Professional Documents
Culture Documents
Franzetal Adv Mater Interfaces20202000552
Franzetal Adv Mater Interfaces20202000552
net/publication/342020372
CITATION READS
1 52
8 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Maciej Krystian on 09 June 2020.
www.advmatinterfaces.de
Adv. Mater. Interfaces 2020, 2000552 2000552 (1 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
strong increase of their donor density. This was accompanied Table 1. Grain size (determined from SEM images) and hardness data
by an increased electronic conductivity of the oxide layers under for titanium and Ti6Al4V in the states AR and ECAP.
anodic polarization. Material Grain size ± SD [µm] Hardness ± SD [HV10]
Microstructural changes of the materials associated with
Ti AR 11.1 ± 0.3 177 ± 1
the ECAP treatment are illustrated in Figure 1, that compares
Ti ECAP 1.55 ± 0.02 236 ± 2
Ti6Al4V AR α phase 0.19 ± 0.005 300 ± 3
β phase 1.46 ± 0.02
Ti6Al4V ECAP α phase 0.19 ± 0.003 373 ± 3
β phase n.d.
Adv. Mater. Interfaces 2020, 2000552 2000552 (2 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
Figure 2. Flatband potentials A,B) EFB ± SE and donor densities C,D) Nd ± SE of air-formed passive layers on titanium (Ti) and the bi-phasic alloy
Ti6Al4V both in the AR and ECAP-treated state. Diagrams (A) and (C) present data in 0.133 m phosphate buffers (pH 4.4, 7.4, and 9.2), diagrams (B)
and (D) compare data from phosphate buffer 7.4 (phosphate) with 0.2 m NaClO4 electrolyte (perchlorate) and cell medium. The gray line in diagram
(A) with the assigned text “60 mV per pH” indicates the theoretical slope of EFB per pH curves; ⁎, ⁎⁎, ⁎⁎⁎ for p < 0.05, 0.01, 0.001.
details see Figures S1–S4, Supporting Information). Fitting the (Figure 2A) and the media composition for pH 7.4 (Figure 2B)
resulting intensity at the surface by a Gaussian resulted in oxide are in a range between −100 and −900 mVAgCl with the trend in
layer thickness data of 7 ± 1 and 23 ± 2 nm for Ti in the AR and diagram (Figure 2A) following the theoretical slope of −60 mV
ECAP state, respectively; for Ti6Al4V these values are 6 ± 1 nm per pH.[32] In phosphate buffers the values for ECAP-treated
for AR and 21 ± 1 nm for ECAP. These more than threefold materials are more positive by 300–400 mV compared to the AR
thicker oxide layers for ECAP-treated materials are contrary to state (exception Ti6Al4V ECAP). Additionally, the values for tita-
Garbacz et al.[28,29] who did not find an effect of ECAP treatment nium are significantly more positive compared to Ti6Al4V for
of oxide layer thickness. In addition, a strong modification of the both treatments in all investigated media.
Ti-L fine structure was visible due to the altered oxidation state Comparing the effects on AR and ECAP materials of per-
in the surface area (Figure 1). To increase the signal-to-noise chlorate electrolyte and cell culture medium to phosphate
ratio, spectra are averaged parallel to the surface. Figure 1 shows buffer in Figure 2B, opposing results are found for the different
the Ti-L edge spectra for the individual samples of the oxide sur- materials. While for AR cell culture medium results in more
face layer as well as of the bulk. The absorption peaks at the L3 positive flatband potentials compared to phosphate buffer, for
and L2 edges are due to excitations from the Ti 2p3/2 and 2p1/2 both ECAP materials a ranking in the order of phosphate > per-
core level states into the unoccupied 3d t2g and eg states, respec- chlorate > cell medium is observed. These differences are all
tively. A strong chemical shift between the two spectra (red and significant for Ti, for Ti6Al4V this is true only for phosphate
blue curves in Figure 1) reveals a strong oxidation of Ti in the buffer compared to the two other electrolytes.
surface layer. This also manifests in the fine structures (i.e., the The data for the donor densities in Figure 2C,D show
“sharpening”) of the L3 and L2 peaks. an extremely significant effect of the ECAP treatment.
