Biometals (2011) 24:1099–1114 DOI 10.1007/s10534-011-9469-7
Complexation- and ligand-induced metal release from 316L particles: importance of particle size and crystallographic structure Yolanda Hedberg • Jonas Hedberg • Yi Liu Inger Odnevall Wallinder
•
Received: 23 April 2011 / Accepted: 4 June 2011 / Published online: 18 June 2011 Ó Springer Science+Business Media, LLC. 2011
Abstract Iron, chromium, nickel, and manganese released from gas-atomized AISI 316L stainless steel powders (sized \45 and \4 lm) were investigated in artificial lysosomal fluid (ALF, pH 4.5) and in solutions of its individual inorganic and organic components to determine its most aggressive component, elucidate synergistic effects, and assess release mechanisms, in dependence of surface changes using atomic absorption spectroscopy, Raman, XPS, and voltammetry. Complexation is the main reason for metal release from 316L particles immersed in ALF. Iron was mainly released, while manganese was preferentially released as a consequence of the reduction of manganese oxide on the surface. These processes resulted in highly complexing media in a partial oxidation of trivalent chromium to hexavalent chromium on the surface. The extent of metal release was partially controlled by surface properties (e.g., availability of elements on the surface and structure of the outermost surface) and partially by the complexation capacity of the different metals with the complexing agents of the different media. In general, compared to the coarse powder (\45 lm), the fine (\4 lm) powder displayed significantly higher released amounts of metals per surface area, increased with increased solution complexation
Y. Hedberg J. Hedberg Y. Liu I. O. Wallinder (&) Division of Surface and Corrosion Science, Royal Institute of Technology (KTH), Drottning Kristinas va¨g 51, 10044 Stockholm, Sweden e-mail:
[email protected]
capacity, while less amounts of metals were released into non-complexing solutions. Due to the ferritic structure of lower solubility for nickel of the fine powder, more nickel was released into all solutions compared with the coarser powder. Keywords Stainless steel Powder Complexation Metal release Dissolution Inhalation
Introduction Occupational effects induced by the potential inhalation of ferrochromium and stainless steel particles at production settings have gained significant interest during the last years. Few epidemiological studies assessing potential adverse effects as a result of inhalation of dust particles during stainless steel production exist and show no apparent results (Huvinen et al. 1997, 2002; Nurminen 2005; Moulin et al. 1993, 2000). Some in vitro toxicity studies have been conducted on stainless steel particles with contradictive results (Hedberg et al. 2010a; Lanone et al. 2009). To enable a REACH-compliant chemical safety assessment (EU 2007) of ferrochromium and stainless steel alloys, significant knowledge gaps have recently been addressed and relevant data generated, both from a bioaccessibility and a toxicity perspective (Santonen et al. 2010). The same gas-atomized stainless steel
123
1100
powders (grade AISI 316L) of different size investigated in this study have previously been subject for testing in several in vitro bioaccessibility studies generating quantitative metal release data and aspects of the chemical form of released chromium, which is an important factor for bioavailability (Hedberg et al. 2010a, b; Midander et al. 2010). The finest powder investigated has also been subject for in vitro toxicity testing on cultivated lung cells, i.e., DNA damage, cytotoxicity and hemolysis from an inhalation perspective (Hedberg et al. 2010a), and used in an in vivo inhalation toxicity study on rats (Dhinsa et al. 2008). Relatively low amounts of released iron, chromium, and nickel, predominantly highly complexed and not in a bioavailable form, have previously been observed in artificial lysosomal fluid, ALF (Hedberg et al. 2010a). ALF simulates in a simple way intracellular inflammatory conditions in lung cells following phagocytosis (Stopford et al. 2003). Despite these low amounts of highly complexed metals released, an increase in DNA damage and a non-significant increase in cytotoxicity were observed for the ultrafine (\4 lm) stainless steel powder (Hedberg et al. 2010a). The same powder showed no adverse effects in the in vivo study of rats after 28 days of repeated daily inhalation of aerosol particle dose levels up to 1.0 mg/l (Dhinsa et al. 2008). In ALF, the strongly complexing medium, significant differences in metal release between the ultra-fine (\4 lm) and the coarse 316L powder (\45 lm) were observed, effects not evident in other non-complexing media (Hedberg et al. 2010a, b; Midander et al. 2007, 2010). Five times more metals, normalized to the surface area, were released from the ultra-fine powder compared with the coarse powder, while pure metal particles of iron and chromium showed the expected opposite effect due to agglomeration and equilibrium aspects (Hedberg et al. 2010a). This observation could not be explained by differences in bulk or surface composition, or to particle morphology of the different 316L powders (Hedberg et al. 2010a). Subsequent in-depth studies have therefore been conducted to assess (1) crystallographic structures of the differently sized 316L powders (Hedberg et al. 2011), (2) electrochemical properties in ALF and in other media (unpublished data), and (3) effect of individual chemical components of ALF (6 inorganic and 6 organic chemicals) and certain combinations on the metal release process (this study).
123
Biometals (2011) 24:1099–1114
Metal release can be induced by different physical (e.g., wear), chemical (e.g., protonation), or electrochemical (corrosion-induced) processes or their combinations. A probable mechanism reported for the dissolution of metal oxides in citric acid is as an example complexation-induced reductive dissolution (Carbonaro et al. 2008; Zhang et al. 1985). For most of these processes, the particle surface is of utmost importance. The crystallinity, chemical composition, and order of surface atoms determine the surface layer conductivity (Marcus and Oudar 1995), stability constants with different complexing agents on the surface (Carbonaro et al. 2008), hardness, and dissolution properties of the surface oxide layer (Marcus and Oudar 1995). Since these surface properties are largely influenced by particle characteristics (size, size distribution, shape, surface charge, agglomeration, etc.) as well as the pre-history of the particles (manufacturing, treatments, transport and storage), results of metal release from different particle types and materials are more difficult to compare, as compared with massive materials that can be treated to a defined surface finish, i.e., abrasion, polishing, cleaning, and ageing (Herting 2008; Midander et al. 2006). Complexation-induced metal release has previously been investigated for 316L massive materials (Herting et al. 2006, 2008a, b; Okazaki and Gotoh 2005; Cieslik et al. 2009; Slemnik and Milosˇev 2006; Milosˇev 2002; Milosev and Strehblow 2000; Kocijan et al. 2003) and for iron and chromium compounds (Zhang et al. 1985; Amirbahman et al. 1997; Ballesteros et al. 1998; Dos Santos Afonso et al. 1990; Carbonaro et al. 2008; Schwertmann 1991) using similar complexing agents in different solutions (e.g., citric acid or citrate) as used in this study. The importance of ligand—adsorption on the metal oxide/ solution interface has been addressed (Zhang et al. 1985; Amirbahman et al. 1997; Dos Santos Afonso et al. 1990; Carbonaro et al. 2008) and the importance of stability constants (Slemnik and Milosˇev 2006; Milosˇev 2002; Milosev and Strehblow 2000) highlighted. Synergistic effects of different complexing agents and metal ions have been observed in several of these studies (Amirbahman et al. 1997; Ballesteros et al. 1998; Carbonaro et al. 2008; Dos Santos Afonso et al. 1990). In this study, ultra-fine (\4 lm) and coarse (\45 lm) gas-atomized 316L powders of similar
Biometals (2011) 24:1099–1114
particle history, morphology and bulk composition are compared from a metal release perspective in artificial lysosomal fluid, ALF, investigating the effect of its individual chemical components (19 different solutions) and potential synergistic effects in relation to elemental and chemical surface changes prior to and after exposure. The influence of particle size and structure on the metal release mechanism is moreover discussed as well as complexation of metals with complexing agents. The release of iron, chromium, nickel, and manganese was determined by means of Graphite Furnace Atomic Absorption Spectroscopy, GF-AAS, and surface compositional changes investigated by means of X-ray Photoelectron Spectroscopy, XPS, and Raman spectroscopy. Complexation of chromium by individual organic complexing components within the ALF media was evaluated by means of stripping voltammetry.
