Quantification of Nanoparticles via Flow Injection Analysis Using Asymmetric Flow Field-flow Fractionation (AF4)

Abstract

To determine the number density of fine precipitates in steels by asymmetric flow field-flow fractionation (AF4) with inductively coupled plasma–mass spectrometry (ICP–MS), an analysis method employing flow injection was investigated. For accurate calibration, matrix matching was performed by mixing the standard solution and AF4 carrier solution in front of a nebulizer. Two surfactants were used for AF4 separation; it was found that the appropriate selection of surfactants based on their acidity constant is essential to avoid salt precipitation. In addition, the effect of the AF4 retention time on recovery was investigated. A long retention time led to the adsorption and aggregation of the samples in the AF4 separation channel. Results showed that an AF4 retention time within 20 min facilitated superior recovery. Moreover, five types of AuNPs were analyzed via AF4–ICP–MS and quantified using flow injection analysis. Good analytical performance was achieved for all AuNPs and the recoveries exceeded 93%, and the coefficient of variation was within 5%. The effect of particle size on the recovery was not confirmed.

Furthermore, the developed flow injection analysis for AF4–ICP–MS was applied to evaluate niobium carbide (NbC) precipitates in steels. The number density of nanometer-sized NbC was quantified to be within 10¹³ to 10¹⁴ particles per 1 g of Fe. It was quantitatively confirmed that the long-duration heat treatment led to an increase in the number density of nanometer-sized NbC. Hence, this method can be useful for quantitatively analyzing the size and number density of nanoprecipitates in steels.

1. Introduction

Nanoparticles, which are generally defined as particles smaller than 100 nm, offer advantages that are distinct from those of their bulk materials. Their unique features include the quantum size effect and large surface area per volume. For example, CdSe¹⁾ and InAs²⁾ quantum dots can absorb light of different wavelengths depending on their size. In the steel industry, the characteristics of nanometer-sized precipitates are utilized effectively. These precipitates are important for strengthening steels³⁾ and controlling the crystal grain size of steels.⁴⁾ Their size distribution and number density affect the properties of steel, and hence, must be evaluated accurately to determine the design of the alloy composition and conditions of the steel manufacturing processes. For many steels, the precipitates exhibit a wide size distribution and a non-uniform dispersion state. Hence, electron microscopy cannot be employed for analysis because the observation areas are locally limited. Conventional chemical analysis of precipitates and inclusions has been performed for various steel samples; however, the precipitates and inclusions are dissolved by acids⁵⁾ or alkalis, resulting in the loss of their size information. Meanwhile, asymmetric flow field-flow fractionation (AF4), which is a size-based separation technique, was used to analyze various nanoparticles^6,7,8) and steels.^9,10,11) AF4 does not require chromatographic support and can achieve size-dependent separation. In addition, AF4 can be combined with various detectors, such as light-scattering, ultraviolet absorption, and inductively coupled plasma–mass spectrometry (ICP–MS) to obtain information on the separated samples. In our previous study,⁹⁾ AF4–ICP–MS analysis was applied to niobium carbide (NbC)-precipitated ferritic steel, and the difference in the size distribution of NbC was evaluated between two steel samples subjected to different aging times; however, the number density of NbC could not be quantitatively evaluated.

In this study, a quantitative analytical method for AF4–ICP–MS was investigated using flow injection analysis (FIA).^12,13) A calibration curve was constructed by integrating the flow injection (FI) chromatogram of the metallic standard solutions measured while mixing the AF4 carrier solution. Moreover, the accuracy of this method was evaluated based on the recovery of several gold standard nanoparticles (AuNPs). Furthermore, the developed analytical method was applied to NbC-precipitated ferritic steels, and the applicability of the FI–AF4–ICP–MS analysis was investigated.

2. Experimental

2.1. Nanoparticles and Chemicals

All AuNPs were purchased from nanoComposix (San Diego, CA, USA). Table 1 summarizes the specifications of the reference nanoparticles dispersed in water and those of the specimens. AuNP-5, AuNP-10, AuNP-30, AuNP-50, and AuNP-100 were used for the recovery test via FIA in AF4–ICP–MS. AuNP-2, AuNP-5, and AuNP-10 were used to calibrate the size measurements for the AF4 analysis of the specimens.

