. Wan, J. Hou, Q. He, A. Badu-Tawiah,
[85] E. Buhr, N. Senftleben, T. Klein, D. Bergmann, D. Gnieser, Z. Nie, Nat. Nanotechnol. 2015, 10, 176.
C. G. Frase, H. Bosse, Meas. Sci. Technol. 2009, 20, 084025. [120] D. P. Bishop, M. Grossgarten, D. Dietrich, A. Vennemann,
[86] S. Mourdikoudis, R. M. Pallares, N. T. K. Thanh, Nanoscale 2018, N. Cole, M. Sperling, M. Wiemann, P. A. Doble, U. Karst, Anal.
10, 12871. Methods 2018, 10, 836.
[87] D. J. Stokes, Microsc. Microanal. September 2012, 67. [121] K. S. W. Sing, Colloids Surf. 1989, 38, 113.
[88] P. Luo, I. Morrison, A. Dudkiewicz, K. Tiede, E. Boyes, P. O’Toole, [122] K. A. Cychosz, M. Thommes, Engineering 2018, 4, 559.
S. Park, A. B. Boxall, J. Microsc. 2013, 250, 32. [123] M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier,
[89] Y. Seo, W. Jhe, Rep. Prog. Phys. 2008, 71, 016101. F. Rodriguez-Reinoso, J. Rouquerol, K. S. W. Sing, Pure Appl.
[90] H.-J. Butt, B. Cappella, M. Kappl, Surf. Sci. Rep. 2005, 59, 1. Chem. 2015, 87, 1051.
[91] S. Morita, R. Wiesendanger, E. Meyer, Noncontact Atomic Force [124] A. J. Lecloux, J. Nanopart. Res. 2015, 17, 447.
Microscopy, Springer, Berlin 2002. [125] A. J. Lecloux, R. Atluri, Y. V. Kolen’ko, F. L. Deepak, Nanoscale
[92] M. Lanza, Conductive Atomic Force Microscopy: Applications in 2017, 9, 14952.
Nanomaterials, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, [126] E. A. Van Doren, P.-J. R. De Temmerman, M. Francisco, J. Mast,
Germany 2017. J. Nanobiotechnol. 2011, 9, 17.
[93] F. J. Giessibl, Rev. Mod. Phys. 2003, 75, 949. [127] W. Wohlleben, J. Mielke, A. Bianchin, A. Ghanem, H. Freiberger,
[94] U. Holzwarth, N. Gibson, Nat. Nanotechnol. 2011, 6, 534. H. Rauscher, M. Gemeinert, V.-D. Hodoroaba, J. Nanopart. Res.
[95] P. Scherrer, Nachr. Akad. Wiss. Goettingen, Math.-Phys. Kl. 1918, 2017, 19, 61.
1918, 98. [128] Scientific Committee on Emerging and Newly Identified Health
[96] H. Klug, L. Alexander, X-Ray Diffraction Procedures: For Risks, Scientific Basis for the Definition of the Term “Nanomaterial”,
Polycrystalline and Amorphous Materials, 2nd ed., John Wiley & European Commission, Brussels, Belgium 2010.
Sons, New York 1974. [129] W. G. Kreyling, M. Semmler-Behnke, Q. Chaudhry, Nano Today
[97] S. Britto, S. Joseph, P. Vishnu Kamath, J. Chem. Sci. 2010, 122, 751. 2010, 5, 165.
[98] P. Muhammed Shafi, A. Chandra Bose, AIP Adv. 2015, 5, 057137. [130] R. Pecora, J. Nanopart. Res. 2000, 2, 123.
[99] A. V. Vorontsov, S. V. Tsybulya, Ind. Eng. Chem. Res. 2018, 57, 2526. [131] D. E. Koppel, J. Chem. Phys. 1972, 57, 4814.
[100] I. Robinson, J. Phys. Soc. Jpn. 2013, 82, 021012. [132] I. D. Morrison, E. F. Grabowski, C. A. Herb, Langmuir 1985, 1,
[101] A. Cervellino, R. Frison, N. Masciocchi, A. Guagliardi, in X-Ray 496.
and Neutron Techniques for Nanomaterials Characterization [133] A. Scotti, W. Liu, J. S. Hyatt, E. S. Herman, H. S. Choi, J. W. Kim,