Materials and methods Experimental design To enable comparisons and to investigate the effect of the different solutions and individual components of ALF on the extent of metals released from two differently sized 316L powders, one exposure time period was selected. Theoretically, trends and differences of released individual alloy constituents could vary between different exposure time periods. However, previous kinetic studies conducted by the authors in ALF of the differently sized 316L powders and in different solutions have shown the same trends with significant differences between the differently sized particles for all time periods investigated up to 1 week (Midander et al. 2007). Based on these studies, 24 h of immersion were selected since the metal release rate in ALF changes from an initially high rate during the first hours of exposure to a more steady-state-like rate after 24 h of exposure. To assess the individual effect of each chemical component of artificial lysosomal fluid (ALF, pH 4.5) on the extent of metal release from 316L powders, their same concentration as present in ALF were investigated. For the solution of highest concentration, citric acid, the influence of different concentrations was in addition explored. A reference
1101
solution of ultrapure water adjusted to pH 4.5 was prepared in order to investigate the pH-effect exclusively. To assess the effect of ultrapure water isolated from other effects on the extent of metal release, ultrapure water without any pH adjustment was used as a reference solution. To explore any synergistic effects of different chemicals, two different combined solutions were investigated with special focus on potential inhibiting effects of phosphates in combination with accelerating effects of citrate. Therefore, citrate was combined with phosphate (combined solution 1, Table 3) and with all inorganic chemicals present in ALF (combined solution 2, Table 3). In order to enable surface analytical studies of exposed powders combined with metal release and complexation studies, a significant amount of each powder was required. This resulted consequently in a relatively high exposed particle loading (2 g/l). As previously shown (Midander et al. 2006), agglomeration is probable for these powders at such a high loading. This means that the entire surface area is most probably not exposed homogenously to the test media, which may result in lower amounts of released metals per surface area compared with a lower particle loading. Generated results from this study should hence only be used to compare different solutions and particle sizes. Detailed investigations of lower particle loadings in ALF are presented elsewhere (Hedberg et al. 2010a; Midander et al. 2006, 2007, 2010). Material Gas-atomized 316L stainless steel powders of two different particle size fractions, less than 4 lm (93%) and less than 45 lm (87%), were supplied via the European Confederation of Iron and Steel Industries and International Stainless Steel Forum, Belgium. The specific surface areas [measured by the BET method (nitrogen absorption)] were 0.069 and 0.700 m2/g for the coarse (\45 lm) and the ultrafine powder (\4 lm), respectively. Nominal bulk compositional data, provided by the suppliers, are given in Table 1. The shape and surface morphology of both powders are illustrated in Fig. 1. Further more detailed information on powders and instruments is given elsewhere (Hedberg et al. 2010a, b; Midander et al. 2007, 2010).
123
1102
Biometals (2011) 24:1099–1114
Table 1 Nominal bulk composition of 316L powders based on supplier information Fe (wt%)
Cr (wt%)
Ni (wt%)
Mo (wt%)
Mn (wt%)
Si (wt%)
C (wt%)
S (wt%)
\45 lm
68.9
16.8
10.3
2.1
1.4
0.5
0.03
0.01
\4 lm
65.5
18.5
11.6
2.3
1.4
0.65
0.05
0.008
Fig. 1 Scanning electron microscopy (SEM) images of ultra-fine and coarse 316L powders using the same magnification (left: \45 lm; right: \4 lm)— detailed information in Midander et al. (2007)
Metal release studies Metal release studies were conducted by immersing a specific particle loading (2 g/l: 100 mg/50 ml) in different solutions (individual constituents and combinations of different chemicals of artificial lysosomal fluid, ALF, pH 4.5) at 37°C for 24 h, as compiled in Tables 2 and 3. The pH (mostly 4.5 ± 0.1, see Tables 2 and 3) was adjusted by means of sodium hydroxide and/or nitric acid. Differences in pH prior to and after exposure are reported in Tables 2 and 3. All immersion studies were conducted at dark conditions in acid-cleaned polymethylpentene (PMP) NalgeneÒ jars located in mini incubators (Merck) positioned on a bi-linear shaking table (30 cycles per minute, inclination 12°). Triplicate powder samples and one blank sample (without any powders) were exposed for each solution. After exposure, the upper part of the test solution was poured into tubes and the powders separated from the solution by means of centrifugation at 704 rcf (relative centrifugal force) for 10 min. This separation technique has previously been shown to be the most suitable procedure for bioaccessibility studies of similar metal and alloy particles (Midander et al. 2006). To ensure sufficient particle separation in the supernatant sampled from the centrifuged test medium, dynamic light scattering (Malvern Zetasizer nano ZS instrument) measurements were performed. The supernatant, free of particles, was then poured into 25 ml polyethylene storage flasks and acidified to a pH less than 2 prior to metal analysis. In parallel, the solution with particles
123
remaining was centrifuged for 10 min at identical conditions in order to separate the particles from the solution for further surface analysis. Two of the three replicate samples were re-centrifuged with ultrapure water to rinse the particles from any remaining solution chemicals. Rinsed and non-rinsed particles were dried in a desiccator prior to surface analysis (XPS and Raman, see following section). To avoid any metal contamination, all test vessels and experimental equipment were thoroughly acidcleaned with pure 10% HNO3 for at least 24 h and rinsed at least four times with ultra pure water (18.2 MX/cm). All experimental equipment and vessels were dried in ambient laboratory air prior to use. Methods Graphite furnace-atomic absorption spectroscopy Total iron, chromium, nickel, and manganese concentrations (lg/l) were analyzed by means of graphite furnace atomic absorption spectroscopy, GF-AAS (Perkin Elmer AAnalyst 800), or flame atomic absorption spectroscopy (AAS) for higher (mg/l) concentrations. All measurements are based on three replicate readings of each sample. Quality control samples of known concentration were analyzed consecutively. Calibration was done with one blank (ultrapure water) and at least three standards of known concentration, e.g., 50, 100, and 500 lg/l for iron. All results presented are based on measured metal
Biometals (2011) 24:1099–1114
1103