Table 1. Specifications of nanoparticles.

	AuNP-2	AuNP-5	AuNP-10	AuNP-30	AuNP-50	AuNP-100
Average diameter by TEM (nm)	2.1 ± 0.3	5.0 ± 0.6	9.5 ± 0.4	28.0 ± 0.9	50.3 ± 2.3	102 ± 4.2
Concentration (mg mL−1)	0.053	1.08	0.052	0.052	0.050	0.053
Number concentration (particles mL−1)	5.5×1014	8.7×1014	5.9×1012	2.4×1011	3.9×1010	4.9×109
Hydrodynamic diameter by DLS (nm)	–	–	19	36	60	107
Zeta potential (mV)	–	–	−20	−46	−53	−50
pH	7.0	6.4	7.7	7.7	7.5	7.6
Surface	Glutathione	Sodium citrate	PEG12 carboxylic acid
Solvent	Water		2 mmol L−1 aqueous sodium citrate

All chemicals used for the AF4–ICP–MS measurements were purchased and used without additional purification. Ultrapure water (>18 MΩ: Milli-Q water purification system and Elix UV10, Millipore Corp., USA) was used to dilute the samples and prepare AF4 carrier solutions. Sodium dodecyl sulfate (SDS, purity ≥ 99.0%, FUJIFILM Wako Pure Chemical Corporation) and sodium cholate (SC, purity ≥ 98.5%, FUJIFILM Wako Pure Chemical Corporation) were used as the AF4 carrier solutions. Nitric acid (purity ≥ 60.0–61.0%), hydrochloric acid (purity ≥ 35.0%–37.0%), and Au standard solution (Au1000, 1000 mg L⁻¹, for atomic absorption spectrometry) were obtained from Kanto Chemical Co., Inc. and used to prepare the standard solution. A multi-element standard solution (XSTC-8, 10 mg L⁻¹, SPEX CertiPrep, Metuchen, USA) was used to quantify the NbC precipitates in steels via FI–AF4–ICP–MS. Acetyl acetone (AA, special-grade, FUJIFILM Wako Pure Chemical Corporation), tetramethylammonium chloride (TMAC, purity ≥ 98.0%, Tokyo Chemical Industry Co., Ltd.), and methanol (special-grade, FUJIFILM Wako Pure Chemical Corporation) were used to extract the NbC precipitates from the steel specimens.

2.2. Quantification Method in AF4–ICP–MS via FIA

A path for flowing the standard solution was installed, independent of the AF4 equipment. The path comprised a syringe pump, six-way switching valve, and sample loop. The syringe pump and six-way switching valve were connected to the nebulizer. The sample loop was connected to a six-way switching valve, and the volume was adjusted to 552 μL. A washing path was connected to the six-way switching valve. The path could be changed using the six-way switching valve. Au standard solutions of 0, 0.01, 0.025, 0.05, and 0.1 mg L⁻¹ with 5% (v/v) aqua regia were prepared (denoted as Std1–5). Multi-element standard solutions of 0, 1.0, 3.0, 7.5, and 10 ng L⁻¹ with 1% (v/v) nitric acid were also prepared. These standard solutions were injected from the washing path to the sample loop under conditions that were unconnected to the nebulizer. After preparing the sample loop using the standard solutions, the path was switched and a constant volume of standard solutions was flowed to the nebulizer using a syringe pump. The standard solution was detected via time-resolved analysis, and the peak area of the signals detected via ICP–MS was integrated. The calibration curve was obtained from correlation between the injected amounts and peak areas.

