Analysis and toxicity of nanoparticles: Fate after human exposure

(1)

Analytical Sciences

Literature Thesis

Analysis and toxicity of

nanoparticles

Fate after human exposure

by

Lex Hendriks

12201421

May 2020

12 ECTS

February 2020 – May 2020

Daily Supervisor:

First Examiner:

Second Examiner:

(2)

Vrije Universiteit Amsterdam Faculty of Science

Amsterdam Institute of Molecules, Medicines and Systems Division of BioAnalytical Chemistry

(3)

Abstract

Many modern-day products, such as clothing, cosmetics, food, and packaging contain nanoparticles (NPs). However, the possible (adverse) effects on human health are mostly not well understood yet. This review therefore gives an update on the current fate of NPs on human health, and summarizes the most common methodologies to analyse NPs.

Even though many research towards NP toxicity has been performed, most of the studies are focused towards metal and metal oxide NPs. It is therefore recommended to do more research on the toxicity of the non-metal NPs as well. Especially toxicity of nanoplastics should be investigated.

The properties of NPs and thus their effects on human health strongly depend on certain factors, including particle size, size distribution, and shape. For this reason, the characteristics of NPs are first determined using mainly microscopy and spectroscopic techniques.

Based on the current knowledge on NPs, it is possible to predict the site of accumulation as well as the excretion mechanism; as the body tries to metabolize the particles to facilitate excretion, the main organs of accumulation seem to be liver and spleen. The filtration-size threshold in the kidneys is 8 nm. Metal (oxide) NPs that are smaller than this threshold can be filtered out of the blood in the kidneys, causing them to end up in the urine. Bigger particles will end up in the faeces. The size limit for non-metal NPs is not determined. For the measurement in human tissues and body fluids, immunoassays, all types of separation techniques, and ICP-MS are often used. The method chosen is based on the compatibility of the technique with the analyte particles, as well as the matrix. After human uptake, proteins may bind to the NPs and form a protein corona. Measuring these particles using ELISA or affinity chromatography could give valuable information. In addition to ICP-MS, other mass spectrometric methods can give more information about the NPs.

Many research to the toxicity of metal (oxide) NPs has been performed on tissues and cells. However, the fate of NPs in body fluids is still barely covered. It is recommended to do more research on this part of toxicology. In addition, the possibility of excretion via sweat or saliva has not been determined yet. This should also be investigated.

(4)

List of Abbreviations

AAS atomic absorption spectroscopy

AES atomic emission spectroscopy

ADME absorption, distribution, metabolism, excretion – the four main steps in toxicology

AF4 asymmetric flow field-flow fractionation

AFM atomic force microscope/microscopy

ANUC analytical ultracentrifugation

CAD charged aerosol detector

CE capillary electrophoresis

CLSM confocal laser scanning microscopy

CF(U)F cross flow (ultra)filtration

DDSN dye-doped silica nanoparticle

DLS dynamic light scattering

DTA differential thermo analysis

EELS electron energy loss spectroscopy

ELISA enzyme-linked immunosorbent assay

EM electron microscope/microscopy

ESI electrospray ionisation

FCS fluorescence correlation spectroscopy

FFF field-flow fractionation

FTIR Fourier-transform infrared spectroscopy

HDC hydrodynamic chromatography

ICP inductively coupled plasma

ISO International Organization for Standardization

IUPAC International Union of Pure and Applied Chemistry

LC liquid chromatography

LIBD laser-induced breakdown detection

LIF laser-induced fluorescence

LOD limit of detection

MALDI matrix-assisted laser desorption ionisation

MS mass spectrometry

MDR multiple drug resistance / multidrug resistance

NMR nuclear magnetic resonance

NP nanoparticle

NSOM near-field scanning optical microscopy

PEG polyethylene glycol

PBS phosphate-buffered saline

PVP polyvinylpyrrolidone

QD quantum dot

RFU relative fluorescence unit

ROS reactive oxygen species

RPLC reversed phase liquid chromatography

(5)

SAED selected area electron diffraction

SANS small-angle neutron scattering

SAXS small-angle X-ray scattering

SCENIHR Scientific Committee on Emerging and Newly Identified Health Risks

SDS-PAGE sodium dodecyl sulphate polyacrylamide gel electrophoresis

SEC size exclusion chromatography

SEM scanning electron microscope/microscopy

SLS static light scattering

TEM transmission electron microscope/microscopy

TFFF thermal field-flow fractionation

TGA thermogravimetric analysis

TOFMS time-of-flight mass spectrometry

UC ultracentrifugation

UF ultrafiltration

UFP ultrafine particle

UV-Vis ultraviolet-visible spectroscopy

XAS X-ray absorption spectroscopy

XPS X-ray photoelectron spectroscopy

XRD X-ray diffraction

XRF X-ray fluorescence spectroscopy

(6)

1. Introduction

Although there is a recent increase in attention towards nanoparticles (NPs), the concept is not new. In the ninth century, Mesopotamian artisans already used NPs to give glittering effects to the surface of pottery [1]. In addition, pottery from the Middle Ages often still retain a distinct metallic glitter, called

lustre, when found. This lustre contains copper and silver NPs which are homogeneously dispersed in the

glassy ceramic glaze. The lustring technique originates from the Islamic world, because Muslims were not allowed to use gold in their crafts [1]. They had to find an alternative method to produce a similar glittering effect, without using real gold. The first scientific article about NPs is written by Michael Faraday in 1857 [2]. George Beilby and Thomas Turner expanded Faraday’s work in the early 1900s [3, 4]. After these articles, not much progress was made until the invention of the scanning tunnelling microscope, at the early 1980s. This invention lead to the discovery of both fullerenes and carbon nanotubes [1]. At that time, these particles were generally called ultrafine particles (UFP) [5, 6]. During the 1990s, the word

nanoparticle (NP) became more common to use instead of UFP [7-10]. In Figure 1 a document analysis

from Web of Science is shown. Here it is clear that research into UFPs began around the early 1980s. The term NP, however, came into use around the mid-1990s. In this figure, it is also clear that around the time NP was used instead of UFP, research started to increase significantly.

As the name already implies, a NP is a particle with dimensions in the nanometre-range [10, 11]. Although in terminology “nano” is anything below the micron range, some scientists define a lower upper limit to NPs, usually being particles < 100 nm [1, 9, 11-13]. This is due to the fact that below this limit, the NPs have some special physical properties. However, others argue that since the limit for these properties is compound dependent, the definition should contain everything up to 1000 nm [10]. According to the International Organization for Standardization (ISO), a structure is considered a nanoparticle if all three dimensions are in the nanometre range [13]. If only two dimensions are in this range, it is defined as a nanofibre, and if it is only one dimension, it is a nanoplate.

A special class of NPs are the particles with a diameter of a few nm, called Quantum Dots (QDs) [1, 10, 14]. QDs are in definition smaller than the Bohr radius [15]. To prevent any confusion, the most commonly accepted upper limit of 100 nm [11, 14, 16] will be used in this review as being NPs. Particles in the nanometre range, but above this limit, will be considered UFP, and particles below 10 nm will be named QDs. Table 1 gives an overview of the definitions used in this thesis.

Table 1 - The relative sizes of the particles as used in this thesis. These limits are based on common limits in literature. However, since there is not a universal agreement about the limits, the definition of these terms could differ between articles.