Table 2 Chemical composition of individual (12) constituents of ALF, amount of acids and bases added to adjust the pH, and information on the initial and final solution pH Solution name
Composition
HNO3 added
NaOH added
Initial pH
Final pH (after exposure)
NaCl
NaCl 3.21 g/l
38 ll/l, 5%
0
4.52
4.42 (blank) 4.56–4.60 (\45 lm) 6.17–6.34 (\4 lm)
NaOH
NaOH 6.01 g/l
12.18 ml/l, 65%
1.5 ml/l, 50%
4.50
4.55 (blank) 4.68–4.73 (\45 lm)
2.03 ml/l, 5%
6.24–6.44 (\4 lm) MgCl2
MgCl2 0.0498 g/l
35 ll/l, 5%
0
4.48
4.51–4.53 (blank) 4.66–4.69 (\45 lm) 6.09–6.33 (\4 lm)
Na2SO4
Na2SO4 0.389 g/l
35 ll/l, 5%
0
4.49
4.48–4.53 (blank) 4.65–4.67 (\45 lm) 6.04–6.38 (\4 lm)
NaHPO4
NaHPO4 0.0715 g/l
510 ll/l, 5%
0
4.44
4.41–4.46 (blank) 4.55–4.58 (\45 lm)
CaCl2
CaCl2 0.129 g/l
35 ll/l, 5%
0
4.48
6.04–6.14 (\4 lm) 4.44–4.55 (blank) 4.45–4.57 (\45 lm) 6.09–6.25 (\4 lm)
Citric acid
C6H8O7 20.855 g/l
0
6.933 g/l
4.48
Pyruvate
C3H3O3Na 0.86 g/l
38 ll/l, 5%
0
4.56
4.47 (blank) 4.48 (\45 lm and \ 4 lm) 4.55–4.59 (blank) 4.73–4.76 (\45 lm) 6.35–6.50 (\4 lm)
Glycine
H2NCH2COOH 0.059 g/l
35 ll/l, 5%
0
4.44
4.56–4.62 (blank) 4.69–4.75 (\45 lm) 6.31–6.77 (\4 lm)
Lactate
C3H5NaO3 0.085 g/l
160 ll/l, 5%
0
4.52
4.55–4.59 (blank) 4.59–4.62 (\45 lm) 5.35–5.50 (\4 lm)
Tartrate
C4H4O6Na22H2O 0.091 g/l
170 ll/l, 5%
0
4.54
4.55 (blank) 4.62–4.64 (\45 lm) 5.88–5.92 (\4 lm)
Citrate
C6H5Na3O72H2O 0.077 g/l
420 ll/l, 5%
0
4.53
4.49–4.51 (blank) 4.58–4.63 (\45 lm) 5.68–5.76 (\4 lm)
ALF
Sum of all chemicals above
0
1.7 ml/l, 50%
4.48
4.45–4.46 (blank) 4.46–4.48 (\45 lm) 4.45–4.46 (\4 lm)
concentrations in the supernatants (particles separated) with the contribution from blank reference samples (matrix effects), if any, subtracted. The limits of detection (based on 3 times the standard deviation
of blank samples) were less than 2.4, 1.2, and 1.5 lg/l for iron, chromium, and nickel, respectively (GFAAS); and 18, 15, 39, and 21 lg/l for iron, chromium, nickel, and manganese, respectively (AAS).
123
1104
Biometals (2011) 24:1099–1114
Table 3 Chemical composition of reference solutions (6), partially of different concentration and different combinations of chemicals compared with ALF Solution name
Composition
HNO3 added
NaOH added
Initial pH
Final pH (after exposure)
MQ
Ultrapure water
0
0
5.91
5.92–5.99 (blank) 6.01–6.24 (\45 lm) 6.66–7.03 (\4 lm)
MQ pH 4.5
Ultrapure water
35 ll/l, 5%
0
4.49
No data available
Citric acid 0.05
C6H8O7 0.05 g/l
85 ll/l, 5%
0
4.59
4.43–4.50 (blank) 4.52–4.54 (\45 lm) 5.39–5.56 (\4 lm)
Citric acid 1
C6H8O7 1.0 g/l
1,650 ll/l, 10%
3,400 ll/l, 5%
4.61
4.87–4.88 (blank) 4.87–4.88 (\45 lm) 5.00–5.03 (\4 lm)
Combined solution 1
C6H5Na3O72H2O 0.077 g/l;
995 ll/l, 10%
20 ll/l, 5%
4.43
4.28–4.29 (blank) 4.34–4.40 (\45 lm)
NaHPO4 0.071 g/l
5.13–5.67 (\4 lm) Combined solution 2
C6H5Na3O72H2O 0.077 g/l;
11,060 ll/l, 65%
2,920 ll/l, 50%
NaHPO4 0.071 g/l; NaCl 3.21 g/l;
40 ll/l, 10%
140 ll/l, 5%
4.51
4.38–4.40 (blank) 4.32–4.45 (\45 lm) 5.14–5.15 (\4 lm)
NaOH 6.00 g/l; MgCl2 0.050 g/l; Na2SO4 0.39 g/l; CaCl2 0.128 g/l
Differential pulse adsorptive cathodic stripping voltammetry Blank samples of the solutions pyruvate, glycine, lactate, tartrate, citrate, and citric acid at different concentrations (see Table 2), were investigated on their complexation capacity for chromium, in comparison to ultrapure water. A chromium(VI) standard solution of known concentration was added and the free (active) non-complexed chromium concentration in the solution measured. By adding chromium and comparing the slope of the peak heights (directly proportional to the active chromium concentration in solution) versus the added Cr(VI) concentration, qualitative information on weak complexes can be obtained. Also, quantitative information on any strong complexes can be obtained analyzing the curve characteristics. The complexation capacity for strong complexes is equal with the amount of added chromium(VI) that is not resulting in any chromium(VI) peak (no active chromium measured). Concentrations of active Cr(VI) were determined by means of differential pulse adsorptive cathodic stripping
123
voltammetry (DPAdCSV) using a Metrohm 797 VA Computrace instrument (using a hanging drop mercury electrode working electrode, an Ag/AgCl sat. KCl reference electrode and Pt auxiliary electrode). The supporting electrolyte (2.5 ml added to 10 ml of solution) was 0.2 mol/l sodium acetate, 0.05 mol/l diethylenetriaminepentaacetate (DTPA), and 2.5 mol/ l sodium nitrate. The detection limit was 0.02 lg/l Cr(VI). Analysis was performed with a pre-purging time of 300 s followed by deposition (60 s, 1.0 V) in the differential pulse mode (pulse amplitude 50 mV) starting and ending at -1.0 and -1.45 V, respectively (sweep rate 33.3 mV/s). The pH was adjusted to pH 6.2 ± 0.1 by using 30% NaOH. All solutions (nonacidified) were frozen prior to analysis and all chemicals used were of puriss. p.a. grade. X-ray Photoelectron Spectroscopy (XPS) Compositional analysis of surface oxides of selected samples prior to and after exposure (water-rinsed samples of the solutions ALF, citric acid, NaCl, ultrapure water-MQ pH 4.5, and NaHPO4, see
Biometals (2011) 24:1099–1114
Tables 2 and 3) was performed by means of X-ray photoelectron spectroscopy, XPS (UltraDLD spectrometer from Kratos Analytical, Manchester, UK) using a monochromatic Al X-ray source (150 W) on areas approximately sized 700 9 300 lm. Wide spectra were run to detect elements present in the outermost surface oxide (information depth of a few nanometers) at five different locations, and detailed high resolution spectra (20 eV pass energy) were acquired for the main compositional elements. Experimental details are given elsewhere (Midander et al. 2010). Raman spectroscopy Confocal Raman measurements were performed with a WITec Alpha300 system. Raman spectra were collected ex-situ with an integration time per spectrum of the order of 1 min. A Nikon objective (209 NA 0.4) was used for the measurements together with a pinhole of 50 lm in diameter. Spectra from five different spots (approximately sized 0.5 lm) were collected and averaged into one spectrum for each material. The laser intensity at the surface was carefully adjusted both by defocusing and reducing the output laser intensity in order to avoid laser induced damage of the sample material. This was ensured both by visual inspection of the samples by means of optical microscopy and comparison with previous investigation which showed peak shifts and broadening upon laser heating damage of iron oxides (de Faria et al. 1997).
Results and discussion Metal release mechanisms In the following, the contribution of the three different principal metal release mechanisms, (1) complexation-induced (either adsorption-controlled or controlled by complexation in solution), (2) protonation-induced (solution pH controlled), and (3) electrochemically or corrosion-induced without any complexation, will be discussed for iron, chromium, nickel and manganese released from 316L powders of different size exposed to ALF and its individual chemical components in relation to changes in surface oxide composition.