2.3. Sample Preparation of Precipitates in Steels via Selective Potentiostatic Etching by Electrolytic Dissolution

The previously reported NbC-precipitated ferritic steel,⁹⁾ whose main chemical composition is 0.1% Nb and 0.01% C (mass%), was used in this study. Heat treatment was conducted at 873 K for 1 or 10 h after solid solution treatment (1523 K, 24 h) and water quenching. The sizes of the NbC precipitates were similar; however, the number density was not quantified in the previous study. The small specimens were cut into dimensions of 25 mm × 25 mm × 2 mm for sample preparation.

To obtain precipitates in the steel specimens, selective potentio-static etching by electrolytic dissolution (SPEED)^14,15) was performed, for which a solution of 10% (v/v) AA, 1% (w/v) TMAC, and 500 μg mL⁻¹ SDS–methanol was used (known as “SPEED solution”) was used. Herein, SDS as a surfactant, which adsorbs on the surface of extracted precipitates was added to prevent their precipitates from aggregating in the SPEED solution. As in the previous study, electrolytic extraction was conducted in two steps at a constant current of 500 mA. The first step was electrolytic dissolution, which was conducted for 15 min to dissolve the pollutants and oxide layer on the surface of the steel specimens. After completing the first step, the sample specimen was transferred to another SPEED solution with the same components to separate the dissolved surface pollutants and oxide layer from the steel specimen. Next, a second electrolytic dissolution step was conducted for 120 min. After 120 min of the electrolytic treatment, almost 1.0 g of the steel specimen dissolved in the SPEED solution. After that, an ultrasonic treatment was performed among 1 min, all precipitates were collected and dispersed in the SPEED solution. This solution sample including dispersed precipitates was subjected to following AF4–ICP–MS measurements.

2.4. Instruments

Figure 1 shows a schematic of the Wyatt Eclipse AF4 system (Wyatt Technology Europe, Germany), which was combined with the ICP–MS and FI equipment. This system was equipped with a high-performance liquid chromatography pump, an integrated degassing system, and an autosampler (Agilent Technologies 1260 Infinity series, Agilent Technologies, USA). After injecting the NPs into the AF4 separation channel, they were diffused and transported using the channel flow and separated in this system. During this time, 300 mg L⁻¹ of SDS solution or 500 mg L⁻¹ of SC solution was used as the AF4 carrier solution. Moreover, a regenerated cellulose (RC) ultrafiltration membrane (Microdyn–Nadir, Germany) with a molecular weight cutoff of 30 kDa (RC 30 kDa) was used as the accumulation wall for the AF4 separation channel. Subsequently, a separation device with a 275 mm-long channel was employed, which was equipped with a 350 μm-thick or 490 μm-thick asymmetric diamond-shaped channel spacer. The AF4 separation channel was directly connected to the ICP-quadrupole MS equipment (Agilent8800, Agilent Technologies, USA). The online elemental information on the nanoparticles separated by AF4 was obtained using ICP–MS.

Fig. 1.

Schematic of AF4–ICP–MS and FI apparatus. (a) Sample load; (b) sample injection. (Online version in color.)

The sample introduction system comprised a self-aspirating nebulizer (Conikal DC Nebulizer, Glass Expansion, Port Melbourne, Australia) and a Scott-type chamber equipped with Peltier cooling apparatus. Table 2 summarizes the operating conditions in ICP–MS. Prior to the AF4 measurement, the membrane was conditioned with an AF4 carrier solution of SDS or SC for at least 30 min. Each AuNP-dispersed solution was directly analyzed without dilution to maintain a good dispersion state in the sample vials. Precipitates dispersed in the SPEED solution were also analyzed without dilution. Additionally, these samples were injected into the AF4 separation channel using an autosampler and analyzed. Table 3 summarizes the AF4 separation conditions.

Table 2. ICP–MS operating conditions.

Agilent8800 ICP–MS
RF power (W)	1550
Nebulizer gas flow (L min−1)	0.82
Make-up gas flow (L min−1)	0.23
Cooling gas (L min−1)	15
Auxiliary gas (L min−1)	0.90
Sampling depth (mm)	7.0
Element (m/z)	Nb (93), Au (197)
Duration time (s)	0.10
Spray chamber temperature (K)	275.15

Table 3. AF4 separation condition and schematic of AuNPs and NbC precipitates.