(Ed: C. Kumar), Springer, Berlin 2016, pp. 545–608. L. A. Lyon, U. Gasser, A. Fernandez-Nieves, J. Chem. Phys. 2015,
[102] T. Li, A. J. Senesi, B. Lee, Chem. Rev. 2016, 116, 11128. 142, 234905.
[103] A. G. Kikhney, D. I. Svergun, FEBS Lett. 2015, 589, 2570. [134] J. Stetefeld, S. A. McKenna, T. R. Patel, Biophys. Rev. 2016, 8, 409.
[104] H. Takeno, in X-Ray and Neutron Techniques for Nanomate- [135] F. Varenne, A. Makky, M. Gaucher-Delmas, F. Violleau, C. Vauthier,
rials Characterization (Ed: C. Kumar), Springer, Berlin 2016, Pharm. Res. 2016, 33, 1220.
pp. 717–760. [136] S. K. Brar, M. Verma, TrAC Trends Anal. Chem. 2011, 30, 4.
[137] H. Qian, M. P. Sheetz, E. L. Elson, Biophys. J. 1991, 60, 910. [176] S. F. Tan, U. Anand, U. Mirsaidov, ACS Nano 2017, 11, 1633.
[138] V. Filipe, A. Hawe, W. Jiskoot, Pharm. Res. 2010, 27, 796. [177] E. Nakamura, Acc. Chem. Res. 2017, 50, 1281.
[139] S. Bhattacharjee, J. Controlled Release 2016, 235, 337. [178] R. M. Gorgoll, E. Yücelen, A. Kumamoto, N. Shibata, K. Harano,
[140] B. R. Ware, W. H. Flygare, Chem. Phys. Lett. 1971, 12, 81. E. Nakamura, J. Am. Chem. Soc. 2015, 137, 3474.
[141] R. Xu, Particuology 2008, 6, 112. [179] M. Koshino, T. Tanaka, N. Solin, K. Suenaga, H. Isobe,
[142] R. P. Carney, J. Y. Kim, H. Qian, R. Jin, H. Mehenni, F. Stellacci, E. Nakamura, Science 2007, 316, 853.
O. M. Bakr, Nat. Commun. 2011, 2, 335. [180] R. Fernandez-Leiro, S. H. W. Scheres, Nature 2016, 537, 339.
[143] P. Schuck, H. Zhao, C. Brautigam, R. Ghirlando, Basic Principles of [181] P. L. Stewart, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol.
Analytical Ultracentrifugation, CRC Press, Boca raton, FL 2015. 2017, 9, e1417.
[144] Ž. Krpetić, A. M. Davidson, M. Volk, R. Lévy, M. Brust, [182] T. Gonen, Y. Cheng, P. Sliz, Y. Hiroaki, Y. Fujiyoshi, S. C. Harrison,
D. L. Cooper, ACS Nano 2013, 7, 8881. T. Walz, Nature 2005, 438, 633.
[145] A. Bekdemir, F. Stellacci, Nat. Commun. 2016, 7, 13121. [183] Y. Cheng, Science 2018, 361, 876.
[146] J. Walter, K. Löhr, E. Karabudak, W. Reis, J. Mikhael, W. Peukert, [184] A. Wittemann, M. Drechsler, Y. Talmon, M. Ballauff, J. Am. Chem.
W. Wohlleben, H. Cölfen, ACS Nano 2014, 8, 8871. Soc. 2005, 127, 9688.
[147] L. Pitkänen, A. M. Striegel, TrAC Trends Anal. Chem. 2016, 80, 311. [185] J. Radon, Math. Phys. Kl. 1917, 69, 262.
[148] A. Podgornik, S. Yamamoto, M. Peterka, N. L. Krajnc, J. Chroma- [186] Z. Saghi, P. A. Midgley, Annu. Rev. Mater. Res. 2012, 42, 59.
togr. B 2013, 927, 80. [187] J. Miao, P. Ercius, S. J. L. Billinge, Science 2016, 353, aaf2157.
[149] G.-T. Wei, F.-K. Liu, C. R. C. Wang, Anal. Chem. 1999, 71, 2085. [188] C. Bonnaud, C. A. Monnier, D. Demurtas, C. Jud, D. Vanhecke,
[150] K. M. Krueger, A. M. Al-Somali, J. C. Falkner, V. L. Colvin, Anal. X. Montet, R. Hovius, M. Lattuada, B. Rothen-Rutishauser,