1105
Complexation-induced reductive metal release and concordant surface compositional changes The influence of the different chemical components of ALF on the extent of the total amount of metal release (iron, chromium, nickel, and manganese) per amount of particles loaded is summarized in Fig. 2 (\4 lm—left, \45 lm—right). The results clearly show that the metal release increases significantly with increasing complexation capacity of the solutions for both particle size fractions. It should be underlined that the concentrations of the constituents of the different solutions are not identical as they reflect their real concentrations in ALF. Na lactate, Na2 tartrate and Na3 citrate have all similar concentration and show a clear trend of increasing metal release with increasing number of functional groups. The highly concentrated citric acid is evidently the predominant constituent of ALF stimulating metal release, an effect most apparent for the ultra-fine powder (\4 lm). The release of metals was strongly selective when normalized to the bulk composition (Fig. 2—top). Manganese was preferentially released and significantly more selectively compared with iron, chromium, and nickel. Chromium and nickel were released only to a minor extent. This observation is strongly related to the surface properties of the 316L powders, factors that will be discussed later. The metal release normalized to the total surface area (BET area) and time of exposure (24 h) is displayed in Fig. 3 for all solutions investigated. The figure includes in addition metal release rates determined for citric acid of different concentration (0.05, 1, and 20.86 g/l), where 0.05 g/l corresponds to the concentration of the Na3 citrate solution. The influence of the complexation capacity of the different solutions on the metal release rate is apparent for iron (Fig. 3a), chromium (Fig. 3b), and nickel (Fig. 3c), in all cases more significant for the ultra-fine powder, while it is of minor importance for the release of manganese (Fig. 3d). Manganese was released to a relatively large extent also in inorganic solutions and in ultrapure water of neutral pH. The release of all metals increased with increasing complexing ability of ligands (increasing number of functional groups of organic chemicals) and their concentration. However, when reaching a certain limit of ligand concentration (approximately 1 g/l for citric acid), the metal release rate increased non-significantly with further increase
123
1106
Biometals (2011) 24:1099–1114 < 45µm, selective release
Mn
Fe
Ni
Cr
< 4µm, selective release
Mn
Fe
Ni
Cr
Fig. 2 Top: Selective release of manganese, iron, nickel, and chromium from 316L powders (left: \45 lm; right: \4 lm) in ALF and MQ pH 4.5 (ultrapure water), normalized to their corresponding bulk metal composition. Bottom: Total metal release per amount of particles loaded (lg/lg) of manganese,
iron, nickel, and chromium from 316L powders into different solutions (left: \45 lm; right: \4 lm), with the blank contribution, if any, subtracted. Inset graphs are of higher magnification
of ligand concentration. These findings indicate an adsorption-induced metal release process, as already proposed in the literature for Fe2O3 particles (Zhang et al. 1985) and for amorphous chromium hydroxide (Carbonaro et al. 2008) in citric acid. Complexation capacities of solutions of citric acid of different concentration towards chromium are displayed in Fig. 4 based on stripping voltammetry measurements. The highest concentrated citric acid solution (16.6 g/l) revealed a complexation capacity of approximately 1.5 lg Cr(VI)/L (strong chromiumligand complexes). For higher chromium concentrations, the solution still complexes chromium to a large extent, but with weaker complexes, indicated by the very small slope. The capacity of citric acid to form strong chromium complexes was approximately 0, 0.16, 0.31, 1.5 lg/l for citric acid concentrations of 0, 3.3, 7.5 and 16.6 g/l, respectively. Thus, the
complexation capacity increased non-linearly with increasing citric acid concentration, which indicates that the process is governed by diffusion and that thermodynamic equilibrium is not attained. As evident from Figs. 2 and 3, the complexationor ligand-induced metal release was most important for iron, followed by nickel, chromium, and to minor extent for manganese, as manganese release showed a weaker dependence on concentration of complexing agents. This trend is in agreement with the stability of complexes formed by manganese, iron, and nickel, as manganese is known to form the least stable complexes (Stumm and Morgan 1996). Stability data for chromium complexes is scarce in the scientific literature. However it has been concluded that chromium most likely can form complexes with organic compounds (Nakayama et al. 1981). While the trend of the complexation dependency between
123
Biometals (2011) 24:1099–1114
1107
Fig. 3 Release rates of iron (a), chromium (b), nickel (c), and manganese (d) (lg/cm2/h) for ultra-fine (\4 lm) and coarse (\45 lm) 316L powders after 24 h of exposure into different solutions, c.f. Tables 2 and 3 for chemical composition, with
the blank contribution, if any, subtracted. The error bars reflect the standard deviation between triplicate samples. The inset graph is of higher magnification
the released metals of iron, chromium, nickel, and manganese can be somewhat related to the stability of the metal and organic complexes, more accurate modeling calculations are not possible. The availability of the different metals on the surface, the crystallinity and surface structure, changes in surface composition and structure over time, and different diffusion mechanisms make it difficult or even impossible to predict or calculate the extent of metal release based on chemical equilibrium constants only. Furthermore, the system is not in equilibrium during the investigated time of immersion (24 h), an effect
clearly seen in previous kinetic metal release studies (Midander et al. 2007). Recent findings have identified the crystallographic structure, measured by means of electron backscattered diffraction of the coarse (\45 lm) 316L powder (Hedberg et al. 2011) as mainly austenitic, and comparable to massive 316L and expected from the bulk composition. The ultra-fine powder (\4 lm) revealed on the other hand a mostly ferritic structure with single crystalline particles, despite a bulk composition which thermodynamically would result in an austenitic structure. As will be
123
1108
Fig. 4 Complexation capacity of citric acid of different concentration towards chromium. The peak height is proportional to active chromium (non-complexed) in solution of known chromium (added) concentration
discussed later, these differences in crystallographic structure are of high importance for the metal release process. Raman spectra are presented in Fig. 5 for both powders prior to and after exposure to citric acid (20.86 g/l). Goethite, a-Fe2O3, was in all cases identified on the surface with bands located at 227, 293, 411, 493, 612, and 1,300 cm-1 (de Faria et al. 1997). The broad band at around 670 cm-1 cannot be unambiguously assigned as several different possible oxides can be attributed to this band, i.e., Fe3O4, MnCr2O4, FeCr2O4, d-FeOOH (de Faria et al. 1997; Lutz et al. 1991; Farrow et al. 1980). In addition, a weak band was occasionally observed at 550 cm-1corresponding to Cr2O3 (Oblonsky and Devine 1995), indicating sample heterogeneity. A band located at 860 cm-1, corresponding to Cr(VI) (Ramsey and McCreery 2004; Sudesh et al. 2006), was observed for both powders exposed to citric acid or ALF, but not after exposure to the other solutions of no or low complexing ability. The particle rinsing procedure with water after exposure was shown to be successful, since non-rinsed particles clearly revealed bands assigned to the solution chemicals (e.g., citric acid), not observed for rinsed powders. This indicates that citric acid does not form strong inner-sphere complexes during the time of immersion in ALF. Previous findings have shown such formation to take longer time than the 24 h (Carbonaro et al. 2008). Furthermore, XPS findings (Fig. 6) revealed no significant change in the outermost (\5 nm) surface
123
Biometals (2011) 24:1099–1114
composition of unexposed and exposed powders in most solutions except for citric acid and ALF. Oxidized iron, chromium, and manganese but no nickel was detected in the outermost surface of unexposed powders. Manganese was strongly enriched compared with the bulk (1.44 wt%) in the outermost surface of unexposed samples with 12.7 wt% Mn/(Fe?Cr?Ni?Mn) for the coarse powder (\45 lm) and with 46.6 wt% for the ultra-fine powder (\4 lm). Since manganese has a high affinity to oxygen (Gaskell 1973) and is known to diffuse to the outermost surface at high temperatures (Lindell and Pettersson 2010), the strong surface enrichment of manganese of the unexposed powders is hence most probably a consequence of the manufacturing procedure (gas atomization). After exposures to ALF and citric acid in which manganese was released to a large extent, no manganese was detected in the outermost surface oxide. Instead, metallic nickel was detected by means of XPS, indicative of its presence beneath a thin surface oxide (Femenia et al. 2004; Marcus and Oudar 1995). Oxidized chromium was strongly enriched in the outermost surface oxide after exposure to these solutions with 50–60 wt% Cr/(Fe?Cr?Ni?Mn) for both particle size fractions in both ALF and citric acid, compared with 17–19 wt% in the bulk. Such enrichment is expected for aggressive solutions and previously reported for massive stainless steels of different grades (Olsson and Landolt 2003; Herting et al. 2005, 2008b). In solutions without high concentrations of citric acid, non-significant surface compositional changes were observed after immersion. Observed surface compositional changes for selected solutions with different extent of metal release (Figs. 2, 3) are presented in Fig. 6. It is evident that the release of low amounts of metals in all solutions, except for the highest concentrated citric acid and ALF, induced minor changes in chemical composition and thickness of the surface oxide (Figs. 5, 6). The relative high release of manganese in the non-aggressive solutions was the only exception. However, this did not result in a surface depletion of manganese. Exposures to citric acid and ALF resulted in contrast to a total surface depletion of manganese, an effect consistent with the high release of manganese into the solutions. According to the Raman investigation, exposures to citric acid and ALF resulted in a surface with chromium present as Cr2O3 but also as hexavalent
Biometals (2011) 24:1099–1114
1109
Fig. 5 Raman spectra for ultra-fine (\4 lm) and coarse (\45 lm) nonexposed and exposed 316L powders, in citric acid (20.86 g/l) for 24 h. The spectra are offset for clarity
chromium, Cr(VI). Cr(VI) has previously been reported to exist in a stable form in the passive layer of stainless steel after treatment in oxidative environments (Brooks et al. 1986; Isaacs et al. 2002). As citric acid causes a high metal release of iron, we propose that this has a destabilizing effect on chromium thereby enabling a redox reaction with oxidized manganese, known to oxidize Cr(III) to Cr(VI) (Richard and Bourg 1991). This hypothesis was supported by XPS findings showing manganese to be totally removed from the surface after exposure in citric acid and ALF. As a consequence, no Cr(VI) was detected on the surface after exposure to the non-aggressive solutions without high concentrations of citric acid. Citric acid and Fe(III) ions have an inhibiting effect on the oxidation of Cr(III) to Cr(VI) (Nakayama et al. 1981), however the reaction can proceed as citric acid also forms stronger complexes with iron resulting in the aforementioned destabilization of the surface oxide (Zhang et al. 1985). This explains why the Cr(VI) peak is not very intense in the Raman spectra, as most of the chromium in hexavalent state is rapidly reduced back to Cr(III) by citric acid and iron. Parallel studies by the authors with chromium metal particles immersed in citric acid detected no Cr(VI) on the surface by means of Raman spectroscopy. This shows that the presence of manganese is instrumental for the oxidation of Cr(III) to Cr(VI). Possible synergistic effects of different chemical constituents of ALF on the metal release from 316L
powders were investigated (see Table 3 for solution compositions). Of particular interest were potential inhibiting effects of phosphate, observed in the literature for iron compounds (Benali et al. 2002; Amirbahman et al. 1997) and massive 316L (Sousa and Barbosa 1991), in combination with the accelerating effect of complexing agents. In the first pooled solution, Na3 citrate was combined with NaHPO4, and in the second solution it was combined with all inorganic chemicals of ALF (in concentrations as in ALF). No significant differences or clear trends were observed for any of the metals released when comparing the metal release in the individual solutions (after subtraction with metal release data in ultrapure water at pH 4.5) with the combined solutions. Protonation and electrochemically induced metal release without complexation Protonation, defined as metal release induced by the relative low pH (4.5) of ALF, was considered to be of minor importance considering the relative small difference in metal release into ultrapure water of neutral pH compared with ultrapure water adjusted to pH 4.5, Figs. 2 and 3. However, the solution pH increased over time, see Table 3. Significant differences in metal release due to this pH effect were only identified for chromium and manganese, and only for the coarse powder (\45 lm) of austenitic structure.
123
1110
Biometals (2011) 24:1099–1114
(Carbonaro et al. 2008; Zhang et al. 1985). The influence of electrochemically induced metal release without the contribution of complexation was considered to be of minor importance since the metal release was significant higher in solutions with high compared to low content of complexing agents, Figs. 2 and 3. Parallel studies conducted to investigate the passivity of the different 316L powders, have shown high open circuit potentials in ALF and relative higher passivity of both powders compared to massive 316L (unpublished data). However, further studies are required, including the influence of oxidized manganese on the surface on the open circuit potential and the oxygen reduction process. Previous findings have shown that oxidized manganese particles on the surface of massive stainless steel can have a large effect (ennoblement) on the electrochemical and corrosion properties of stainless steel (Linhardt 2010). Effect of particle size and crystallographic structure on the metal release process
Fig. 6 Relative chemical composition of iron, chromium, nickel and manganese in the outermost surface oxide (\5 nm) of non-exposed (top) and exposed (24 h immersion in different solutions, rinsed with ultrapure water and dried) 316L powders; a \45 lm; b \4 lm
The reason may be related to the significantly higher amount of oxidized manganese in the outermost surface of the ultrafine powder (c.f. Fig. 6) compared to the coarse powder, and possibly to a more stable form of manganese oxide, e.g., a spinel with chromium. Such a hypothesis would explain both the significantly lower release of manganese from the smaller sized particles and the lack of significant pH effect on the release of manganese for the ultra-fine powder. Changes in pH may however contribute to an improved adsorption of complexing agents and hence be indirectly important for the metal release process
123
Previous sections on the different mechanisms of metal release from 316L powders of different size fractions into ALF have elucidated significant differences both from a metal release and a surface perspective. The complexation-induced metal release process was found to be of significantly higher importance for the ultra-fine powder (\4 lm), compared with the coarse powder (\45 lm), c.f. Fig. 3. For all elements released, the ultra-fine powder displayed an enhanced extent of metal release with increased solution complexation capacity. In the case of released iron, chromium, and manganese, the metal release per surface area was low in all solutions but significantly enhanced for citric acid (1, 20.86 g/l) and ALF, solutions of high complexation capacity. This behavior was especially pronounced for iron and chromium being released to a significantly higher extent per given surface area from the coarse powder compared to the ultra-fine powder in weakly or non- complexing solutions, whereas vice versa in the case of highly complexing solutions, c.f. Fig. 3. This significantly stronger dependence on surface complexation of the ultra-fine, ferritic powder compared with the coarser, austenitic powder could mainly be related to differences in crystallographic structure, important for the
Biometals (2011) 24:1099–1114
adsorption of complexing agents (Carbonaro et al. 2008). In the case of nickel, being an austenite stabilizer, the importance of the crystallographic structure is obvious. Compared to the coarse powder, the release of nickel per surface area from the ultrafine powder of ferritic structure (but with a composition corresponding to austenitic structure) was higher in all solutions despite a lower extent of iron, chromium, and manganese released into non-complexing solutions, c.f. Fig. 3 and the fact that the outermost surface of the ultra-fine powder did not contain any oxidized nickel (also the case for coarse powder), see Figs. 5 and 6. The reason is believed to be related to the lower solubility of nickel in a ferritic structure compared with an austenitic structure (Blair 2005). While both powders were more electrochemically passive compared with massive 316L (unpublished data), the release of iron, chromium, and nickel per surface area in ALF was lower for the coarse powder and higher (in the case of iron) for the ultrafine powder compared with massive 316L (Midander et al. 2007). This could be a result of higher adsorption of complexing agents on the ultra-fine ferritic powder compared with the coarse austenitic powder and massive sheet of 316L. All in all, it is evident that the complexing agents of ALF (increasing complexation capacity with increasing number of functional groups and concentration) were the main components governing the extent of metal release from the 316L powders investigated. Protonation and electrochemically induced metal release (corrosion-induced) without the contribution of complexing agents were considered to play a minor role for this system. A schematic overview of prevailing metal release processes and changes in surface properties after immersion of the 316L powders in ALF is given in Fig. 7. The metal release is depicted as metal-citrate complexes, as those are believed to be the dominating metal– organic complexes, based on metal release and complexation capacity data.
Conclusions Prevailing metal release mechanisms for 316L gasatomized powders when immersed into the highly complexing medium of ALF (artificial lysosomal fluid, pH 4.5) have been discussed and the most
1111 Metal release
[Fe-cit] [Mn-cit]
<4 m
[Cr-cit] [Ni-cit]
ALF
[Fe-cit] [Mn-cit]
< 45 m
[Cr-cit] [Ni-cit]
Surface Mn(ox.)
Cr(ox.) > Fe(ox.) > Ni(met.) Cr(III), Cr(VI)
Fe(ox.)>Cr(ox.) Cr(III)
<4 m
ALF Fe(ox.) > Mn(ox.) Cr(III)
Cr(ox.)
Cr(ox.) > Fe(ox.) > Ni(met.) Cr(III), Cr(VI)
< 45 m
Fig. 7 Schematic overview of the extent and complexation of released metals from the 316L powders of different size distribution immersed in ALF for 24 h and concomitant changes in surface oxide composition
aggressive chemicals from a metal release perspective of this solution identified and assessed from a surface compositional standpoint. In addition, the effect of the different particle size fractions and crystallographic structure of the powders on the metal release mechanisms were investigated and discussed. The following main conclusions were drawn: –
–
Metal complexation, most probably adsorptioncontrolled complexation-induced reductive metal release, is the main process governing the extent of metal release from 316L particles immersed in ALF. Protonation (pH effect) and electrochemically (corrosion) induced metal release without the contribution of complexing agents are considered to be of minor importance. The complexation-induced metal release depends on the concentration of complexing agents and increases non-linearly with increasing concentration, indicative of saturation of possible adsorption sites. The process also depends on the number of functional groups (higher complexation-induced metal release with increasing number of functional groups) and on the chemical species. Manganese is significantly more selectively released compared with iron, chromium, and nickel. All oxidized manganese present at the non-exposed outermost surface was released in strongly complexing solutions by reduction processes. This reduction resulted in the subsequent
123
1112
surface oxidation of trivalent chromium to hexavalent chromium upon immersion in strongly complexing solutions. – Metal release from 316L powders into ALF was partially controlled by surface properties (e.g., availability of elements on the surface and structure of the outermost surface) and partially by the complexation capacity of the different metals with complexing agents of ALF. This resulted in a strongly selective dissolution in ALF (Mn Fe [ Ni & Cr) when compared to the bulk composition. – Changes in elemental and chemical surface composition could be correlated to the extent of metal release, but many aspects such as crystallographic structure and chemical information (e.g., spinels) need to be taken into consideration for accurate interpretation of the correlation between surface changes and metal release. – Differences in particle size distribution and also crystallographic structure (\4 lm: ferritic; \45 lm: austenitic) of the 316L powders strongly influence the extent and mechanisms of metal release. In general, the ultra-fine powder (\4 lm) released significantly higher amounts of metals (normalized to the surface area) with increasing solution complexation capacity, but less metals into non-complexing solutions when compared with the coarse powder (\45 lm). As a result of the ferritic structure of the ultra-fine powder of lower solubility for nickel, more nickel was released into all solutions compared with the coarse powder. Acknowledgments Cusanuswerk, Germany, is highly acknowledged for the financial support of Yolanda Hedberg. The authors are members of the Stockholm Particle Group, an operative network between three universities in Stockholm: Karolinska Institutet, Royal Institute of Technology, and Stockholm University, supported by the Swedish Research Council. Instrumental grants from Knut and Alice Wallenberg foundation (XPS), Jernkontoret and Carl Trygger Foundation (Voltammetric equipment) are acknowledged. The grant from Nils and Dorthi Troe¨dsson Foundation for the combined confocal Raman AFM equipment is gratefully acknowledged. Ashkan Reza Gholi is highly acknowledged for parts of the exposures. Prof. Paul Linhardt, Vienna University of Technology, Austria, is highly acknowledged for his scientific support.
123
Biometals (2011) 24:1099–1114
References Amirbahman A, Sigg L, von Gunten U (1997) Reductive dissolution of Fe(III) (hydr)oxides by cysteine: kinetics and mechanism. J Colloid Interface Sci 194:194–206 Ballesteros MC, Rueda EH, Blesa MA (1998) The influence of iron (II) and (III) on the kinetics of goethite dissolution by EDTA. J Colloid Interface Sci 201:13–19 Benali O, Abdelmoula M, Refait P, Ge´nin J-MR (2002) The behaviour of phosphate ions as corrosion inhibitor for Fe(II)-Fe(III) hydroxycarbonate green rust. Hyperfine Interact 139(140):223–230 Blair M (2005) Corrosion of cast stainless steels, corrosion: materials. In: Cramer SD, Covino BS Jr. ASM Handbook, vol 13B. ASM International, Materials Park, Ohio, pp 78–87 Brooks AR, Clayton CR, Doss K, Lu YC (1986) On the role of Cr in the passivity of stainless steel. J Electrochem Soc 133(12):2459–2464 Carbonaro RF, Gray BN, Whitehead CF, Stone AT (2008) Carboxylate-containing chelating agent interactions with amorphous chromium hydroxide: adsorption and dissolution. Geochim Cosmochim Ac 72(13):3241–3257 Cieslik M, Reczynski W, Janus AM, Engvall K, Socha RP, Kotarba A (2009) Metal release and formation of surface precipitate at stainless steel grade 316 and Hanks solution interface—inflammatory response and surface finishing effects. Corros Sci 51(5):1157–1162 de Faria DLA, Venaˆncio Silva S, de Oliveira MT (1997) Raman microspectroscopy of some iron oxides and oxyhydroxides. J Raman Spectrosc 28(11):873 Dhinsa NK, Griffiths DR, Brooks PN (2008) Stainless steel powder (Grade 316L): twenty-eight day repeated dose exposure inhalation (nose only) toxicity study in the rat, SPL project number: 2476/0001. SafePharm Laboratories Dos Santos Afonso M, Morando PJ, Blesa MA, Banwart S, Stumm W (1990) The reductive dissolution of iron oxides by ascorbate: the role of carboxylate anions in accelerating reductive dissolution. J Colloid Interface Sci 138:74–82 EU (2007) REACH in brief. European Commission Environment Directorate General, http://.ec.europa.eu/environ ment/chemicals/reach/pdf/2007_02_reach_in_brief.pdf Farrow RL, Benner RE, Nagelberg AS, Mattern PL (1980) Characterization of surface oxides by Raman spectroscopy. Thin Solid Films 73(2):353 Femenia M, Pan J, Leygraf C (2004) Characterization of ferrite-austenite boundary region of duplex stainless steels by SAES. J Electrochem Soc 151(11):B581–B585 Gaskell DR (1973) Introduction to metallurgical thermodynamics. McGraw-Hill, New York Hedberg Y, Gustafsson J, Karlsson HL, Mo¨ller L, Odnevall Wallinder I (2010a) Bioaccessibility, bioavailability and toxicity of commercially relevant iron- and chromiumbased particles: in vitro studies with an inhalation perspective. Part Fibre Toxicol 7:23 Hedberg Y, Midander K, Odnevall Wallinder I (2010b) Particles, sweat, and tears: a comparative study on bioaccessibility of ferrochromium alloy and stainless steel particles, the pure