Flatband potentials and donor densities were derived from Whereas the values for both materials in the AR state are
a Mott-Schottky analysis of the impedance data in analogy to between 2.3 × 1020 cm−3 (for titanium in cell medium) and
ref. [19]. The relative dielectric constant was set to 50 based on 6.1 × 1020 cm−3 (for Ti6Al4V in phosphate buffer 7.4), donor
literature[30,31] and in agreement with earlier work.[19] As no lit- densities increase by more than two orders of magnitude to
erature data are available for oxide layers on Ti6Al4V, a relative 1.8 × 1021 – 1.1 × 1022 cm−3 and 3.6 × 1022 – 2.8 × 1023 cm−3 for
dielectric constant of 50 was also used here. Ti ECAP and Ti6Al4V ECAP, respectively.
With the exception of Ti6Al4V ECAP in cell medium, The donor densities for ECAP-treated Ti6Al4V are at least one
the EFB values as a function of the pH for phosphate buffers order of magnitude higher than the corresponding values for Ti
Adv. Mater. Interfaces 2020, 2000552 2000552 (3 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
ECAP over all studied media. Further, for both materials the same After 300 s at the open circuit potential (EOCP), galvano-
ranking of perchlorate > cell medium≈phosphate (7.4) is observed. static polarization with increasing current densities from 0.5 to
No other groups have so far compared the semiconductor 5 mA cm−2 for 120 s each was applied until a maximum poten-
properties of air-formed oxide layers on ECAP-treated and tial of 10 VAgCl was reached (Figure 3A–D). The dotted lines
AR titanium based materials. Although the flatband poten- with a slope of 2.5 mC V−1 cm−2 in these diagrams, assigned to
tials in Figure 2A,B, are shifted to more positive values due to 100% Faraday efficiency (FarEff), indicate the theoretical gain of
the ECAP processing, the values remain negative to the body the potential given the case that all charge transferred through
potential given with 400 to 500 mVAgCl by Velten et al.[33] and the system is used for oxide formation.[39] Thus, higher poten-
thus in the “biologically safe region”. Other groups, how- tials at comparable time points of polarization can be assigned
ever, report redox potentials measured in plasma as being in to higher Faraday efficiencies for the oxide formation for a given
the range of −334 ± 9 mVAgCl for healthy patients aged below material, corresponding to a lower amount of oxygen formation
60 years.[34] These values increase to with age and/or diseases as a side reaction due to the lower electronic conductivity of the
such as diabetes type 2 or atherosclerosis to values above oxide layer. Further, for galvanostatic polarization the time axis
−280 mVAgCl.[35,36] Given these data, the EFB-values measured for in the diagrams can also be considered as a charge density axis.
both ECAP-treated materials in phosphate buffer pH 7.4 come For phosphate buffer 7.4 (Figure 3A,B) the AR state of both
close the redox potential and indicate potentially increased materials results in different shapes of the potential/time (charge
electronic conductivity of the oxide layers under these condi- density) curves. Whereas a continuously rising curve for current
tions. Interestingly, the flatband potentials determined in cell densities ≥ 1 mA cm−2 is observed for Ti AR, after a similar start
medium become more negative after ECAP treatment, which a voltage maximum is detected for Ti6Al4V AR at the end of the
might be related to the cysteine/cysteine and glutathione/glu- 1.5 mA cm−2 polarization interval between 600 and 650 s. This
tathione disulfide redox systems present in this medium[34] in corresponds to a transferred charge density of ≈350 mC cm−2.
combination with the increased electronic conductivity of the Irrespective of these differences, the Faraday efficiency for the
oxide layers at potentials positive to the flatband potential. formation of anodic oxide of both AR materials is substan-
Although a general increase in the dislocation and grain tially above the corresponding ECAP materials throughout
boundary density is expected as a consequence of an ECAP the whole experiment. These materials show flatter curves
treatment,[37] little is known about specific values for either that start to rise at higher current densities (≥1.5 mA cm−2),
CP titanium or Ti6Al4V as a function of the ECAP treatment indicative of a higher inherent electronic conductivity of the
parameters.[12] For the strong increase in the donor densities of oxide layers.