Chem. 2005, 77, 3511. A. Petri-Fink, ACS Nano 2014, 8, 3451.
[151] Z. Aspanut, T. Yamada, L. W. Lim, T. Takeuchi, Anal. Bioanal. [189] D. Armitage, F. P. Price, Chem. Phys. Lett. 1976, 44, 305.
Chem. 2008, 391, 353. [190] C. L. Jackson, G. B. McKenna, J. Chem. Phys. 1990, 93, 9002.
[152] S. Süß, C. Metzger, C. Damm, D. Segets, W. Peukert, Powder [191] G. Fagerlund, Mater. Constr. 1973, 6, 215.
Technol. 2018, 339, 264. [192] M. Brun, A. Lallemand, J.-F. Quinson, C. Eyraud, Thermochim.
[153] J. D. Robertson, L. Rizzello, M. Avila-Olias, J. Gaitzsch, C. Contini, Acta 1977, 21, 59.
M. S. Magoń, S. A. Renshaw, G. Battaglia, Sci. Rep. 2016, 6, 27494. [193] J. H. Strange, M. Rahman, E. G. Smith, Phys. Rev. Lett. 1993, 71, 3589.
[154] J. C. Giddings, Science 1993, 260, 1456. [194] J. Mitchell, J. B. W. Webber, J. H. Strange, Phys. Rep. 2008, 461, 1.
[155] G. Moreno-Martin, J. Sanz-Landaluze, Y. Madrid, in Encyclopedia of [195] J. Park, A. Byun, D.-H. Kim, S. S. Shin, J.-H. Kim, J. W. Kim, J. Col-
Analytical Chemistry (Ed: R. A. Meyers), John Wiley & Sons, Ltd, loid Interface Sci. 2014, 426, 162.
Chichester, UK 2017, pp. 1–24. [196] N. Rescignano, M. Amelia, A. Credi, J. M. Kenny, I. Armentano,
[156] S. K. R. Williams, J. R. Runyon, A. A. Ashames, Anal. Chem. 2011, Eur. Polym. J. 2012, 48, 1152.
83, 634. [197] N. Gopinathan, B. Yang, J. P. Lowe, K. J. Edler, S. P. Rigby, Int. J.
[157] K.-G. Wahlund, J. Chromatogr. A 2013, 1287, 97. Pharm. 2014, 469, 146.
[158] K. Bacia, P. Schwille, Methods 2003, 29, 74. [198] S. J. Sahl, S. W. Hell, S. Jakobs, Nat. Rev. Mol. Cell Biol. 2017, 18,
[159] K. Bacia, S. A. Kim, P. Schwille, Nat. Methods 2006, 3, 83. 685.
[160] N. Pal, S. D. Verma, M. K. Singh, S. Sen, Anal. Chem. 2011, 83, [199] D. Wöll, C. Flors, Small Methods 2017, 1, 1700191.
7736. [200] J. J. Richardson, H. Ejima, S. L. Lörcher, K. Liang, P. Senn, J. Cui,
[161] K. Starchev, J. Zhang, J. Buffle, J. Colloid Interface Sci. 1998, 203, 189. F. Caruso, Angew. Chem. 2013, 125, 6583.
[162] A. Zimpel, N. Al Danaf, B. Steinborn, J. Kuhn, M. Hoehn, T. Bauer, [201] A. P. H. Gelissen, A. Oppermann, T. Caumanns, P. Hebbeker,
P. Hirschle, W. Schrimpf, H. Engelke, E. Wagner, M. Barz, S. K. Turnhoff, R. Tiwari, S. Eisold, U. Simon, Y. Lu, J. Mayer,
D. C. Lamb, U. Lächelt, S. Wuttke, ACS Nano 2019, 13, 3884. W. Richtering, A. Walther, D. Wöll, Nano Lett. 2016, 16, 7295.
[163] G. R. Willmott, Anal. Chem. 2018, 90, 2987. [202] G. M. Conley, S. Nöjd, M. Braibanti, P. Schurtenberger,
[164] A. K. Pal, I. Aalaei, S. Gadde, P. Gaines, D. Schmidt, F. Scheffold, Colloids Surf. A 2016, 499, 18.
P. Demokritou, D. Bello, ACS Nano 2014, 8, 9003. [203] A. M. Clemments, P. Botella, C. C. Landry, J. Am. Chem. Soc. 2017,
[165] E. Weatherall, G. R. Willmott, Analyst 2015, 140, 3318. 139, 3978.