Biometals (2011) 24:1099–1114 metals and their metal oxides, in simulated skin and eye contact. Integr Environ Assess Manag 6(3):456–468 Hedberg Y, Karlsson O, Szakalos P, Odnevall Wallinder I (2011) Ultrafine 316L stainless steel particles with frozenin magnetic structures characterized by means of electron backscattered diffraction. Mater Lett 65(14):2089–2092 Herting G (2008) Bioaccessibility of stainless steels—importance of bulk and surface features. Doctoral Thesis, Royal Institute of Technology (KTH), Stockholm Herting G, Odnevall Wallinder I, Leygraf C (2005) A comparison of release rates of Cr, Ni, and Fe from stainless steel alloys and the pure metals exposed to simulated rain events. J Electrochem Soc 152(1):B23–B29 Herting G, Odnevall Wallinder I, Leygraf C (2006) Factors that influence the release of metals from stainless steels exposed to physiological media. Corros Sci 48(8):2120–2132 Herting G, Odnevall Wallinder I, Leygraf C (2008a) Corrosion-induced release of chromium and iron from ferritic stainless steel grade AISI 430 in simulated food contact. J Food Eng 87(2):291–300 Herting G, Odnevall Wallinder I, Leygraf C (2008b) Corrosion-induced release of the main alloying constituents of manganese–chromium stainless steels in different media. J Environ Monitor 10:1084–1091 Huvinen M, Oksanen L, Kallioma¨ki K, Kallioma¨ki P-L, Moilanen M (1997) Estimation of individual dust exposure by magnetopneumography in stainless steel production. Sci Total Environ 199(1–2):133–139 Huvinen M, Makitie A, Jarventaus H, Wolff H, Stjernvall T, Hovi A, Hirvonen A, Ranta R, Nurminen M, Norppa H (2002) Nasal cell micronuclei, cytology and clinical symptoms in stainless steel production workers exposed to chromium. Mutagenesis 17(5):425–429 Isaacs HS, Virtanen S, Ryan MP, Schmuki P, Oblonsky LJ (2002) Incorporation of Cr in the passive film on Fe from chromate solutions. Electrochim Acta 47(19):3127–3130 Kocijan A, Milosˇev I, Pihlar B (2003) The influence of complexing agent and proteins on the corrosion of stainless steels and their metal components. J Mater Sci Mater Med 14(1):69–77 Lanone S, Rogerieux F, Geys J, Dupont A, Maillot-Marechal E, Boczkowski J, Lacroix G, Hoet P (2009) Comparative toxicity of 24 manufactured nanoparticles in human alveolar epithelial and macrophage cell lines. Part Fibre Toxicol 6(1):14 Lindell D, Pettersson R (2010) Pickling of process-oxidised austenitic stainless steels in HNO3-HF mixed acid. Steel Res Int 81(7):542–551 Linhardt P (2010) Twenty years of experience with corrosion failures caused by manganese oxidizing microorganisms. Mater Corros 61:1034–1039 Lutz HD, Mu¨ller B, Steiner HJ (1991) Lattice vibration spectra. LIX. Single crystal infrared and Raman studies of spinel type oxides. J Solid State Chem 90(1):54–60 Marcus P, Oudar J (eds) (1995) Corrosion mechanisms in theory and practice. Marcel Dekker, Inc., Paris Midander K, Pan J, Leygraf C (2006) Elaboration of a test method for the study of metal release from stainless steel particles in artificial biological media. Corros Sci 48(9):2855–2866
1113 Midander K, Pan J, Odnevall Wallinder I, Leygraf C (2007) Metal release from stainless steel particles in vitro—influence of particle size. J Environ Monitor 9:74–81 Midander K, de Frutos A, Hedberg Y, Darrie G, Odnevall Wallinder I (2010) Bioaccessibility studies of ferrochromium alloy particles for a simulated inhalation scenario: a comparative study with the pure metals and stainless steel. Integr Environ Asses Manag 6(3): 441–455 Milosˇev I (2002) Effect of complexing agents on the electrochemical behaviour of orthopaedic stainless steel in physiological solution. J Appl Electrochem 32(3): 311–320 Milosev I, Strehblow H–H (2000) The behavior of stainless steels in physiological solution containing complexing agent studied by X-ray photoelectron spectroscopy. J Biomed Mater Res 52(2):404–412 Moulin JJ, Wild P, Mantout B, Fournier-Betz M, Mur JM, Smagghe G (1993) Mortality from lung cancer and cardiovascular diseases among stainless-steel producing workers. Cancer Cause Control 4(2):75–81 Moulin JJ, Clavel T, Roy D, Dananche´ B, Marquis N, Fe´votte J, Fontana JM (2000) Risk of lung cancer in workers producing stainless steel and metallic alloys. Int Arch Occup Environ Health 73(3):171–180 Nakayama E, Kuwamoto T, Tsurubo S, Fujinaga T (1981) Chemical speciation of chromium in sea water: part 2. Effects of manganese oxides and reducible organic materials on the redox processes of chromium. Anal Chim Acta 130(2):401–404 Nurminen M (2005) Overview of the human carcinogenicity risk assessment of metallic chromium and trivalent chromium. Internet J Epidemiol 2(1) Oblonsky LJ, Devine TM (1995) A surface enhanced Raman spectroscopic study of the passive films formed in borate buffer on iron, nickel, chromium and stainless steel. Corros Sci 37(1):17–41 Okazaki Y, Gotoh E (2005) Comparison of metal release from various metallic biomaterials in vitro. Biomaterials 26(1):11–21 Olsson C-OA, Landolt D (2003) Passive films on stainless steels—chemistry, structure and growth. Electrochim Acta 48:1093–1104 Ramsey JD, McCreery RL (2004) Raman microscopy of chromate interactions with corroding aluminum alloy 2024-T3. Corros Sci 46(7):1729–1739 Richard FC, Bourg ACM (1991) Aqueous geochemistry of chromium: a review. Water Res 25(7):807–816 Santonen T, Stockmann-Juvala H, Odnevall Wallinder I, Darrie G, Zitting A (2010) Use of read-across in the health risk assessment of ferro-chromium alloys under REACH. In: The Twelfth International Ferro Alloy Congress (INFACON XII), Helsinki (FI), 6–9 June 2010 Schwertmann U (1991) Solubility and dissolution of iron oxides. Plant Soil 130:1–25 Slemnik M, Milosˇev I (2006) An impedance study of two types of stainless steel in Ringer physiological solution containing complexing agents. J Mater Sci Mater Med 17(10):911–918
123
1114 Sousa SR, Barbosa MA (1991) Electrochemistry of AISI 316L stainless steel in calcium phosphate and protein solutions. J Mater Sci Mater Med 2(1):19–26 Stopford W, Turner J, Cappellini D, Brocka T (2003) Bioaccessibility testing of cobalt compounds. J Environ Monitor 5:675–680 Stumm W, Morgan JJ (1996) Aquatic chemistry, 3rd edn. Wiley, New York
123
Biometals (2011) 24:1099–1114 Sudesh TL, Wijesinghe L, Blackwood DJ (2006) Characterisation of passive films on 300 series stainless steels. Appl Surf Sci 253(2):1006–1009 Zhang Y, Kallay N, Matijevic E (1985) Interaction of metal hydrous oxides with chelating agents. 7. Hematite-oxalic acid and -citric acid systems. Langmuir 1:201–206