Introduction

Over the last few years there has been a strong increase in interest in nanotechnology and the application of nanoparticles (NPs) in the food, cosmetics, and pharmaceutical industries. Because of their nanometre size, NPs have specific characteristics as regards transfer across tissue barriers and can be engineered for subsequent release of bioactive compounds at a desired target tissue. However, many uncertainties exist and issues regarding safety assessment and tissue targeting remain to be resolved (reviewed elsewhere [1]). Generally, two groups of NPs can be distinguished—inorganic and organic. Inorganic NPs contain only inorganic components, e.g. silica, gold, silver, titanium, and others. Organic NPs are mostly assemblies composed of organic molecules and can be further classified on the basis of the chemical nature of the building blocks: lipid, protein, and carbohydrate [2, 3].

Until now, inorganic NPs have received much more attention in analytical and safety research than organic NPs, especially because the latter consist of building blocks that are generally regarded as safe (GRAS). Hence, the development of strategies for chemical characterisation of inorganic NPs in food and other sources is much further developed [4]. As far as we are aware, analytical strategies for quantification of organic NPs are not available at present and certainly not validated [3]. A challenge in investigating organic NPs is the similar chemical nature to the food and tissue matrices in which they are incorporated. Another aspect which should be considered is the tendency of organic NPs to disassemble in the course of sample preparation in organic solvents or under the conditions in most analytical chemical techniques, e.g. MALDI–TOF MS and LC–MS–MS. Thus, to preserve the assembled state, sample preparation strategies and subsequent separation steps should be mild The upper molecular weight limit of the above-mentioned analytical chemical techniques does not enable characterisation of intact organic NPs with molecular masses of the order of megadaltons [5], but it is possible to identify the building blocks.

In this study we investigated a two-step concept as a general strategy for characterisation of organic NPs. In the first step, organic NPs were separated from the matrix in their native state on the basis of size in the nanometre range by use of hydrodynamic chromatography (HDC). In the second step, the building blocks from the separated NPs were characterized by use of MALDI–TOF mass spectrometry. In this study we investigated the validity of this two-step concept with commercially available liposomes, suspended in a relatively simple food matrix—a commercial orange-flavoured beverage. In the EU-financed consortium NanoLyse this matrix is used to investigate analytical procedures for organic nanoparticles in general, with the assumption that organic nanoparticles may serve as vehicles to enrich such beverages with apolar additives, e.g. carotenoids and pharmaceutical drugs. The NanoLyse project focuses on the development of validated methods and reference materials for analysis of engineered organic and inorganic nanoparticles in food and beverages.

Liposomes are the most thoroughly investigated class of organic nanoparticles and are used in a wide range of pharmaceutical drug-delivery strategies [6, 7]. For preliminary sample preparation μm-pore-size filters were investigated, because they are likely to safeguard the native nanostructure and remove matrix debris, which might interfere with subsequent chemical separation and characterisation steps. We chose HDC for separation of these liposome NPs from small matrix components, because it is a relatively rapid technique and more accessible to most research facilities, compared with alternative techniques such as field-flow fractionation (FFF) [8] or gas-phase electrophoretic mobility molecular analysis (GEMMA) [9]. For chemical characterisation we evaluated the use of MALDI–TOF MS as a versatile, sensitive, robust, and rapid analytical technique. Because of its soft ionisation principle, it has a wide molecular weight range from approximately 300–300,000 Da and thus enables characterisation of intact lipids [10–14], proteins [15], and polysaccharides [16]. It has already been used for analysis of intact polyamidoamine dendrimers, a class of synthetic organic nanoparticles with pharmaceutical applications [17], but not for organic NPs composed of natural components. Finally, the combination of HDC as a separation technique and MALDI–TOF as a characterisation strategy was validated as outlined by the European Decision on Methods Validation for Contaminants [18] and the International Union for Pure and Applied Chemistry (IUPAC) [19].