the air-formed oxide layers shown in Figure 2C,D, we hypoth- For perchlorate electrolyte (Figure 3C,D) the potential/time
esize a correlation with the dislocation densities in the metallic (charge density) curves for Ti AR and Ti ECAP are in general
phase that forms the substrate for the oxides layers. Zhilyaev comparable to those for phosphate buffer. However, the current
et al.,[23] have reported an increase in the dislocation density density at which potentials start to rise continuously (the point
from ≈0.03 × 1015 m−2 for the AR material to above 0.4 × 1015 m−2 where the electronic conductivity of the oxide layer requires the
depending on the location within the ECAP-treated titanium formation of additional oxide to be able to transport all charge)
sample, with an oxygen content corresponding to grade 1 for is shifted toward higher values. This shift is to 1.5 instead of
ECAP conditions similar to those used in our work. Schafler 1 mA cm−2 for Ti AR and to 2 instead of 1.5 mA cm−2 for Ti
et al.[38] report even higher dislocation densities of 2 × 1015 m−2 ECAP, with the slopes for higher current densities being steeper
for titanium grade 2 after an ECAP treatment with 8 passes BC for both materials compared to those in phosphate buffer.
in the same temperature range as used here. For Ti6Al4V an For both Ti6Al4V states a very limited increase of the poten-
increase in the dislocation density by almost an order of mag- tial is observed for polarization in perchlorate electrolyte. A
nitude (0.8 and 7.5 × 1015 m−2) is reported for AR and ECAP voltage maximum similar to the one in phosphate buffer occurs
states for six ECAP passes at 650 °C under conditions identical for Ti6Al4V AR at current densities ranging from 2.5 to 4 mA
to those applied in this work.[26] cm−2 between individual samples. For Ti6Al4V ECAP potentials
Galvanostatic anodic polarization in phosphate buffer 7.4 and stay below 3.5 VAgCl even for the highest current density, inter-
perchlorate electrolyte was applied as an independent method secting the 2 VAgCl line for current densities between 2.5 and
to study the effect of ECAP treatment on the electronic conduc- 3 mA cm−2. Whereas for titanium the Faraday efficiency for the
tivity of oxide layers on the different materials (Figure 3). oxide formation of the AR state is clearly above the ECAP state,
Possible (electrochemical) reactions during polarization can no such conclusion can be drawn for Ti6Al4V.
be described with: Figure 3E shows a quantification of the required charge den-
sities to reach potentials of 5, 8, and 10 VAgCl as a function of
Ti → Ti4+ + 4 e− for the formation of titanium ions in the interface the material treatment (AR or ECAP) and the electrolyte for all
metal/oxide investigated samples. A value of 2500 mC cm−2 with no error
Ti4+ + 2 O2−→ TiO2 and their reaction with oxygen ions formed in the bars in the diagram indicates that the target voltage was not
interface oxide/electrolyte via dissociation of water for reached for the given combination of material and electrolyte.
the oxide formation
For Ti AR the charge densities required to reach potentials
and with: of 5, 8, and 10 VAgCl increase continuously in both electrolytes,
2 H2O → O2 + 4 H+ + 4 e− for the formation of oxygen by oxidation of water in with the values in phosphate buffer being above those of per-
the interface oxide/electrolyte due to the electronic chlorate electrolyte. All data stay well below 1000 mC cm−2 even
conductivity of the oxide layer as side reaction.
for 10 VAgCl. Controversially, the required charge densities for
Adv. Mater. Interfaces 2020, 2000552 2000552 (4 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
Figure 3. Galvanostatic anodic polarization: Typical potential/time (charge density) curves for stepwise galvanostatic polarization with current densi-
ties increasing from 0.5 to 5 mA cm−2 for A,C) titanium and B,D) Ti6Al4V in the AR and ECAP state in phosphate buffer 7.4 (A,B) and perchlorate elec-
trolyte (C,D). The dotted lines correspond to a theoretical curve for 100% Faraday efficiency for oxide formation. Diagram (E) quantifies the potential/
time (charge/density) curves presenting charge density data required to reach different potentials. Columns with a charge density of 2500 mC cm−2
without error bars indicate that this potential was not reached for the given combination of material and electrolyte; ⁎, ⁎⁎, ⁎⁎⁎ for p < 0.05, 0.01, 0.001.