Lower limit Upper limit Term used in thesis Abbreviation

1 nm 10 nm quantum dot QD

10 nm 100 nm nanoparticle NP

100 nm 1000 nm ultrafine particle UFP

> 1000 nm Bulk

Figure 1 - Number of articles published about NPs. These graphs are based on data obtained from Web of Science, and contain the number of articles published with the term “ultrafine particle” (A), or “nanoparticle” (B). The same analysis is performed with Scopus and PubMed (results not shown), with similar trends observed. In these graphs it is also clear that research into NPs is still increasing each year.

(8)

NPs can be further divided into different classes, based on their characteristics or shape. Most NPs are classed by their shape, such as nanoboxes [17], nanocapsules [18], nanocages [19] or nanospheres [10, 11]. These shapes are often based on shapes found in everyday life, and are intrinsic to their function [13]. Unfortunately, not all NPs have a well-defined shape [16]. For these NPs, ‘shape factors’, such as

sphericity or convexity, are devised. However, the definitions of these shape factors are not standardized

[16], which increases the difficulty in NP classification. Similar to a size distribution, NPs can also have a “shape distribution” [16]. This makes classification of NPs even more difficult.

NPs are also classed based on their characteristics. One kind of NP is the nanocomposite, defined as a composite in the nanometre range [20]. According to the International Union of Pure and Applied Chemistry (IUPAC), a composite is a “multi component material comprising multiple, different

(non-gaseous) phase domains in which one phase domain is a continuous phase” [20]. Another definition is

that nanocomposites are hybrid materials at the nanometre scale [21]. Another frequently used NP is the core-shell NP [8]. This is a NP consisting of a core, and one or more shells around it [13]. In this case, the entire particle, meaning the core diameter plus the shell thickness, has to be in the nanoscale to be considered a core-shell NP. If the core is spherical, and it has multiple concentric shells, the NP is defined as a nano-onion [13]. The advantage of a core-shell NP could be that the shell, when it is an inert material of a few nm thick, can protect the core from oxidation [22]. A lot of articles also mention nanocrystals [9, 10, 12, 23]. A nanocrystal is a NP that has a crystalline structure [13]. Another way to define nanocrystals is as nanoscale crystals [1]. Even though nanocrystals have some good properties, it is quite challenging to predict what agents in what combination leads to a certain size and shape of crystal [24].

The two main synthesis routes for NPs are top-down and bottom-up [25], which strongly influence the size distribution of the formed NPs. Since top-down synthesis usually is performed using physical techniques, the synthesis is relatively hard to control, leading to NPs with a wide size distribution [25]. On the other hand, in bottom-up synthesis NPs are formed through chemical and biological techniques, where size distributions, as well as morphology and possible surface defects depend on the method used [25]. On average, NPs synthesized with a chemical technique have a narrower size distribution, and less surface defects [25]. Natural events, such as a volcanic eruption, sea spray, or erosion, may also create and release NPs [12]. Silica NPs are popular due to their simple synthesis method, and their hydrophilic character [8]. In addition, silica can be easily (bio)conjugated, making it an interesting NP for functionalization [8]. For biological targeting studies, it is also possible to conjugate proteins to the surface of NPs, as showed by Åkerman et al. [15].

NPs have some differences compared to their bulk materials, which mainly causes their special properties [14, 25]. Faraday performed a method similar to the lustring technique on different metals and tested the effects on the reflection, absorption and transmission of light [2]. This was the first documented examination of the difference in physicochemical properties of NPs. One of these differences is the relative larger surface area of NPs [10, 14, 25]. Buzea et al. illustrate this by comparing a carbon microparticle with a diameter of 60 μm with a carbon NP with a diameter of 60 nm [10]. The carbon microparticle has a surface area of 0.01 mm2_{, whereas the NP has a surface area of 11.3 mm}2_{, which is a} 1000-fold increase. Since a larger surface area provides more surface for chemical reactions, this means that reactivity increases roughly a 1000-fold as well [10]. Another difference between NPs and their bulk is the quantum effects in NPs, which appear more strongly in QDs [10, 14]. The electronic behaviour of QDs is similar to those of atoms and small molecules. Several unpaired electronic spins in particles of only a few hundred atoms give QDs a magnetic moment, even if the same molecule is non-magnetic in

(9)

bulk [10]. Due to their quantum confinement, QDs also have a different electron affinity compared to the bulk [10]. Suspensions of NPs, often referred to as nanofluids, have a higher effective thermal conductivity compared to the common coolants [26]. This good heat transfer coefficient make nanofluids a suitable use as high performance heat transfer fluid. For example, the heat transfer coefficient of water goes up when gallic acid-treated graphene nanoplatelets were added, while the viscosity did not change significantly [26]. Organic NPs have some properties which make them applicable for gene therapy [27]. These properties include protection of target genes against degradation, specific targeting of tissues or cells, improvement of DNA stability, and enhancement of transformation efficiency.

The special properties of NPs make them very popular in various applications. For example, silver NPs have antibacterial, antiviral and antifungal properties, which makes it one of the most widely applied NPs [28]. ZnO can reflect, scatter, and absorb UV radiation, which is why it is added to cosmetics and sunscreens [23]. The applications of NPs are not mainly limited to metal-based particles, as applications of fullerenes and nanoclays have been reported as well [14]. Other examples include silica-based NPs as drug delivery systems [29]. Additionally, a lot of NPs are produced unintended. There are much more examples about the application of NPs. For clarity, a summary of these applications is given in Table 2. This table clearly shows that NPs are used practically everywhere.

Table 2 - Overview of the published applications of various NPs, together with their respective sources. The column “Type of NP used” lists a few examples as named in literature (if any). These particles are, however, not the only applicable/used NPs.

Application Type of NP used Source

Agriculture * [14]

Antibiotics ZnO, CuO, Au [1, 30, 31]

Antimicrobial agents

Ag [1, 14]

Bioassays Ru(II)- and Ir(III)-complexes, DDSN, polymer-coated nanospheres [1, 8]

Biological imaging metal NPs [1, 25]

Catalysis metal and metal oxide NPs [17, 25, 32]

Clothing nano-fibres [14, 33]

Cosmetics fullerene C60, silica, ZnO [14, 34] Diagnostics fluorescent QDs, hybrid organic/inorganic NPs [1, 15, 35]

Diffusion barriers nano-clay [14]

Drug delivery liposome NPs, hybrid organic/inorganic NPs, Au [1, 15, 31]

Electronics metal NPs, ZnO [25, 33, 36]

Engineering * [14]

Environmental protection

zero-valent iron [14, 32]

Food essential elements (Cu, Zn, and Mn), ZnO [1, 14, 36, 37]

Fuel additives Ce, CeO2 [38]

Gene therapy organic NPs [1, 27, 35]

Information storage metal NPs [17]

Medicine protein-based, polysaccharide-based, lipid-based NPs, Au [1, 14, 31, 39]

(10)

Photography metal NPs [17] * No type of NP mentioned for this application.