Experimental

Materials

All chemicals were reagent grade. MALDI–TOF matrices 2,5-dihydroxybenzoic acid and 9-aminoacridine were purchased from Sigma–Aldrich Chemie (Zwijndrecht, Netherlands). The solvents acetonitrile and methanol were from Biosolve (Valkenswaard, Netherlands). Coatsome liposomes, types A, N, and C, were purchased from NOF Europe (Grobbendonk, Belgium). After weighing, the lyophilized material (approximately 225 mg) was suspended in 2 mL water and diluted further as indicated. According to the manufacturer’s specifications suspended liposomes have diameters in the 100–300 nm range. Transmission electron microscopy (TEM) using aqueous uranyl acetate (2 %, by weight) as contrasting agent, with a Jeol 1010 electron microscope (Jeol Worldwide, Alcobendas, Spain) revealed high polydispersity of the Coatsome type A, N, and C liposomes, with size ranges 50–300 nm, 50–500 nm, and 50–400 nm, respectively. Orange-flavoured beverage, commercially available as “Aquarius Orange” (Coca Cola Company), was purchased from a local supermarket. According to the manufacturer’s specifications it contains vitamins B3, B5, B6, and E at the mg L−1 level, vitamin B7 at 30 μg L−1, monosaccharides at 79 g L−1, sodium ions at 240 mg L−1, unspecified concentrations of vitamin C, carotene, citric acid, natural orange aromas, glycerol esters of wood resin, calcium, and phosphate ions, and no fibre, proteins, or lipids.

Sample preparation (Ultrafiltration)

To remove insoluble components in the orange-flavoured beverage Acrodisc syringe filters (diameter 25 mm; Pall, Ann Arbor, MI, USA) of different pore size were tested: 0.2 μm, 0.45 μm, 1.2 μm, and 5 μm. Samples (0.5 mL) were passed through the filters by use of a 1-mL syringe. Recovery during filtration was quantified by measurement of UV absorption at 280 nm with an unfiltered control as reference.

Hydrodynamic chromatography (HDC)

HDC was performed at ambient temperature on a PL-PSD Type 1 column (800 mm × 7.5 mm), containing unspecified non-porous polymer, as stationary phase (purchased from Agilent Technologies, Amstelveen, Netherlands). Samples (50 μL) were injected with a Gilson type 234 autoinjector (Gilson International, The Hague, Netherlands) and eluted with 0.01 mol L−1 sodium dodecyl sulfate at a flow rate of 1.5 mL min−1, provided by a Gilson 307 pump. The HDC system was calibrated daily by use of polystyrene particles of defined sized: 20 nm, 30 nm, 40 nm, 46 nm, 102 nm, and 203 nm, from Duke Scientific (Palo Alto, CA, USA; purchased via Distrilab, Leusden, The Netherlands). Larger polystyrene particles, e.g. 350 nm and 700 nm, did not elute from the column which is consistent with previous observations [20]. Analytes were quantified by measurement of UV absorption at 280 nm by use of a Separations UV-detector, type K-2600 (Hendrik Ido Ambacht, The Netherlands) and evaluated with Totalchrom software, version 6.3.2, purchased from Perkin–Elmer (Groningen, The Netherlands). The values measured may be a result not only of UV absorption but also of light scattering. Fractions to be processed further for MALDI–TOF analysis were collected manually between 7 and 10 min (0.1 min fractions). For method validation a single fraction was collected from 8.3–8.7 min, encompassing more than 80 % of the Coatsome A peak (Fig. 1).

Fig. 1
figure 1

HDC chromatograms obtained from Coatsome A (anionic), Coatsome N (non-ionic), and Coatsome C (cationic) liposomes. Nanoparticle sizes, indicated in the chromatograms, are derived from a calibration curve (inset), obtained by binomial regression of retention time in HDC and diameter of polystyrene nanoparticles of defined diameters: 20, 30, 40, 46, 102, and 203 nm. The vertical line serves to indicate small differences in retention time (Rt). A 280nm refers to light absorption, which may also include light scattering, at 280 nm

Full size image

MALDI–TOF analysis

MALDI–TOF analysis was performed on Bruker Ultraflextreme II equipment (Bruker Daltonics, Bremen, Germany). The instrument is equipped with a nitrogen laser (337 nm). Laser energy, administered as 200-Hz pulses, was adjusted for each sample to achieve optimum signal-to-noise ratios. Samples were prepared by mixing analyte and appropriate MALDI–TOF matrix solution, at a 1:1 (v/v) ratio, from which 1 μL was applied to a Bruker 384 ground steel target plate and air-dried. For MALDI–TOF analysis in positive and negative modes 2,5-dihydroxybenzoic acid at 154 mg mL−1 in 90 % aqueous methanol was used as matrix. In contrast with observations by Dannenberger et al. [11] we obtained higher sensitivity in MALDI–TOF negative mode with this matrix than with 9-aminoacridine at 10 mg mL−1 in 2-propanol–acetonitrile (60:40, v/v; Fig. S1 in Electronic Supplementary Material), which may be related to the analyte lipid composition. For each MALDI–TOF MS spectrum 3 × 500 shots were recorded in reflector mode. Ions were accelerated at 20 kV. Detector gain voltage was set at 2400 V.