		For AuNPs	For AuNPs	For NbC precipitates
Channel parameters	Membrane nature	Regenerated Cellulose (RC)	Regenerated Cellulose (RC)	Regenerated Cellulose (RC)
	Membrane cut-off	30 kDa	30 kDa	30 kDa
	Spacer	350 μm	350 μm	490 μm
	Elution solvent	1.04 mmol L−1 SDS	1.04 mmol L−1 SDS	1.04 mmol L−1 SDS
Fractionation time	Elution time	1 min	35 min	1 min
	Focusing time	1 min	0 min	1 min
	Focus+Injection time	2 min	0 min	2 min
	Focusing time	3 min	0 min	1 min
	Elution time	35 min	0 min	45 min
Fractionation step, flow, and volume	Injection volume	1 μL	1 μL	100 μL
	Injection flow	0.2 mL min−1	0.2 mL min−1	0.2 mL min−1
	Channel flow (Vout)	1.0 mL min−1	1.0 mL min−1	1.0 mL min−1
	Cross flow (Vc)	Conditon 1: 0.5→0 mL min−1	Conditon 0: 0 mL min−1	2.0→0.1 mL min−1 (linear gradient)
		Conditon 2: 1.5→0 mL min−1
		Conditon 3: 2.0→0 mL min−1
		Conditon 4: 3.0→0 mL min−1
		Conditon 5: 4.0→0 mL min−1 (linear gradient)
	Focus flow	3.0 mL min−1	0 mL min−1	3.0 mL min−1

3. Results and Discussion

3.1. Development of Calibration Method Using FIA for AF4–ICP–MS

The Au standard solution and AuNP-2 were analyzed using only AF4–ICP–MS, as shown in Fig. 2, and the peak areas of each chromatogram were compared. The peak area of AuNP-2 was approximately 3.3 times larger than that of the Au standard solution despite half the injection amount of Au to the AF4 separation channel. It was inferred that the introduced amounts of metallic standard solution in ICP–MS decreased significantly because a significant amount of metallic ions had passed through the membrane in the AF4 separation channel and then removed as waste solution. Hence, the injection of a metallic standard solution into the AF4 separation channel was not suitable for obtaining an accurate calibration curve.

Fig. 2.

AF4 chromatogram of Au standard solution and AuNP-2. The concentrations of the Au standard solution and AuNP-2 were 0.1 and 0.053 mg L⁻¹, respectively. (Online version in color.)

Therefore, FIA was further performed for the accurate quantification from AF4–ICP–MS. First, the calibration method was investigated using a standard solution. To conduct matrix matching for ICP–MS, the concentration of the AF4 carrier solution was adjusted to twice that used in the AF4 analysis, and the AF4 carrier solution was mixed with the standard solution involving 5% (v/v) aqua regia or 1% (v/v) nitric acid at the same volume ratio immediately before the nebulizer. Figure 3 shows the FI chromatogram of the Au standard solution using ultrapure water as the AF4 carrier solution. Each peak of the Std chromatogram was integrated, and a calibration curve was constructed using the peak areas. Figure 4 shows the FI calibration curves of the Au standard solution using each AF4 carrier solution. The calibration curve of the SDS solution indicated good linearity, similar to that of ultrapure water. By contrast, that of the SC solution was inferior and indicated significant variations. This was caused by the difference in the acidity constants of each surfactant. The acidity constants of SDS and SC were 1.84 and 4.98, respectively. When Std with 5% (v/v) aqua regia as the matrix was mixed with each of the surfactant solutions, the pH decreased below the acidity constant of SC. Figure 5 shows mixed solutions of 5% (v/v) aqua regia and SDS or SC solutions at the same volume ratio. The SDS solution was transparent; however, the precipitation of salts was observed in the SC solution. The significant variation in signal intensity presented in the FI calibration curve was caused by the blockage of the nebulizer and/or PFA tubes via the precipitation of SC salts. Thus, FIA using the SC solution is not preferable because of the unstable signal intensity. Hence, the acidity constant of the surfactants should be considered when selecting the surfactants to be used for AF4–ICP–MS involving FIA.

Fig. 3.

FI chromatogram of Au standard solution (Std). Ultrapure water was applied as carrier solution at 1.0 mL min⁻¹. Std 1–5 concentrations of 0, 0.01, 0.025, 0.05, and 0.1 mg L⁻¹ were prepared.

Fig. 4.

Comparison of FI calibration curves of the Au standard solution among ultrapure water (△), SDS solution (□), and SC (○) solution. Measurements for the FIA were performed three times.

Fig. 5.

Mixed solutions of 5% (v/v) aqua regia and (a) 1.16 mmol L⁻¹ SC solution or (b) 1.04 mmol L⁻¹ SDS solution at the same volume ratio. (Online version in color.)

3.2. Quantification of AuNPs Using AF4-ICP-MS via FIA

To investigate the analytical accuracy and precision of the FIA, AF4–ICP–MS was conducted using AuNPs. The AuNPs detected in AF4–ICP–MS were quantified using the FI calibration curve. The analytical accuracy and precision were evaluated using the recovery equation shown in Eq. (1). In addition, the injected amounts were determined by quantifying the detected amounts using the FI calibration curve when an equal amount of AuNPs was analyzed via AF4–ICP–MS under only channel flow, while considering the uncertainty of sampling.   

$$\begin{array}{l}\text{Recovery}\mathrm{\hspace{0.17em} } \text{(\%)}=\\ \frac{The\mathrm{\hspace{0.17em} } amounts\mathrm{\hspace{0.17em} } quantified\mathrm{\hspace{0.17em} } by\mathrm{\hspace{0.17em} } FIA\mathrm{\hspace{0.17em} } analysis\mathrm{\hspace{0.17em} } (g)}{The\mathrm{\hspace{0.17em} } injected\mathrm{\hspace{0.17em} } amounts\mathrm{\hspace{0.17em} } of\mathrm{\hspace{0.17em} } samples\mathrm{\hspace{0.17em} } (g)}\times 100\end{array}$$(1)

In this study, the effect of the retention time of the samples on the recovery achieved by AF4 was investigated. The AF4 separation conditions could be determined by changing the channel flow rate, cross-flow rate, channel volume, temperature, and solvent viscosity, as shown in Eq. (2).^16,17) In particular, the channel flow rate and cross flow rate were the easiest parameters to adjust, to change the AF4 separation conditions.   

$$t_R=\frac{w^2{R}_h\pi \eta }{kT}ln\left(1+\frac{V_c}{V_0}\right)$$(2)

where V₀ is the channel flow rate, V_c is the cross-flow rate, w is the thickness of the channel spacer, η is the solvent viscosity, k is the Boltzmann constant, T is the absolute temperature, and R_h is the hydrodynamic diameter.

The AF4 separation conditions were changed using only the cross-flow rate, as shown in Table 3. Figure 6(a) shows the AF4–ICP–MS chromatogram of AuNP-50 under each AF4 separation condition. The higher the cross-flow rate, the longer was the retention time. Additionally, Fig. 6(b) shows the recovery of AuNP-50. The longer retention time led to the decreased recovery of AuNP-50, and a retention time longer than 25 min significantly affected the recovery of the samples. This was caused by the aggregation of the samples and adsorption to the membrane in the AF4 separation channel. Aggregates of the samples were observed for AF4 separation conditions 2 to 5. The peaks of the aggregates were not considered in the recovery evaluation; however, even if these peaks were considered, the recovery did not reach 100%. It was discovered that a portion of AuNP-50 adsorbed to the membrane in the AF4 separation channel, resulting in no detection of a portion of AuNP-50 by ICP–MS. Therefore, the retention time of the samples in the AF4 analysis should be adjusted to within 20 min.

Fig. 6.

(a) AF4–ICP–MS chromatogram of AuNP-50 under each AF4 separation condition, (b) Recovery of AuNP-50 under each AF4 separation condition. (Online version in color.)