[166] M. Platt, G. R. Willmott, G. U. Lee, Small 2012, 8, 2436. [204] B. Rühle, M. Davies, T. Bein, C. Bräuchle, Z. Naturforsch., B: J.
[167] M. J. Williamson, R. M. Tromp, P. M. Vereecken, R. Hull, Chem. Sci. 2013, 68, 423.
F. M. Ross, Nat. Mater. 2003, 2, 532. [205] T. Lebold, J. Michaelis, C. Bräuchle, Phys. Chem. Chem. Phys. 2011,
[168] S. F. Tan, S. W. Chee, G. Lin, U. Mirsaidov, Acc. Chem. Res. 2017, 13, 5017.
50, 1303. [206] A. Zürner, J. Kirstein, M. Döblinger, C. Bräuchle, T. Bein, Nature
[169] H. Zheng, R. K. Smith, Y.-w. Jun, C. Kisielowski, U. Dahmen, 2007, 450, 705.
A. P. Alivisatos, Science 2009, 324, 1309. [207] B. Rühle, M. Davies, T. Lebold, C. Bräuchle, T. Bein, ACS Nano
[170] N. de Jonge, L. Houben, R. E. Dunin-Borkowski, F. M. Ross, Nat. 2012, 6, 1948.
Rev. Mater. 2019, 4, 61. [208] M. Davies, B. Rühle, C. Li, K. Müllen, T. Bein, C. Bräuchle, J. Phys.
[171] L. R. Parent, E. Bakalis, M. Proetto, Y. Li, C. Park, F. Zerbetto, Chem. C 2014, 118, 24013.
N. C. Gianneschi, Acc. Chem. Res. 2018, 51, 3. [209] F. Feil, V. Cauda, T. Bein, C. Bräuchle, Nano Lett. 2012, 12, 1354.
[172] N. D. Loh, S. Sen, M. Bosman, S. F. Tan, J. Zhong, [210] T. Lebold, L. A. Mühlstein, J. Blechinger, M. Riederer,
C. A. Nijhuis, P. Král, P. Matsudaira, U. Mirsaidov, Nat. Chem. H. Amenitsch, R. Köhn, K. Peneva, K. Müllen, J. Michaelis,
2017, 9, 77. C. Bräuchle, T. Bein, Chem. — Eur. J. 2009, 15, 1661.
[173] U. Anand, J. Lu, D. Loh, Z. Aabdin, U. Mirsaidov, Nano Lett. 2016, [211] T. Lebold, A. Schlossbauer, K. Schneider, L. Schermelleh,
16, 786. H. Leonhardt, T. Bein, C. Bräuchle, Adv. Funct. Mater. 2012, 22,
[174] G. Lin, S. W. Chee, S. Raj, P. Král, U. Mirsaidov, ACS Nano 2016, 106.
10, 7443. [212] E. Sage, M. Sansa, S. Fostner, M. Defoort, M. Gély, A. K. Naik,
[175] Q. Chen, H. Cho, K. Manthiram, M. Yoshida, X. Ye, A. P. Alivisatos, R. Morel, L. Duraffourg, M. L. Roukes, T. Alava, G. Jourdan, E. Colinet,
ACS Cent. Sci. 2015, 1, 33. C. Masselon, A. Brenac, S. Hentz, Nat. Commun. 2018, 9, 3283.
[213] J. E. Sader, M. S. Hanay, A. P. Neumann, M. L. Roukes, Nano Lett. [216] M. M. Modena, Y. Wang, D. Riedel, T. P. Burg, Lab Chip 2014, 14, 342.
2018, 18, 1608. [217] M. M. Modena, P. Hirschle, S. Wuttke, T. P. Burg, Small 2018, 14,
[214] T. P. Burg, M. Godin, S. M. Knudsen, W. Shen, G. Carlson, 1800826.
J. S. Foster, K. Babcock, S. R. Manalis, Nature 2007, 446, 1066. [218] A. Cox, P. Andreozzi, R. Dal Magro, F. Fiordaliso, A. Corbelli,
[215] S. Olcum, N. Cermak, S. C. Wasserman, K. S. Christine, L. Talamini, C. Chinello, F. Raimondo, F. Magni, M. Tringali,
H. Atsumi, K. R. Payer, W. Shen, J. Lee, A. M. Belcher, S. N. Bhatia, S. Krol, P. Jacob Silva, F. Stellacci, M. Masserini, F. Re, ACS Nano
S. R. Manalis, Proc. Natl. Acad. Sci. USA 2014, 111, 1310. 2018, 12, 7292.

Nanoparticle Characterization What To Measure

Uploaded by

Copyright:

Available Formats

Nanoparticle Characterization What To Measure

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nanoparticle Characterization What To Measure

Uploaded by

Copyright:

Available Formats

15214095, 2019, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.201901556 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/10/2022].

Nanoparticle Characterization: What to Measure?

yet, the characterization methods themselves can directly affect

sedimentation-based methods are commonly employed for the

Figure 3. Surface charge and zeta potential of nanoparticles in suspen-

3.1.1. Transmission Electron Microscopy (TEM)

Transmission electron microscopy is undoubtedly one of the

Table 1. Summary of characterization methods for dry nanoparticles.

Technique NPa) state Parameters Advantages Limitations Reviews

a)NP stands for nanoparticle.

SEM typically requires conductive substrates for high-resolu-

3.1.3. Atomic Force Microscopy (AFM)

Atomic force microscopy is a scanning probe microscopy tech-

3.1.4. X-Ray Diffraction (XRD) 3.1.5. Small-Angle X-Ray Scattering (SAXS)

Dynamic light scattering estimates the particle size from the

Table 2. Summary of characterization methods for nanoparticles in suspension.

Technique NPa) state Parameters Advantages Limitations Reviews

a)NP stands for nanoparticle.

compositions render DLS extremely suitable for routine and kBT

3.3.1. In Situ Liquid-Cell (LC) and Atomic-Resolution (AR) TEM

Table 3. Summary of advanced characterization methods.

Technique NPa) state Parameters Advantages Limitations Reviews

a)NP stands for nanoparticle.

reconstruction), and a high level of complexity (especially the

3.3.4. NMR Cryoporometry and DSC Thermoporosimetry

Liquids confined within nanometer-sized pores experience a

the application include the investigation and

3.3.7. Nanomechanical Resonators

Micromechanical resonators present a

© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2019, 31, 1901556

You might also like