Validation

For validation three series of eight replicates were performed. Each of the series, which were separated by a two-week interval, was performed with a new batch of Coatsome A liposomes. Batches were weighed (225.8 ± 1.3 mg), suspended in 2 mL water, and diluted fivefold, tenfold, or twentyfold with the beverage. For assessment of linearity and limit of quantification further dilutions to 30, 100, 300, and 1000-fold of the original concentration were prepared. The beverage concentration (v/v) was 80 % for the fivefold-diluted samples and 90 % for the other dilutions. Each sample was filtered through 5-μm Acrodisc syringe filters of 25 mm diameter (Pall). The filtered solution (50 μL) was analysed by HDC as described above. From each series of eight the first result was consistently lower than the subsequent seven. Therefore, the first result was omitted from the validation analysis. Blanks were run of beverage without Coatsome A and of beverage replaced by water for the five, ten, and twentyfold dilutions of Coatsome A. Quantitative results were evaluated by use of an in-house developed format for validation of chemical analyses by analysis of variance (ANOVA) embedded in Excel software (Microsoft).

Results

Hydrodynamic chromatography (HDC) of Coatsome-type liposomes

Coatsome A (A = anionic) liposomes consist of cholesterol–l-(α)-dipalmitoylphosphatidylglycerol (DPPG)–l-(α)-dipalmitoylphosphatidylcholine (DPPC) (7:10:10, by weight). They spontaneously form liposomes upon suspension in water. HDC of these liposomes gives a single peak at 8.47 ± 0.006 min (n = 3, with three days between replicates) corresponding to a nanoparticle size of 153 nm (Fig. 1). The total peak width from 8.2–8.9 min for Coatsome A liposomes reflects a particle size range from 20–350 nm which may be because of polydispersity and/or lateral diffusion (see below). A 50 % larger peak width at half height for Coatsome A liposomes compared with that for the 203 ± 5 nm polystyrene standard (not shown) also suggests polydispersity of the liposomes. A binomial trend line, fitted by using the retention times of the polystyrene nanoparticles used for calibration, was applied to calculate the diameter of liposomes with high predictability (R 2 = 0.9995) [19, 20] (Fig. 1, inset). Linear fitting of these data points results in a lower R 2 of 0.9613.

Similar chromatograms were obtained from the other two types of Coatsome liposomes, Coatsome N (N = non-ionic) and Coatsome C (C = cationic), with slightly different compositions: Coatsome N contains cholesterol–DPPG–DPPC 17:2:7 (by weight) and Coatsome C contains cholesterol–stearylamine–DPPC 7:1:17 (by weight). HDC of Coatsome N and Coatsome C liposomes each gave a single peak at 8.41 ± 0.006 min and 8.37 ± 0.01 min, respectively, (n = 3, with three days between replicates), reflecting diameters of 187 and 205 nm, respectively (Fig. 1). Peak widths of Coatsome types A, N, and C are consistent with the polydispersity indicated by the provider and with TEM characterisation (Experimental section). In addition, lateral diffusion may have contributed to peak broadening. Molecules smaller than approximately 18 nm, eluting after 8.9 min, were almost undetectable in these samples. When water was used as a mobile phase instead of 10 mmol L−1 sodium dodecylsulfate no detectable peaks eluted from the column. The behaviour of polystyrene particles on HDC separation with water as mobile phase was also highly aberrant in the sense that particles larger than 30 nm had retention times much shorter than expected and no correlation could be established between retention time and diameter. Retention times in HDC varied by approximately 0.05 min between months for the three Coatsome types, but their ranking order remained the same.

MALDI–TOF analysis of Coatsome-type liposomes

MALDI–TOF-MS analysis in positive mode of Coatsome A liposomes revealed major peaks at m/z 734.5, 756.6, 1468.1, and 1490.1 amu (atomic mass units) and smaller peaks at 991.6 and 1229.8 (Fig. 2). The major peaks correspond to [M + H]+ and [M + Na]+ adducts of monomeric l-(α)-dipalmitoylphosphatidylcholine (DPPC) and its dimer. On dilution to 1 mg mL−1 the monomer-to-dimer peak-height ratio did not change. The smaller peaks could not be explained on the basis of the chemical composition given by the provider. Other components of Coatsome A nanoparticles, DPPG and cholesterol, were not observed. MALDI–TOF mass spectra of Coatsome N were very similar to that of Coatsome A (data not shown). Coatsome C gave the same major peaks but the peaks at 991.6 and 1229.8 were of similar intensity to the 734.6 and 1468.1 peaks (Fig. 2). In the negative mode Coatsome A and Coatsome N nanoparticles gave a peak at m/z = 721.3 amu corresponding to the [M − H]− adduct of l-(α)-dipalmitoylphosphatidylglycerol (DPPG) (Fig. 2). Because of the absence of this phospholipid, MALDI–TOF MS spectra of Coatsome C lack the 721.3 peak but contain major peaks at m/z = 648.2 and 886.4. The former could not be explained on the basis of the chemical composition of Coatsome C; the latter is very likely to be an [M − H]− adduct of DPPC (m/z = 733.5) and the MALDI–TOF matrix 2,5-dihydroxybenzoic acid (m/z = 154.0) [12].

Fig. 2
figure 2

MALDI–TOF MS spectra of Coatsome A and Coatsome C liposomes, in both positive and negative modes. DPPC = l-(α)-dipalmitoylphosphatidylcholine, DPPG = l-(α)-dipalmitoylphosphatidylglycerol. R1 = R2 = CH3(CH2)14

Full size image

Coatsome A was used for further evaluative quantitative analyses after suspension in a beverage matrix, because it contained similar proportions of DPPC and DPPG, which give clear signals that can be used for MALDI–TOF MS characterisation.

Sample preparation for Coatsome A nanoparticles in a beverage matrix

To remove residues from the matrix, e.g. fibres in beverages, which might interfere with consecutive analysis, we investigated ultrafiltration as a first sample-preparation step. Filters with 0.2-μm and 0.45-μm pores removed such residues but also approximately 30 % of the Coatsome A liposomes. With filters with 1.2-μm and 5-μm pores the beverage solution was transparent, and for Coatsome A nanoparticles recovery was higher than 95 %. To maximise recovery we chose 5-μm filters for sample preparation.

Combined separation by HDC and characterisation by MALDI–TOF MS of Coatsome A liposomes in a beverage matrix

To investigate the two-step concept for characterisation by MALDI–TOF MS after separation by HDC suspensions of Coatsome A liposomes in beverages were first filtered through 5-μm filters to remove insoluble components from the beverage. HDC of the filtered beverage alone gave a large UV absorbance peak at the solvent front, indicating the presence of components smaller than 18 nm (Fig. 3, top panel). As indicated by the commercial provider, the beverage contains a variety of UV-absorbing, low-molecular-weight components, for example vitamins, carotene, natural orange aromas, and glycerol esters of wood resin. HDC of a suspension of Coatsome A in the beverage gave a peak of relatively small particles of the beverage and the Coatsome A peak at 8.47 min with almost baseline separation (Fig. 3, bottom panel).

Fig. 3
figure 3

HDC chromatograms obtained from a blank beverage (top panel) and a suspension of Coatsome A liposomes in this matrix (bottom panel) after filtration through 5-μm filters. A 280nm refers to light absorption, which may also include light scattering, at 280 nm

Full size image

Fractions of 0.1 min (= 0.15 mL) were collected between 7 and 10 min and processed for MALDI–TOF MS in the positive mode. Signals with m/z values of 734.5 and 756.5, which are typical for l-(α)-dipalmitoylphosphatidylcholine, were observed in the fractions collected from 8.2 to 8.9 min (Fig. 4, top four panels). This corresponds to the HDC retention times for Coatsome A as detected by UV absorption at 280 nm. The characteristic 721.3 signal of l-(α)-dipalmitoylphosphatidylglycerol was seen in the negative mode in fractions collected between 8.4 and 8.7 min (Fig. 4, bottom four panels).