Furthermore, five types of AuNPs were analyzed via AF4–ICP–MS thrice under appropriate separation conditions and quantified using FIA. Figure 7 shows the recovery of each AuNP. Good analytical performance was achieved by the developed FIA, all recoveries exceeded 93%, and the coefficient of variation was within 5%. The effect of particle size on the recovery was not confirmed; however, it was speculated that the sampling variation and dispersion state of the samples caused a slight recovery loss.

Fig. 7.

Recovery of AuNPs in AF4–ICP–MS analysis. Analyses were conducted three times for each AuNP.

3.3. Quantification of Niobium Carbides in Steels by FI–AF4–ICP–MS

To investigate the applicability of the FIA for AF4–ICP–MS, NbC precipitates extracted from the steel samples were measured via AF4–ICP–MS and quantified using the FIA. Consequently, FIA was discovered to be effective for quantitatively evaluating nanometer-sized precipitates in steels.

The NbC precipitates were measured via AF4–ICP–MS under the conditions listed in Table 3. A calibration method for AF4 size measurements was conducted as follows. First, a calibration curve was made by plotting each average size of several AuNPs acquired using transmission electron microscopy (TEM) and their AF4 retention times. After measuring NbC precipitates via AF4–ICP–MS, the size distribution was estimated by converting each AF4 retention time to the size using the above calibration curve. Additionally, the width of size distribution was corrected by a broadening coefficient of AF4 measurement as reported method.⁹⁾ In the previous report,⁹⁾ the accuracy of the size estimation was confirmed by comparing the result of AF4–ICP–MS, TEM, and small angle X-ray scattering. Figure 8 shows the number density of the NbC precipitates measured via FI–AF4–ICP–MS. The number density of the NbC precipitates was calculated using the average diameter, which was almost 2 nm, measured by AF4⁹⁾ and the density of NbC. The steels were subjected to heat treatment at 873 K for 1 or 10 h. This increased the number density of the NbC precipitates by approximately six times. Furthermore, TEM observations showed that these NbC precipitates were not uniformly dispersed in steels; therefore, the precise number density could not be evaluated using TEM. By contrast, FI–AF4–ICP–MS enabled the analysis of more precipitates in the steels than TEM and atom probe tomography (APT), and was advantageous in terms of sample representativeness. Therefore, FI–AF4–ICP–MS is effective for quantitatively measuring the size and number density of precipitates in steel.

Fig. 8.

Quantification of number density of NbC precipitates in NCA5-1 and NCA5-3 steels. Heat treatment was performed at 873 K for 1 and 10 h on NCA5-1 and NCA5-3, respectively.

In conclusion, the FIA for AF4–ICP–MS could facilitate the quantification of various nanoprecipitates in steels. However, the location information of the precipitates was lost owing to extraction from steels in this analysis. Hence, AF4–ICP–MS should be used complementarily with local observation techniques such as TEM and APT, and a multiphase evaluation should be conducted to interpret the various phenomena in steel manufacturing processes.

4. Conclusions

The quantification method via FIA for AF4–ICP–MS was investigated using AuNPs and NbC precipitates in steels. The findings of this study are summarized as follows:

(1) The surfactant for the AF4 carrier solution should be selected considering the acidity constant because the standard solutions required to obtain the calibration curve must be mixed with AF4 carrier solutions to conduct appropriate matrix matching.

(2) The retention time of samples in AF4 analysis should be adjusted to within 20 min to avoid the aggregation and adsorption of samples to the membrane in the AF4 separation channel.

(3) FI–AF4–ICP–MS analysis quantitatively revealed that the change in the holding time of the heat treatment of the steel specimens from 1 to 10 h increased the amount of fine NbC precipitates by approximately six times.

The FIA was effective for the number density evaluation of various nanoparticles using AF4–ICP–MS analysis. Hence, this analysis should be useful for determining the size distribution and number density of various precipitates in steels.