The field of science studying the nature, adverse effects, detection and treatment of poisons on organisms is called toxicology [10]. Toxicology has as fundamental principle that everything is poisonous as long as the concentration is high enough. Since there are many different substances, with different properties, the field of toxicology is further divided into smaller fields focused around a certain class of poisons. The toxicology of NPs is often referred to as nanotoxicology [38]. In the past decade, a lot of diseases have been associated with interference of cellular functions that appear to be correlated with exposure to NPs [10]. These diseases include autoimmune diseases, Crohn’s, Parkinson’s and Alzheimer’s diseases, but also various types of cancer. However, even though there seems to be a correlation between NP exposure and these diseases, the effects of NPs on human health are not well understood. Despite this lack of understanding of possible adverse health effects, the amount of NPs used in various applications is increasing rapidly [28] Therefore, it is important to get a clear understanding about the possible toxicity of NPs. Furthermore, the possible health risks of NPs in cosmetics are still widely unknown [1, 40].

In order to assess nanotoxicology, it is necessary to be able to detect and characterize NPs in a great variety of matrices [14]. Access to different, robust analytical methodologies is essential. Since nanotoxicology studies the interactions between NPs and biological materials [38], these analytical techniques have been widely applied on biological samples, such as tissues and body fluids, as well. The general aim of this thesis is twofold; a toxicological aim as well as an analytical one:

First, this thesis aims to obtain an overview of the nature and toxicity of NPs that enter the human body, as well as the current knowledge on the mechanisms of absorption, distribution, metabolism, and excretion (ADME). In addition, this review will also collect the published analytical techniques and methodologies used to characterize and study NPs in human tissues and body fluids.

(11)

2. Nature and toxicity of NPs

NPs have different properties compared to their bulk materials [14, 25]. Metal oxides that are considered nontoxic, such as ZnO, show toxic effects when the particles are at nanoscale [23]. There is even evidence that NPs that are not directly toxic, may alter the functionality of immune cells, thereby causing a great health risk altogether [32]. These different properties could be the reason some materials show toxic effects when the particles are in the nano-range. Since it is important to take these differences into account when testing a compound for toxicity, the physicochemical properties as well as their influence on toxicity will be discussed in this chapter.

2.1. Physicochemical properties and toxicity of NPs

Unlike the bulk, the intrinsic properties of metal NPs seem to depend on size, structure, composition, crystallinity, and shape [17]. This subsequently means that the effects of NPs on human health strongly depend on these factors as well, which makes the comparison of toxicological data difficult. Especially since not all research articles describe these factors. This could lead to pseudo-inconsistencies; some articles seem to have different conclusions, because the research involved NPs of different morphologies. The first observations about the effects of small particles on light were described by Michael Faraday in 1857 [2]. In this experiment, Faraday describes a visible trend that finer particles tend to absorb and transmit more light, while reflecting less, compared to coarse particles [2]. Faraday also observed a shift in the light Au particles transmitted; going from blue for the largest particles to ruby for the smallest particles [2]. Using Faraday’s experiments as a basis, the research into NPs started at the beginning of the twentieth century [3, 4]. In 1903, George Beilby performed the synthesis methods described by Faraday, and through the use of microscopy discovered the metals were reduced to small, spherical aggregates [3]. Three years later, professor Thomas Turner extended these results by observing that, unlike Au films, Ag and Cu films only could be reduced in the presence of oxygen [4]. In 1976, Granqvist et al. concluded that the GE theory describing electromagnetic properties of small metallic particles was inconsistent for particles with diameters below 10 nm [5]. These observations indicate that NPs have some different properties than their bulk compounds. The optical properties of metal NPs depend on their size, shape and its surrounding medium [14, 25], which explain the findings of Faraday in his article. Furthermore, metal NPs are able to enhance fluorescence, known as metal-enhanced fluorescence [39]. Interactions between fluorophores and metal NPs result in a greater photostability, enhanced fluorescence, and decreased lifetimes caused by increase in system radiation decay. These spectral properties may not necessarily influence toxic responses, but it is important as it does influence spectroscopic analysis techniques.

The different properties of NPs relative to their bulk form could give some information about the cause of nanotoxicity. While nanotoxicology is a relative new field of toxicology, it is confirmed that one important property of NPs causing toxicity is their increased reactivity, caused by a high surface area [23]. Another important property of NPs in respect with toxicity is size [41]. This is mainly due to the fact that the size of a NP determines if and how it will be taken up by a cell. Since the size of NPs is in the same range as for instance proteins and viruses, it is possible that NPs can enter cells as well [42]. On the other hand, research has also shown that NPs do not readily enter cells and need help of certain mechanisms that transports them across the cell membrane [42].

(12)

NPs have particle sizes that are lognormally distributed [7]. This size-distribution of NPs is not only dependent on the synthesis method, but also on the state of dispersion [16]. In nanoparticulate systems, dispersion is the extent to which the particles are agglomerated. Agglomeration is the formation of clusters, caused by interparticle forces. The most important forces between NPs are van der Waals forces [16]. The degree of agglomeration is strongly related to size in nanotoxicology studies [41]. Additionally, research has indicated that for some kinds of NPs, such as polystyrene [35], toxicity depends on the shape of the NP [41]. Toxicity of NPs also seem to depend on crystal structure [41]. An example is TiO2. TiO2 has a variety of crystal structures, although the rutile, anatase, and brookite forms are most extensively researched. In one such experiment, anatase TiO2 NPs turned out to be roughly 100 times more toxic than the rutile form [41].

The characteristics of NPs can also play a major role in allergies. Allergens, which in most cases are proteins, are hypothesized to bind to NPs, which in turn may change the overall structure of the allergen [32]. Binding to NPs can therefore both enhance or reduce allergic reactions to these allergens.

The toxicity of NPs also depends on the route of exposure [42]. Whereas most toxicological studies are focussed around inhalation and dermal uptake, the few studies performed on oral uptake suggest that NPs do not enter the human body as well as via inhalation. However, the amount of available data is not enough to draw definite conclusions [42]. There is, however, some evidence that indicates the mechanism of uptake seems to influence the toxic effects observed [43]. Research has indicated that NPs can be translocated from the nasal cavity to the brain [43-45]. This effect, however, was not observed in other routes of exposure.

2.2. Examples of toxic NPs

Even though there is no clear summary of NPs that are proven to be toxic, Zoruddu et al. stated that both Ag and TiO2, which both are considered nontoxic/inert in bulk form, are extremely toxic in NP form [38]. Au NPs with size below 13 nm are considered toxic as well, with the particles below 2 nm even showing an unexpected degree of toxicity [41]. Additionally, recent experiments have indicated that polystyrene NPs can penetrate cellular membranes and interfere with cellular processes [46]. On top of that, polystyrene NPs are not tested for carcinogenicity, while its monomer styrene is proven to be carcinogenic.

In fact, even though the toxicity of metal NPs has practically been proven [33, 38, 47], it is very likely that all NPs are toxic. This is in agreement with the basic principle of toxicology, that states that everything will become toxic if the dose is high enough [10]. To conclude: even though metal NPs are proven to be toxic, there is not enough toxicity data to claim non-metal NPs are not toxic. It is, at this stage, therefore too soon to divide NPs as “toxic” or “non-toxic”.

(13)

2.3. Testing NP toxicity

Since NPs could be toxic, it is necessary to perform tests to determine how toxic a NP is. A challenging aspect in NP toxicity testing is the variability in manufacturing methods [42]. It is difficult to keep all NP properties consistent, to enable reliable cross comparisons. It requires the use of a classification scheme, as the toxicological differences between spherical ZnS-capped CdSe QDs and “nail shaped” QDs were evident [42]. There is a strong need for consistency in this front before reliable comparisons can be made.

Another challenging aspect in NP toxicity is the correlation between properties [42]. Most research is focussed on testing the effect of only one property. Most often the effect of either size or shape is assessed, but the effect of changing both size and shape simultaneously is not tested.

The challenging aspects described above clearly show it is important to have guidelines in toxicity testing. The majority of research in nanotoxicology is performed as cell culture studies [41]. However, since research has proven that certain NPs may interfere or react with one of the assays, it is recommended to perform different assays instead of just one [41]. An important feature to assess the toxicity of NPs, is measuring its degree of cellular uptake [41]. Some experiments have indicated that the mechanism of cellular uptake can also influence NP toxicity [41].

According to the Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR), to assess health risks caused by NPs, toxicological testing should be performed [48]. Based on available nanotoxicity data, the SCENIHR has recommended for NPs to test at least [48]:

- Blood cell damage

- Acute phase responses of hepatocytes or lung cells - Permeability tests of endothelial cells

- Effects on the autonomic nervous system

- Immune activation by T cell activity and cytokine induction in lymph nodes - Induction of oxidative stress

- Induction of cellular DNA damage - Toxicity on various cell lines in vitro - Biopersistance

This list of recommendations already indicates that a single “toxicity test” does not exist. Furthermore, since this list contains recommendations, based on the already examined toxic effects of NPs, this list will likely change as the knowledge in nanotoxicology progresses. In addition, after exposure NPs can be metabolized [42]. Since the physicochemical properties and thus the toxicity of NPs are mainly related on size and shape, it is difficult to predict possible effects of the metabolites of NPs. It is therefore necessary to perform in vivo tests as well [42].

(14)

3. Fate of NPs in the human body

Ag NPs have antimicrobial properties. Another frequently used type of NP is ZnO, which light scattering properties makes it common in many sunscreens. ZnO also has some antibiotic properties. For both Ag and ZnO NPs, however, there is evidence that it also has toxic effects. In addition, the antibiotic properties of for instance ZnO can also mean it targets human cells. Toxicity testing therefore is required. For these reasons, a summary of published articles about toxic effect of NPs will be given in this chapter. All steps in the ADME will be discussed separately.

3.1. Absorption pathways

According to Greulich et al., exposure of Ag NPs can occur through the skin, inhalation, injection, and ingestion [33]. This means that all common absorption pathways can introduce NPs into the human body. Of these pathways, inhalation is considered the most important route of exposure, due to the lungs being in constant contact with the environment through air [47].

Even though the skin is a defensive barrier and should prevent particles from penetration further than the hair follicles, several NPs are reported to have crossed the derma and reached the vascular system through the lymph nodes [38]. Furthermore, medication that is used to treat skin burns, which is coated with Ag NPs also proves NPs can be absorbed through the skin [38]. That is, the Ag NPs that are released from this coating have been related to liver malfunctioning. It has also been reported that Ag NPs are able to penetrate the skin through the pores of both sweat glands and sebaceous glands [49].

Uptake from ingestion is the rarest uptake pathway for NPs. The only NPs that enter this way are the intentional uptakes, both as food supplements and as medicine, and the contaminations in food and drinking water [38]. From the gastro intestinal tract these particles are absorbed into the bloodstream. Just like the other routes of uptake, the amount that enters the bloodstream depends on the particle size; small particles enter the bloodstream more easily compared to larger NPs. Especially the particles between 150 nm and 500 nm are reported in literature to be absorbed [38]. Negatively charged particles can cross the mucus layers of the intestine more easily than positively charged ones, as the mucus layer is negatively charged, causing the positive charged NPs to be trapped here [38].

Some recent research has shown that NPs can cross the placental barrier as well, leading to exposure and possible health effects to the foetus [40].

On cellular level, due to the small size, NPs are able to cross cell membranes [1, 14, 33]. NPs can enter through passive diffusion, membrane disruption, non-specific endocytosis, and receptor-mediated endocytosis [35]. These uptake mechanisms seem to be dependent on the size, charge, and functionalisation of the surface of the particles [33, 38]. Shape of the NPs also influences its uptake, as 20 nm polystyrene nanospheres could cross artificial membranes, but nanodisks of the same material and size are shown to bind to the membrane [35]. By making computational simulations, Thake et al. could conclude that the transport of polystyrene NPs across a membrane seems to depend on the composition of the membrane as well [35]. Thus, in all these mechanisms, nanoparticle-membrane interactions play a role. However, it is currently not known how these interactions relate to the translocation into the cell [35].

(15)

Some NPs have the tendency to cause inflammatory responses in lung cells [47, 50]. An example is asbestos. Since some kinds of asbestos particles are too big to enter a lysosome, they persist in the lungs where they induce an inflammatory response [38]. This could result in the development of lung cancer [38]. The same effects as with asbestos has been reported with both TiO2 and Ni NPs [38]. Furthermore, exposure to NPs in air causes a significant increase in blood pressure in children [50]. It has also been reported that both single- and multiwall carbon nanotubes activated cytokine and lymphokine productions and increase oxidative stress after crossing cellular membranes [38].

3.2. Distribution

Metal NPs that are directly injected into the bloodstream or inhaled, are distributed and end up in all organs, especially in kidneys, liver, spleen, brain, and heart [33, 38, 47]. The liver, as the metabolic centre of the human body, stores the highest concentrations compared to other organs [47]. Following ingestion, CeCl3 caused the highest cerium levels in the bones and spleen of rats [38]. In mice, the cerium levels were increased in lung, liver, spleen, and kidneys.

Manganese NPs can move from the nasal cavity through the axons of certain nerve cells to the brain, causing a severe progressive damage to the central nervous system [38]. The same transport, with the lungs as starting point, has been reported for MnO2 NPs [38]. These effects have been reported for professional groups at risk of occupational exposure to metal NPs, such as manganese mining or welders [38].

There is proof that certain NPs accumulate in cells [14, 51]. Positively-charged gold NPs are proven to enter the kidneys, while negatively-charged gold NPs remained in the liver and spleen [1]. This shows that charge influences the excretion pathways of NPs. However, the particles can be compartmentalized in peripheral tissues and therefore accumulate in the body over time. There is also evidence of ZnO NP accumulation in the brain and testis, indicating that some NPs can easily cross the brain and blood-testis barriers [36].

In their research on gold NPs, Lopez-Chaves et al. found that Au NPs with particle size of 60 nm did not show a significant increase in DNA damage [31]. However, NPs of 30 nm, as well as QDs with size of 10 nm showed a significant increase in DNA damage. Since the pores in the cell nucleus are about 35 nm in diameter, this effect was probably due to the QDs and 30 nm NPs being able to enter the nucleus, whereas the 60 nm NPs could not. This observation is confirmed with TEM, where the 10 nm QDs were clearly accumulated within the nucleus of liver cells [31]. Furthermore, the storage of the particles changed according to their size: 10 nm accumulated mostly in intestine, while 30 nm accumulated more in spleen, followed by liver. In contrast, 60 nm particles accumulated more strongly in spleen, followed by intestine [31].

Kim et al. found that ZnO NPs tend to distribute to the eyes of rats, where they cause the death of retinal cells [34]. The mechanism that transports these NPs to the eyes is still unknown, though. In addition to the eyes, ZnO is shown to distribute into the liver, kidneys, spleen, lungs, and heart as well [34].

The distribution of NPs after absorption is summarized in Table 3. In this table, it becomes evident that the research towards distribution is mainly focused on metal-based NPs. The information towards distribution of non-metal NPs is limited and requires more research.

(16)

Table 3 – Summary of distribution of NPs. All the information from this section, regarding type of NP, its target and the effect, are summarized in this table.

NP Type Target Effect

Au

10 nm Intestine cell nucleus DNA damage 30 nm Spleen cell nucleus

Liver cell nucleus DNA damage

60 nm Spleen

Intestine Positively charged Kidneys Negatively charged Liver

Spleen ZnO

Brain Testis

Eyes Retinal cell death

Metal Directly injected

Liver Kidneys Spleen Brain Heart CeCl3 Rat bone Rat spleen Mouse lung Mouse liver Mouse kidney Mouse spleen Cerium accumulation

Mn Via nasal cavity

Brain Damage to Central Nervous System MnO2 Via lungs

In case of toxicity testing, it is good to know where to expect certain effects. Since the most reliable tests rely on human test subjects or animal studies, it is not very ethical to test all organs if the target is known; especially if the target organ can be removed or analysed without killing the tested animal. The information from the research as described above suggests that most NPs tend to accumulate in the liver, spleen, and kidneys. In addition, some NPs can be accumulated in the brain or heart. Since the function of the liver is to detoxify and metabolize compounds, it is understandable to find an accumulation of NPs there. As for the kidneys, their function is to filter unnecessary “waste” compounds out of the blood and excrete it via the urine. If the NPs are too big to pass through this filter, they can get stuck in the kidneys, hence the accumulation. The other blood filter, the spleen, may accumulate NPs for the same reason as with the kidneys. These three organs, where accumulation of all NPs seems to take place, are explained. It also makes sense for larger NPs that are inhaled to be found in the lungs, because they get stuck in the alveoli of the lungs. According to several studies, smaller NPs have a bigger tendency to agglomerate [52]. This agglomeration is likely the reason NPs tend to get stuck in the alveoli.

Even though the organs of accumulation can be explained, most of this research is focused towards metal and metal oxide NPs. The accumulation of non-metal NPs does not seem to be examined yet. However, it is crucial for these particles to have some toxicity data as well, because one of the factors where the

(17)

physicochemical properties of NPs depend on, is composition [17]. This ultimately means that there are, to some extent, differences in properties between metal and non-metal NPs. These differences in properties can change reactivity and thus toxic effects of NPs completely. In this case, metal NPs that are smaller than 30 nm can pass the membrane of the cell nucleus and are able to accumulate in there, able to cause DNA damage. However, since non-metal NPs have different properties, this size cut-off can be completely different. If this cut-off is higher for non-metal NPs, meaning bigger particles can enter the cell nucleus, it is possible for these particles to cause mutations even though the particles are bigger than the “safe” size of 30 nm.

On top of this, since the liver functions to metabolize compounds to facilitate excretion, the difference in properties between metal and non-metal NPs is crucial here as well. If metal NPs accumulate in the liver, this indicates that they are not or slowly metabolized or degraded. However, non-metal NPs may be easier to metabolize and therefore do not accumulate as strongly as metals do.

To conclude, the behaviour of non-metal NPs considering distribution within the body and accumulation in certain tissues/organs can be completely different compared to that of metal (oxide) NPs. Therefore the research into the distribution of non-metal NPs should be more extensively studied.

3.3. Metabolism

Mineral and synthetic NPs are coated with proteins from the biological fluids when entering the human body [32, 53]. This protein corona influences the distribution as well as the effects of NPs on the human body. For example, the proteins in the protein corona can be detected by the immune system, leading to various (auto)immune reactions [32]. The main compounds that bind to these NPs seemed to be saturated free fatty acids, lipids, sugars, and amino acids, and DNA likely binds to these NPs as well [53]. Although it has been proven that metal NPs are able to form a protein corona [32], some research indicates this is also true for polystyrene NPs [53]. The proteins bind to the surface of the NPs via weak interactions such as hydrophobic and electrostatic interactions [32]. However, the binding of proteins to the aforementioned NPs is considered an irreversible process. Even though some mechanisms have been suggested, it is currently not known how these weak interactions exactly cause an irreversible corona formation [32]. If a protein corona forms around a NP, this changes the mechanism through which this particle is taken up into the cell [35]. For example, a protein corona changes size, surface charge, aggregation behaviour, and more.

It is known for metal and metal oxide NPs to generate reactive oxygen species (ROS) [38, 47, 54]. This can occur either by a chemical reaction directly at the surface of the NP, intracellularly, or indirectly through processes caused by an immune response [47]. ROS are able to impair the antioxidant defence system and thereby can cause oxidative stress or even cytotoxicity. Furthermore, ROS play a role in multiple diseases, such as failure in the immune system and cancer [34]. Research has also indicated that the ROS production seems to be dependent on both time and dose [31]. When the ROS production increases, the cells increase the antioxidant activity in response, which is most likely the reason for the time dependent trend. This claim is supported by Lopez-Chaves et al., who observed an increase in ROS activity 16 hours after exposure to Au NPs [31]. However, 32 hours after exposure, the ROS activity was only slightly increased.

Heim et al. showed that non-coated ZnO NPs tend to dissolve and be internalized into the cells as Zn2+ [54]. These cations accumulate into the cell nucleus, where they damage the DNA (double strand breaks).

(18)

response, which causes the induction of cytokine expression [23]. Au NPs seem to be dissolved and neutralized as well, although afterwards they are stored in lipid drops in order to attempt to reduce the toxicity in the liver [31]. These Au NPs are most likely broken down to smaller particles in the lysosomes of the macrophages [31]. Jenkins et al. reported that nanoroses (nanoclusters consisting of gold-coated iron) are degraded in the macrophages over 14 days as well [55].

Research by Gosens et al. indicate that metal and metal oxide NPs tend to agglomerate to some extent, in the presence of serum [47]. However, whereas this is true ex vivo, observations suggest this is not the case neither in vivo nor in vitro [31].

Manoharan et al. developed a 5-nitroindole-capped ZnO/CuO nanocomposite to test its potential antibiotic effects on multi drug resistant (MDR) and persistent bacteria [30]. The difference between MDR and persistent bacteria is that persistent bacteria become “dormant” during antibiotic treatment; they cannot reproduce, but survive. MDR bacteria are unaffected by the antibiotic and can reproduce as well. During this experiment, toxicity on both rat and human cells was tested. In both cases there were no toxic effects observed at concentrations up to 1 mM, but cells that were exposed to 2 mM of these NPs showed a significant cytotoxicity, and the morphology of the cells changed completely [30]. Even though Manoharan et al. concluded that the produced nanocomposites were non-toxic at concentrations up to 1 mM, this might not be entirely correct. This is because in their research article, the claims are made that these nanocomposites decrease the bacteria motility, which is regulated by flagella movement [30]. This could mean that the movement of the flagella is somehow affected, and since mammalian sperm cells also have a flagellum, this could mean these NPs decrease sperm quality, and in the worst case cause infertility. This claim is supported by Moridian et al., who showed that ZnO NPs lead to the decrease of testis weight and volume, as well as decrease in testosterone concentration, which both decrease sperm quality [36]. Moridian et al. showed that at least 50 mg/kg ZnO NPs cause a reduction in the weight and volume of testis in mice [36]. It was also clear that the Leydig cells, responsible for testosterone production and consequently testosterone concentrations decreased as well. All these changes lead to a reduced fertility. Another claim in the research of Manoharan et al. is that metal oxides (such as the used ZnO/CuO nanocomposite) can pass cell membranes and damage DNA [30]. However, the human toxicity tests performed do not include carcinogenicity or damage to sperm quality. Since these effects are not documented, and therefore probably not tested, it is not definite to conclude these nanocomposites are non-toxic at a concentration of 1 mM.

In most articles described in this section the following effects occurred. Metal and metal oxide NPs seem to induce the formation of ROS. In addition to that, some NPs, especially the larger metal (oxide) ones that get stuck in the kidneys or lungs, tend to induce an inflammatory response. On top of that, particles that are smaller than 35 nm apparently are able to enter the cell nucleus and most likely will interact with the DNA, leading to DNA damage. These three toxic effects are the most common responses to NPs and could be a good starting point in assessing nanotoxicity. However, these effects are mainly focused on metal and metal oxide NPs. The data on the non-metal NPs is limited, and therefore still requires attention. Even though there is evidence that polystyrene NPs can form a protein corona, there is no definite proof. On top of that, other effects of these NPs are not extensively researched either.

(19)

3.4. Excretion

Jenkins et al. found that Au NPs mainly distribute to liver and spleen up to 1 day after injection [55]. However, 31 days after injection, the recovery from organs decreased to about 10%, indicating that the gold is excreted between these two measurements. Although the excretion pathway is not determined in this experiment, the hypothesis is that particles with sizes up to 8 nm are preferentially excreted via renal clearance. Other researchers agree that particles with sizes below this renal filtration limit are preferentially excreted through the urine [31]. This hypothesis is supported by research to TiO2 NP and Ag NPs, which are excreted in the urine [56, 57].

Most NPs, however, are too big for this excretion. Particles that are bigger than this limit are unable to pass the glomerular capillary wall, hence it is more likely for those particles to be excreted into the bile, and subsequently eliminated in the faeces [55]. It suggested that gold NPs with size above the filtration limit are filtered out of the blood by the reticuloendothelial system, causing them to accumulate in liver and spleen [31]. This suggestion is widely accepted, as concentrations of larger NPs are proven to accumulate in these organs. This makes faecal excretion the main endogenous route of elimination for gold NPs [31].

Even though QDs are theoretically able to be excreted via the urine, there has been little research into this excretion pathway. Additionally, there has been even less, if any, attention towards excretion via the sweat or saliva. The smallest type of sweat glands, the eccrine glands, have a dermal duct of 10 – 20 μm and a secretory tubule of 30 – 40 μm in diameter [58]. It may therefore be possible for NPs to be excreted in sweat. However, according to Wilke et al., there has been close to no research performed towards the mechanism of sweating [58].

As for saliva it may be possible as well. Although no or little research has been done regarding NP excreting in saliva, the salivary glands might be large enough for NPs to cross. This is because one of the components of saliva, amylase, is an enzyme with a mass of 56 kDa [59]. If the salivary glands are able to excrete such a big protein, it is likely a relatively small NP can be excreted by this gland as well. Even though there is no toxicological report considering NP excretion in saliva, Supiandi et al. published a method to detect metal NPs in artificial saliva with a low LOD [60]. This method used isotopically labelled CdSe and ZnS QDs to reduce the background noise. With this approach, the QDs could be detected at the ppt (parts per trillion) level. Even though this method did not assess the excretion of NPs in saliva, it hypothesizes that it could be possible, and simultaneously offers a sensitive method to measure excretion in saliva.

Both saliva and sweat are aqueous matrices. Reverse-phase liquid chromatography (RPLC) should be able to measure NPs in these matrices, if present. In that case it is necessary to remove the salts and proteins from the samples. Since saliva only contains amylase, and sweat only contains salts, a simple sample preparation such as an SPE method should be enough to clean up the matrices before measurement. As reference material sweat or saliva that do not contain NPs can be used. The difference between those samples, would be the NPs, if present, which then would show up as one or multiple extra peak(s). To characterize the NPs, electron microscopy can be used.

(20)

4. Analytical techniques for the characterization of NPs

and their analysis in biological samples

Since the field of nano-research has increased significantly over the past two decades, there are many published articles about the analysis of NPs. To get a good understanding of NPs, and to control the synthesis if needed, it is necessary to be able to characterize NPs. In addition, since the absorption into the human body and consequently the effects depend on these properties, it is important in toxicological studies to start with characterizing the particles [38]. This is not easy, though, as sampling of NPs is more difficult compared to sampling the bulk [16]. The properties that characterize NPs involve size, size distribution, morphology, bulk chemistry, solubility, surface area, and other relevant physicochemical properties [16]. In toxicological studies, the NPs of interest are often synthesized personally. The techniques that will be discussed in this chapter are therefore used to characterize the NPs after synthesis, but also to measure these particles in biological samples.

4.1. Microscopy

Most articles involved with NPs start characterizing using a microscope [1, 14, 16]. Microscopy is used for determination of both size and shape [16]. Since different methods are available, all with their own strengths and weaknesses, it is more reliable to obtain particle size by combining the data of multiple methods [16]. To determine shape, no other techniques are used. In most cases, the particles are too small to observe with an optical microscope, but a confocal laser scanning microscope (CLSM) and near-field scanning optical microscopy (NSOM) are two optical microscopy techniques that are able to achieve proper resolution [14]. In CLSM a laser is focused at one point of the sample holder, which excites a fluorescent sample [61]. The fluoresced light is then reflected, and passes through a pinhole which is placed in the conjugate focal point. This pinhole filters the light which is out of focus, reducing background. Theoretically, an optical microscope has a resolution of λ/2 [62]. NSOM tackles this limitation by creating an aperture a few tenths of nm from the surface, the near-field. With this technique, the spatial resolution approaches the diameter of the aperture, which makes resolutions below 100 nm possible. However, the resolution is still somewhat lacking compared to other microscopes [62].

Determining the degree of dispersion is considered difficult, since agglomeration depends on various factors, such as agglomerate size, strength of the interparticle forces, number of particles, and distribution of agglomerates in the system [16]. To estimate the state of dispersion, however, mostly microscopy is used. Due to the ability to not only visualise the NPs, but also see certain properties such as state of aggregation, shape and size, electron microscopy (EM) [1, 5, 30], and atomic force microscopy (AFM) [1, 8, 14, 16] are most often used.

EM can be divided in two main techniques; scanning electron microscopy (SEM) and transmission electron microscopy (TEM) [14]. In TEM, electrons are transmitted through a sample, which therefore has to be very thin [14]. In SEM, however, an image is made by detecting electrons that are scattered at the sample interface [14]. Although this method can give much information about NPs, EM is more difficult for lighter atoms, because the electrons scatter less efficiently [14]. In addition, EM is a destructive method, and due to the electron irradiation, EM also causes charging effects [14]. An example of characterization of CeO2 NPs using TEM is shown in Figure 2.

(21)

Figure 2 – TEM image showing CeO2 nanorods (NR), nanocrystals (NC), nanoparticles (NP), and submicron particles (SMP). This

image was used in the original research [63] to characterize the different nanostructured CeO2 catalysts they synthesized.

Reprinted by permission from Copyright Clearance Center: Elsevier, Applied Catalysis B: Environmental, “Direct copolymerization of carbon dioxide and 1,4-butanediol enhanced by ceria nanorod catalyst”, Z-J Gong et al.

Using AFM has the advantage that, in contrast to EM, AFM can provide a 3D image [16]. In AFM, an oscillating cantilever scans the specimen surface, and measures the electrostatic forces between the tip and the surface [14]. With AFM, 3D imaging is possible, as the force measurements have a height resolution of approximately 0.5 nm. Furthermore, AFM can scan sub-nanometre structures under moist or wet conditions, something that is not possible with EM, which works under a vacuum [14]. A drawback of AFM, however, is that the tip is often larger than the NPs, leading to a relative big error. This could lead to an overestimation of the size of the NPs, if not used with caution. The error in the z-direction (height) is not influenced by the geometry of the tip, though. The measurements on the height are therefore rather accurate [14].

To obtain a spatial resolution image of an aqueous sample without sample pre-treatment, X-ray microscopy (XRM) can be used [14]. With wavelengths between 2.34 and 4.43 nm, X-rays have a much shorter wavelength than visible light. Since the resolution in microscopy depends on the wavelength of the radiation source, this means that XRM has a better resolution than a light microscope [62]. The X-ray absorption of iron oxides, clay, organic matter, and particles is higher than water, giving XRM a good contrast for imaging in aqueous samples [62]. This removes the need for drying, fixating, staining, etc. of the sample, preventing a lot of artefacts.

(22)

Table 4 – The most commonly used microscopic techniques for the characterization of NPs are summarized here, together with their applicable size range, advantages, disadvantages, and the properties that can be determined with that technique. The data from this table are combined from [14, 16, 61, 64].

Technique Size Range Properties Advantages Disadvantages

CLSM > 200 nm - Size - Shape - 3D imaging - Low background - Needs fluorescent sample - Poor resolution - Slow AFM 0.5 nm – few μm - Size - Electrical and mechanical properties - Structure - Aggregation state - Good resolution - Provides 3D image

- Dry, moist and liquid samples - Ambient environment - Fast - Little sample pretreatment required

- Can only show surface - Overestimations - Artefacts due to moving particles - Particles adhere to the tip TEM 0.5nm – 1mm - Size - Shape - Structure - Visualization - High resolution - High vacuum - Difficult sample preparation - Time consuming - Expensive - Requires thin samples

- Destructive - Causes charging

effects

SEM 1 nm – 1 μm - Size - High resolution

- High vacuum - Sample preparation is difficult - Charging effects - Time consuming - Expensive - Destructive NSOM > 30 nm - Size - Chemical bonding - Visualization - Optical imaging

- Requires thin samples - Poor spatial resolution XRM > 30 nm - Size - Shape - Visualization - Good resolution - Works with aqueous samples - Very little sample

pretreatment

(23)

Microscopy can also be used for specific analyses [31, 33, 34, 47, 54, 55]. The advantage of using microscopy in toxicology, is that it can visualize the specific accumulation of NPs in cells and tissues. This helps to understand in what parts of the cell certain NPs tend to concentrate.

An example is the research performed by Lopez-Chaves et al. [31], where TEM was used to observe the movement of Au NPs in colon cells. In this in vitro study, it was evident that these NPs translocated form outside the cells to the cytosol in 4 hours after treatment. Additionally, 16 hours after treatment, the Au NPs were translocated to the lipid drops and cell nucleus.

To assess the intracellular distribution of Ag NPs in human stem cells, Greulich et al. combined phase contrast microscopy and fluorescence microscopy [33]. In this experiment, where NPs with a hydrodynamic diameter of 80 nm were used, the microscopic techniques confirmed that the NPs could enter the cells. After cellular uptake, the NPs tended to agglomerate in the endo-lysosomal areas. No accumulation in cell nucleus, Golgi complex, or endoplasmic reticulum was observed.

In an in vitro study towards the genotoxic effects of ZnO NPs, alveolar epithelial-like cells were treated with ZnO NP after which intracellular deposition was evaluated using EM [54]. After 15 or 30 minutes of exposure, no internalized NPs were observed. However, after 1 hour of exposure, agglomerated NPs were observed in the endosomes, and single NPs were found in the cytosol and cell nucleus.

Another example is the article written by Jenkins et al. [55], who did research to the degradation of gold-iron nanoclusters. During this experiment, TEM showed that these NPs were present in the macrophages of mice, leading to the hypothesis that these macrophages engulf and digest the NPs.

However, microscopy can also be used to observe changes induced by NPs. This is illustrated by an experiment performed by Kim et al. [34]. In this experiment microscopy was used to assess damage to the retina of rats treated with ZnO NPs. In addition, when using a immunohistochemical staining, the damage to the retinal neurones could also be observed.

4.2. Spectroscopy

Due to their strong absorbing and light scattering properties, spectroscopic techniques are frequently used in addition to microscopy. In fact, some techniques even use a microscope. For example electron energy loss spectroscopy (EELS) uses the loss of the energy of the incident electron to determine the elemental composition, whereas selected area electron diffraction (SAED) provides information about the crystalline structure [14]. Both EELS and SAED can only be used in combination with TEM.

UV-Vis spectroscopy [1, 8, 25, 30], fluorescence spectroscopy [1, 8], Fourier-transform infrared spectroscopy (FTIR) [1, 30], laser diffraction spectroscopy [16], small angle scattering [14, 16], Raman spectroscopy [14, 39], and dynamic light scattering (DLS) [1, 8, 16] are also used. Fluorescent species can also be detected with CLSM coupled with fluorescence correlation spectroscopy (FCS) [14].

DLS can be used to determine particle size as well as state of aggregation of a system [14]. Even though DLS can provide real-time in situ sizing, it is very sensitive and can easily be influenced by artefacts such as dust particles [14]. On the other hand, static light scattering (SLS) can give information about the structure of NPs [14]. A disadvantage of light scattering in general is that it relies on single scattering events [61]. This means that for turbid samples, where multiple scattering steps happen, no information about structure can be obtained. If SLS is combined with either DLS or field-flow fractionation (FFF), it

(24)

Small angle neutron scattering (SANS) and small angle X-ray scattering (SAXS) can be performed on both solids and liquids [14]. In SAXS, if the sample is monodisperse, it is possible to determine structure, shape, and size of the particles [14]. However, if the sample is polydisperse, it is only possible to obtain information about size distribution. On top of that, SAXS is not applicable for organic compounds, as C, H, and O do not scatter X-rays very well [65].

Laser-based spectroscopy techniques, such as Raman spectroscopy, laser-induced breakdown detection (LIBD), and laser-induced fluorescence (LIF) are widely used in characterizing NPs as well [14]. LIF is in principle the same as common fluorescence, but in this case a tuneable laser is used as excitation source [66]. In LIBD, NPs are irradiated with an intense pulsed laser, which generates a plasma [67]. The light this plasma emits, is detected. The breakdown threshold is lower for solids than for liquids, so a laser energy is chosen that ensures breakdown of the NP whilst keeping the medium intact. The spatially detection of the emitted light yields a spatial distribution, which is size dependent [67]. LIBD is capable of analysing the size and concentration of colloids. It has a low detection limit (LOD), which makes it a promising technique in NP characterization [14]. However, it requires particle-specific calibration, because it cannot distinguish different types of particles. In addition, since the technique involves creating a plasma, it is a destructive method.

UV-Vis and infrared spectroscopy are mainly useful for characterizing QDs and organic-based NPs, such as fullerenes [14]. Nuclear magnetic resonance spectroscopy (NMR) can be used to obtain information about the 3D structure of NPs [8, 14]. On top of that, NMR can also be used to characterize size and interactions of colloidal matter [14].

Spectroscopic techniques using X-rays are also widely applied in the field of NPs [14]. X-ray spectroscopy is a class of different techniques, including X-ray photoelectron spectroscopy (XPS), X-ray fluorescence spectroscopy (XRF), X-ray absorption spectroscopy (XAS), and X-ray diffraction (XRD) [1, 14]. These techniques are non-destructive and can provide information about elemental composition, concentration, or crystallographic structures [14]. However, X-ray spectroscopy is sensitive to sample perturbations and has a poor resolution [64, 68]. An example of a powder XRD analysis on different concentrations of Sm3+_{doped ZnS NPs is shown in Figure 3A. The peaks show the angle of diffraction of} the incident X-rays, which enables structure determination. In this case the peaks do not shift at increasing Sm3+_{concentration, suggesting that these ions are added homogeneously [69]. Peak shifts are} indicative for a change in surface morphology. Using the Scherrer’s formula, an average particle size can be obtained from the peak width (the formula is explained in reference [69]). Figure 3B shows the FTIR spectra of the same Sm3+_{doped ZnS NPs. These spectra clearly show that all particles in this case have} the same composition. Molecular structure can be elucidated, and possible contaminations can be visualized in these spectra. Especially when one spectrum differs from the other spectra.

(25)

Figure 3 – Characterization of Sm3+_{doped ZnS NPs at different ratios. A: powder XRD patterns. B: FTIR spectra. Reprinted by}

permission from Copyright Clearance Center: Elsevier, Journal of Luminescence, “Synthesis, characterization and photoluminescence studies of samarium doped zinc sulfide nanophosphors”, K. Ashwini et al.

A summary of the spectroscopic techniques mentioned in this section can be found in .

Table 5 – Most common spectroscopic techniques for characterization of NPs. The applicable size range, properties that can be determined, advantages and disadvantages of each technique is summarized here. The data are obtained from [14, 16, 61, 62, 64, 67, 68].

Technique Size Range Properties Advantages Disadvantages

EELS - Sample composition - Structure - Gives a lot of information - Needs TEM to work - Slow - Destructive - Complex data interpretation UV-Vis - Characterizing QDs and

organic NPs

- In-situ

measurements - Low sensitivity FTIR - Characterizing QDs and

organic NPs - Non destructive Laser diffraction spectroscopy 40 nm – 3 mm - Easy to use - Broad size range

- Valid for dry or suspended samples

- Requires refractive index in the

nano-range - Only valid for spherical particles Raman - Oxidation state - Structure - Size - Compatible with aqueous suspensions - Compatible with - Parameter effects

(26)

DLS 0.7 nm – 7 μm

- Particle size - State of aggregation

- Good for narrow size distributions - In-situ measurements - Rapid - Simple - Sensitive to contaminants - Difficult to interpret - Limited on polydisperse samples SLS 50 nm – few μm - Structure determination

- Does not work on turbid samples SANS 1 nm – 1 μm - Charge density - Structure (pH, ionic strength, solute concentration dependent) - Can be used on both solids and

liquids

SAXS 1 nm – 1 μm

- Monodisperse: Structure, shape, size

- Polydisperse: Size distribution

- Can be used on both solids and

liquids - Requires high concentrations - Poor contrast in organic compounds LIBD 5 nm – few μm

- Colloid size and concentration - Low LOD - No artefacts - Requires particle-specific calibration - No elemental information - Destructive NMR - Obtain 3D structure - Elemental composition - Hydrodynamic diameter - Suitable for colloids in solid or liquid state - Lack of standards XRD 0.5 nm – few μm - Structure - Size - High signal-to-noise ratio - Needs crystal sample XAS - Oxidation state - Elemental composition - Structure - Sample can be solid, liquid or gas

- Specific

- Does not work with thick samples

- Does not work with low concentrations XRF - Morphology - Isotope ratios - Quantitative bulk analysis - Non-destructive

- Easy sample prep - Requires vacuum

XPS - Shape - Size - Elemental composition - Oxidation state - Atomic composition of layers of 1 – 10 nm - Non-destructive - Sensitive - Poor lateral resolution

The antioxidant glutathione exists in the oxidized and reduced form [70]. The reduced form of glutathione is able to reduce and detoxify reactive oxygen species (ROS), converting to the oxidized form. A change in glutathione levels is therefore indicative for an oxidative impact on the liver [47]. For this

(27)

reason, glutathione levels in the liver are sometimes analysed to indicate NP toxicity. Another way to detect the presence of ROS is by using 2,7-dichlorofluorescein diacetate [31, 54]. This non-fluorescent compound is converted to the fluorescent 2’,7’—dichlorofluorescein in the presence of ROS. This conversion, and thus ROS generation can be measured fluorometrically. To do this, Cancino-Bernardi et al. added human serum albumin coated gold NPs to cell medium and incubated for 2 hours [71]. Afterwards, they were rinsed with phosphate-buffered saline (PBS), and 2,7-dichlorofluoroscein diacetate was added and incubated for 30 minutes. The excess was removed by rinsing with PBS and fluorescence was measured. With excitation of 485 nm and emission of 530 nm, the fluorescence was normalized against untreated cells, and plotted as relative fluorescence units (RFU). The results are shown in Figure

4. It is clear in this figure that in both fibroblasts (F C3H) as the carcinoma cells (HTC) no dependence of

NP concentration (25, 75, or 125 μg/mL) on ROS production is visible. In fact, there is even no significant difference between both cell types, leading to the conclusion that the Au NPs did not cause an adverse effect.

Figure 4 – Fluorescence results in ROS testing, according to [71]. The graphs contain the spectroscopic data of Au NP and nanorods (AuNR), with or without human serum albumin (HSA) for a fibroblast (F C3H) and a carcinoma cell (HTC), at 25, 75 and 125 μg/mL.. Reprinted by permission from Copyright Clearance Center: Elsevier, Chemosphere, “Gold-based nanospheres and nanorods particles used as theranostic agents: An in vitro and in vivo toxicology studies”,Cancino-Bernardi, J. et al.

To measure the cellular concentrations of metal or metal oxide NPs, atomic absorption spectroscopy (AAS) can be performed [54]. In this procedure, first the number of cells are counted, washed with PBS, and lysed. During centrifugation the cytoplasmic fraction forms the supernatant, and the nuclear fraction forms the pellet. These fractions are separated, the nuclear fraction is resuspended in an extraction buffer, and finally both separate fractions are diluted in distilled water. These samples are dissolved in nitric acid, and measured. The measured concentrations are then divided by the counted number of cells to get absolute concentrations in cytoplasm and the nucleus of a single cell [54].

Although most research is focussed on the accumulation of NPs in organs, recently there also has been research to NPs in amniotic fluid [40]. In this research, metal NPs were extracted from the amniotic fluid using a dispersion buffer. After incubation for 20 minutes, glycerol was added, and the mixture was centrifuged for 3.5 hours. The supernatant, containing the metal NPs was measured using inductively coupled plasma atomic emission spectroscopy (ICP-AES).

Analysis and toxicity of nanoparticles: Fate after human exposure

Analytical Sciences

Literature Thesis