Fig. 4
figure 4

MALDI–TOF MS spectra, in positive mode (top four panels) and negative mode (bottom four panels), of Coatsome A liposomes in beverage after filtration through 5-μm filters and fractionation by HDC. Times, indicated in the right corner, are retention times of the collected fractions. Arrows indicate MALDI–TOF MS signals which are characteristic for the liposome components DPPC (top four panels) and DPPG (bottom four panels)

Full size image

Validation of quantitative analysis of Coatsome A nanoparticles after ultrafiltration and HDC separation

Relative standard deviations for seven replicates performed on the same day were 7.2 %, 6.9 %, and 3.2 % at 5.6, 11.3, and 22.6 mg Coatsome A mL−1 in beverage, respectively (Table 1), which is indicative of good repeatability. Relative standard deviations between days of 8.2 %, 7.4 %, and 3.3 % are indicative of high within-laboratory reproducibility of the analytical method. As expected, repeatability and reproducibility improve with increasing Coatsome A concentration.

Table 1 Results from validation of HDC analysis of Coatsome A in an orange-flavoured beverage matrix
Full size table

Recovery of Coatsome A nanoparticles from the beverage after ultrafiltration through 5-μm filters and HDC separation from beverage components was 60.9 ± 4.5 % (n = 3) of the values observed when the HDC column was bypassed with empty Peek tubing (length 35 m, Ø inner = 0.004 mm). When Coatsome A liposomes were dissolved in water instead of beverage the recovery was lower: 45.1 ± 3.5 %.

Coatsome A liposomes could be quantified with a good linear relationship (R 2 > 0.99) between Coatsome A concentration and UV280nm-based quantification in a five-point calibration series, 1.1, 3.7, 5.6, 11.3, and 22.6 mg Coatsome A mL−1 beverage. Coatsome A at concentrations below 1 mg mL−1, were not detectable by UV absorption at 280 nm; at concentrations ≥45 mg mL−1 UV absorption was too high for accurate quantification.

For suspensions of Coatsome A in beverage, subjected to ultrafiltration and HDC, MALDI–TOF-MS signals at m/z 734.5 and 756.6 for DPPC could be observed at Coatsome A concentrations of 3.7 mg mL−1 and higher. In an equivalent test for detection capability (CC β) 20 suspensions at 4 mg Coatsome A mL−1 in beverage were separately processed for MALDI–TOF MS including ultrafiltration and HDC. For this purpose HDC fractions eluting from 8.3–8.7 min were pooled. All 20 of these suspensions gave the two MALDI–TOF signals at m/z = 734.5 and 756.6.

Discussion

In this paper we describe a two-step concept for characterisation of organic NPs, in this case liposomes, in a simple food matrix. In the first step hydrodynamic chromatography (HDC) is used to separate NPs from the matrix and other smaller particles on the basis of size in the nanometre range. In the second step, chemical characterisation, MALDI–TOF MS is used for chemical identification of the phospholipid components of the liposomes. HDC and MALDI–TOF MS are readily accessible to most research facilities. HDC requires basic HPLC equipment, with a dedicated HDC column as the only additional component. It enables separation of large structures in the 20–200 nanometre range. MALDI–TOF MS may be regarded as an MS technique which is tolerant of low concentrations of salts, most other food components, and low concentrations of detergents used as constituents in solvents for size separation techniques [14]. Separation of nanoparticles on the basis of size is necessary to ensure the characterisation includes only NPs and not their monomeric building blocks. MALDI–TOF MS will only show the monomeric building blocks, because the upper molecular weight threshold of approximately 300 kDa is too low for detection of intact liposomes with a unit weight of several megadaltons. Furthermore, sample preparation for MALDI–TOF MS analysis includes organic solvents which destroy the assembled state of the intact nanoparticle.

For Coatsome A the two-step separation/characterisation concept was validated as outlined in European Commission Decision 2002/657/EC on method validation for contaminants [18] and by IUPAC [19], and repeatability, within-laboratory reproducibility, linearity, limit of detection and quantification, and recovery were determined. Detection capability (CCβ) was covered by the observation that characteristic MALDI–TOF signals could be acquired for all 20 suspensions of Coatsome A in beverage at a concentration close to the detection limit for MALDI–TOF MS of approximately 3.7 mg mL−1.

Estimation of nanoparticle diameter by HDC using calibration with particles of defined size has been proposed by Takeuchi et al. [21], who used silica colloids for size calibration. In their study a linear, negative size-versus-retention time correlation was used to predict diameters of unknown particles. However, they also proposed a binomial relationship based on theoretical considerations, which led to the equation:

$$ \frac{{{t_i}}}{{{t_j}}}=\frac{{\left( {d_{ch}^2+2{d_j}{d_{ch }}-d_j^2} \right)}}{{\left( {d_{ch}^2+2{d_i}{d_{ch }}-d_i^2} \right)}} $$

where t i and t j are elution times for two spherical particles of diameters d i and d j, respectively, and d ch is the apparent interstitial diameter of the packed bed column. The equation was based on work by Tijssen et al. [22] for HDC with open capillary columns. In our study we compared binomial and linear correlations. The higher R 2 value for the binomial correlation empirically substantiated much higher predictability.

The commercially available liposomes referred to as “Coatsome A”, are composed of l-(α)-dipalmitoylphosphatidylcholine (DPPC), l-(α)-dipalmitoylphosphatidylglycerol (DPPG), and cholesterol; of these, the first two phospholipid components produced a signal in MALDI–TOF mass spectrometry. This predominance in signal intensity of the two phospholipids, with DPPC being the only signal in the positive mode and DPPG in the negative mode, is a well-known phenomenon for mixtures containing these two phospholipids. It is very likely to be a consequence of ion suppression [10–14]. Given the fact that the MALDI–TOF-signals are typical for these two phospholipids and observed only in fractions separated by HDC on the basis of a specific particle size makes the signals highly indicative and selective for the presence of these lipid-based nanoparticles. MALDI–TOF signal intensity is not an accurate measure for quantification. UV-absorption after HDC separation of the liposomes, which is also more sensitive, seems a much better approach to quantification of these nanoparticles, as substantiated by the validation described in this study.

In this exploratory study, we started with a relatively simple matrix, an orange-flavoured beverage, and observed a substantial effect compared with distilled water—a much higher response was obtained for the beverage. Two explanations seem most likely for this observation:

  1. 1.

    higher stability of Coatsome liposomes in the beverage; and/or

  2. 2.

    reduced interaction with the stationary phase because of the greater ionic strength of the beverage.

Because recovery from both matrices was very reproducible, the latter explanation seems most appropriate, because the time between mixing liposomes and beverage or water and analysis on the HDC column varies substantially between samples. This variable time span would have resulted in variability in recovery for at least one of the two matrices. The strongly aberrant chromatographic behaviour upon HDC with water as the mobile phase, as described also by Takeuchi et al. [21], indicates the requirement for a specific ionic strength to prevent undesirable stationary phase–analyte interactions.

Liposomes are the most commonly applied NPs for therapeutic drug delivery, with a wide variety of applications [6, 7]. When administered as pharmaceutical drugs many other, and very likely much more complex, matrices will have to be tested. This study shows that liposomes, when embedded in a simple matrix, can be quantified by a sequence of analytical steps starting with mild sample preparation, e.g. filtration, followed by separation of NPs on the basis of size, and, finally, characterization of the separated NPs by use of mass spectrometry, which provides information on the characteristic building blocks of the organic NPs.

In our laboratory we have tested this analytical concept of characterisation of organic NPs by, first, separation by HDC then analysis by MALDI–TOF MS for identification of other types of organic NP. In this process it became apparent that HDC conditions may lead to dilution and/or the presence of SDS which may hamper generation of MALDI–TOF signals, e.g. for protein-based NPs [23] or may lead to disassembly of unstable NPs, as observed for carbohydrate-based NPs. For these challenges appropriate solutions must be found. An important, additional aspect will be to extend the procedures to enable isolation of the nanoparticles from more complex food and other matrices, e.g. by filtration and/or centrifugation. Much attention will have to be given to effects of the matrix on the stability of the nanoparticles and chromatographic conditions to enable accurate quantification. Other separation techniques, for example field-flow fractionation (FFF) [8] and gas-phase electrophoretic mobility molecular analysis (GEMMA) [9], may be investigated, as well as other chemical analytical strategies, e.g. LC–MS–MS, for chemical characterisation. In addition to alternative separation techniques, other detector strategies which give direct information about size and polydispersity may also be investigated, for example MALLS or DLS, and contribute to the approach described in this study in which size determination was based on calculation from a calibration line. In the recent past different multi-detector approaches have been used in combination with HDC, in particular for size determination of polymers and inorganic nanoparticles [3, 20, 24–27]. Ultimately, these procedures should be validated for quality control of food, cosmetics, and pharmaceutical products.