References

• 1)   C.  Bullen and  P.  Mulvaney: Langmuir, 22 (2006), 3007. https://doi.org/10.1021/la051898e
• 2)   K.  Yamaguchi,  K.  Yujobo and  T.  Kaizu: Jpn. J. Appl. Phys., 39 (2000), L1245. https://doi.org/10.1143/JJAP.39.L1245
• 3)   N.  Kamikawa,  Y.  Abe,  G.  Miyamoto,  Y.  Funakawa and  T.  Furuhara: ISIJ Int., 54 (2014), 212. https://doi.org/10.2355/isijinternational.54.212
• 4)   H.  Ohta,  R.  Inoue and  H.  Suito: ISIJ Int., 48 (2008), 294. https://doi.org/10.2355/isijinternational.48.294
• 5)   S.  Kinoshiro,  T.  Ishida,  M.  Inose and  K.  Fujimoto: Tetsu-to-Hagané, 99 (2013), 362 (in Japanese). https://doi.org/10.2355/tetsutohagane.99.362
• 6)   K. C.  Nwoko,  X.  Liang,  M. A.  Perez,  E.  Krupp,  G. M.  Gadd and  J.  Feldmann: J. Chromatogr. A, 1642 (2021), 462022. https://doi.org/10.1016/j.chroma.2021.462022
• 7)   D.  Ojeda,  M. V.  Taboada-López,  E.  Bolea,  J.  Pérez-Arantegui,  P.  Bermejo-Barrera,  A.  Moreda-Piñeiro and  F.  Laborda: Anal. Chim. Acta, 1122 (2020), 20. https://doi.org/10.1016/j.aca.2020.04.080
• 8)   H.  Hagendorfer,  R.  Kaegi,  J.  Traber,  S. F. L.  Mertens,  R.  Scherrers,  C.  Ludwig and  A.  Ulrich: Anal. Chim. Acta, 706 (2011), 367. https://doi.org/10.1016/j.aca.2011.08.014
• 9)   D.  Itabashi,  R.  Murao,  S.  Taniguchi,  K.  Mizukami,  H.  Takagi and  M.  Kimura: ISIJ Int., 60 (2020), 979. https://doi.org/10.2355/isijinternational.ISIJINT-2019-387
• 10)   D.  Itabashi,  K.  Mizukami,  S.  Taniguchi,  M.  Nishifuji and  H.  Kamiya: Anal. Sci., 35 (2019), 995. https://doi.org/10.2116/analsci.19P087
• 11)   O.  Meili-Borovinskaya,  F.  Meier,  R.  Drexel,  M.  Baalousha,  L.  Flamigni,  A.  Hegetschweiler and  T.  Kraus: J. Chromatogr. A, 1641 (2021), 461981. https://doi.org/10.1016/j.chroma.2021.461981
• 12)   L.  Fischer,  B.  Zipfel,  G.  Koellensperger,  J.  Kovac,  S.  Bilz,  A.  Kunkel,  C.  Venzago and  S.  Hann: J. Pharm. Biomed. Anal., 95 (2014), 121. https://doi.org/10.1016/j.jpba.2014.02.016
• 13)   M. E.  Lagerström,  M. P.  Field,  M.  Séguret,  L.  Fischer,  S.  Hann and  R. M.  Sherrell: Mar. Chem., 155 (2013), 71. https://doi.org/10.1016/j.marchem.2013.06.001
• 14)   K.  Takimoto,  I.  Taguchi and  R.  Matsumoto: J. Jpn. Inst. Met., 40 (1976), 834. https://doi.org/10.2320/jinstmet1952.40.8_834
• 15)   F.  Kurosawa,  I.  Taguchi and  R.  Matsumoto: J. Jpn. Inst. Met., 43 (1979), 1068. https://doi.org/10.2320/jinstmet1952.43.11_1068
• 16)   J. C.  Giddings: Science, 260 (1993), 1456. https://doi.org/10.1126/science.8502990
• 17)   J. C.  Giddings,  F. J.  Yang and  M. N.  Myers: Science, 193 (1976), 1244. https://doi.org/10.1126/science.959835