Ti ECAP increase from about 700 mC cm−2 for 5 VAgCl to ≈1500 lyte. For the two ECAP-treated materials, higher charge densi-
and 2000 mC cm−2 for 8 and 10 VAgCl, respectively, and are com- ties are required in both electrolytes to reach given potentials, a
parable between the two media. A similar situation is observed difference that is already pronounced for 5 VAgCl.
for Ti6Al4V AR in both media, but with a parallel shift to higher We thus can rank the electronic conductivities of the
required charge densities for the perchlorate electrolyte and a forming anodic oxide layers in the initial phase of polariza-
strong increase in the required charge densities between 8 and tion in the order of Ti6Al4V ECAP > titanium ECAP > Ti6Al4V
10 VAgCl. While charge densities required for 5 and 8 VAgCl are AR≈titanium AR, which is in perfect agreement with the Nd
comparable between Ti AR and Ti6Al4V AR this is no longer data presented in Figure 2C,D.
the case for 10 VAgCl. This potential is reached for Ti6Al4V AR The increasing slope of the potential/time (charge density)
in phosphate buffer only at a charge density twice as high as curves for ECAP-treated titanium at potentials > 6 VAg/AgCl indi-
the one for Ti AR and not reached at all in perchlorate electro- cates an increasing Faraday efficiency. Anodic oxide formation on
Adv. Mater. Interfaces 2020, 2000552 2000552 (5 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
Adv. Mater. Interfaces 2020, 2000552 2000552 (6 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advmatinterfaces.de
[24] V. V. Stolyarov, Y. T. Zhu, T. C. Lowe, R. K. Islamgaliev, R. Z. Valiev, E. Miller, K. Takanabe, J. Turner, in Photoelectrochem. Water Spliting
Nanostruct. Mater. 1999, 11, 947. Standards, Springer, New York 2013, pp. 63–85.
[25] M. Eftekhari, G. Faraji, S. Nikbakht, R. Rashed, R. Sharifzadeh, [33] D. Velten, V. Biehl, F. Aubertin, B. Valeske, W. Possart, J. Breme,
R. Hildyard, M. Mohammadpour, Mater. Sci. Eng., A 2017, 703, 551. J. Biomed. Mater. Res. 2002, 59, 18.
[26] I. P. Semenova, J. Modina, A. V. Polyakov, G. V. Klevtsov, [34] D. P. Jones, Rejuvenation Res. 2006, 9, 169.
N. A. Klevtsova, I. N. Pigaleva, R. Z. Valiev, Mater. Sci. Eng., A 2019, [35] S. Ashfaq, J. L. Abramson, D. P. Jones, S. D. Rhodes,
743, 581. W. S. Weintraub, W. C. Hooper, V. Vaccarino, D. G. Harrison,
[27] Q. Shi, Y. Y. Tse, R. L. Higginson, Mater. Sci. Eng., A 2016, 651, 248. A. A. Quyyumi, J. Am. Coll. Cardiol. 2006, 47, 1005.
[28] H. Garbacz, M. Pisarek, K. J. Kurzydłowski, Biomol. Eng. 2007, 24, 559. [36] P. S. Samiec, C. Drews-Botsch, E. W. Flagg, J. C. Kurtz, P. Sternberg,
[29] A. Sotniczuk, D. Kuczyńska, D. Kubacka, A. Królikowski, R. L. Reed, D. P. Jones, Free Radical Biol. Med. 1998, 24, 699.
H. Garbacz, Mater. Sci. Technol. 2017, 0836, 1. [37] R. Z. Valiev, Y. Estrin, Z. Horita, T. G. Langdon, M. J. Zehetbauer,
[30] H. Ha, M. Yoshimoto, H. Koinuma, B. Moon, H. Ishiwara, Appl. Y. Zhu, JOM 2016, 68, 1216.
Phys. Lett. 1996, 68, 2965. [38] E. Schafler, K. Nyilas, S. Bernstorff, L. Zeipper, M. Zehetbauer,
[31] G. P. Burns, J. Appl. Phys. 2012, 65, 2095. T. Ungàr, Z. Kristallogr. Suppl. 2006, 2006, 129.
[32] Z. Chen, T. Deutsch, H. Dinh, K. Domen, K. Emery, A. Forman, [39] C. K. Dyer, J. Electrochem. Soc. 1978, 125, 1032.
N. Gaillard, C. Heske, T. Jaramillo, A. Kleiman-Shwarsctein, [40] N. Khalil, J. S. L. Leach, Electrochim. Acta 1986, 31, 1279.
Adv. Mater. Interfaces 2020, 2000552 2000552 (7 of 7) © 2020 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim