A Comparative Study of the Electro-Assisted Grafting of Mono- and Bi-Phosphonic Acids on Nitinol

Abstract

Over the last few years, Nitinol (NiTi) has become one of the most attractive alloy materials for industrial applications. However, its implementation is still problematic due to its surface nickel content, making it sensitive to pitting corrosion. In applications, it is often necessary to modify NiTi surfaces by using organic coatings, which provides new physico-chemical properties as well as functionalities and often contributes to a reinforcement of the alloy corrosion resistance. In this work, we assess the differences between the molecular layers made out of methylphosphonic acid (C₁P) and the bi-phosphonic acid derivatives: (methylimino)dimethylene-bisphophonic acid (MIP₂) and 1-hydroxyethylidene-1,1-diphosphonic acid (HEP₂) using conventional (CG) and electro-assisted (EG) graftings. The surface modifications with the bi-phosphonic derivatives (MIP₂) and (HEP₂) carried out with the EG process lead to denser layers and a reinforced NiTi corrosion resistance.

1. Introduction

Nitinol (NiTi), a nearly equiatomic nickel titanium alloy, is a major material for industrial applications [1,2,3,4]. It presents highly desirable intrinsic properties including heat, impact and corrosion resistance, high fatigue strength [1,5] super-elasticity, and shape memory [6,7,8]. As a consequence, NiTi is used in a wide range of applications such as self-expending stents, surgical endoscopic instruments, atrial septal occlusion devices, orthodontic wires, orthopedic staples, and plates [9,10,11,12,13,14]. Its corrosion resistance results from the presence of a native TiO₂ passivating top-layer [15]. However, this very thin oxide layer still contains nickel inclusions, making it sensitive to pitting corrosion [16,17,18]. In recent years, various NiTi surface modification processes have been investigated, e.g., plasma treatment [19,20,21], chemical vapor deposition [22,23], hydrothermal reinforcement of the oxide layer [24,25], and the formation of organic self-assembled monolayers (SAMs) [26,27,28].

This last process is an attractive way to modify interfacial properties of NiTi, mainly by providing new physico-chemical surface properties such as wettability, protein adsorption resistance, biocompatibility, and/or bioactivity, and possibly reinforce their corrosion resistance. Among the various surface modifiers, organophosphonic acid derivatives robustly graft on metal oxide surfaces through hydrolysis resistant M-O-P bonds [29,30,31,32,33], thereby contributing to a strong reinforcement of the metal surface against aggressive environments [32,34,35,36,37]. A common way to form such SAMs, referred to as conventional grafting (CG), is by direct immersion of the substrate in the organophosphonic solutions at different temperatures. An approach based on the anodic polarization of the substrate immersed in the organophosphonic solution, referred to in the sequel as electro-assisted grafting (EG), has been proposed recently by Mettoki et al. on Ti-6Al-4V [38]. Very recently, NiTi has been modified with organophosphonic derivatives by EG [39,40]. This approach has the advantage of leading to the formation of high quality SAMs within short modification times and at room temperature, which preserves temperature sensitive terminal functions of the organic surface modifiers [40].

Some molecules of interest present bulky terminal functions, which can lead to disordered and less protective layers. The use of mixed monolayers resulting from the co-adsorption of two organophosphonic acid derivatives differing in length and/or terminal functional groups has been shown to improve the corrosion resistance as well as the layer organization [41,42,43,44]. The present experimental work investigates the grafting on NiTi of short tail mono- and bi-phosphonic acid derivatives, which could be an alternative to the mixed monolayers approach, to find the most efficient anchoring group to be used in molecules of biological interests. For this purpose, surface modifications based on bi-phosphonic acid derivatives (i.e. (methylimino)dimethylene-bisphophonic acid, MIP₂) and (1-hydroxyethylidene-1,1-diphosphonic acid, HEP₂)) have been studied using both conventional (CG) and electro-assisted (EG) grafting approaches and compared to a mono-phosphonic acid derivative (methylphosphonic acid, C₁P) (Figure 1).

2. Experimental

2.1. Chemicals

Disolol® (ethanol 99% denatured with isopropanol 2% and butanone 2%, ChemLab, Zedelgem, Belgium), methylphosphonic acid (C₁P, 98%, Sigma-Aldrich, Overijse, Belgium), (methylamino)dimethylene-bisphophonic acid (MIP₂, Specific Polymer, Castries; France), 1-hydroxyethylidene-1,1-diphosphonic acid (HEP₂, 95%, ABCR, Karlsruhe, Germany), potassium nitrate (98%, Overijse, Belgium ), sodium hydroxide (98.5%, Acros Organics, Molinions, France) and sodium chloride (≥ 99.5%, Fluka, Hampton, United States) were used without further purification. All aqueous solutions were prepared in ultra-pure milli-Q water (18.2 MΩ cm).

2.2. Nitinol Substrates Preparation

Nitinol (Ni 56%/Ti 44%) rectangular-shaped plates (20 mm × 10 mm × 0.3 mm), purchased from AMF (Lury sur Arnon, France), were mechanically polished (Ecomet 300 instrument, Buehler, Esslingen, Germany) down to 0.1 µm roughness on silicon carbide papers using diamond pastes (9 µm) (Struers, Champigny sur Marne, France), a mixture of colloidal silica (MasterMet^TM 2, Buehler, Esslingen, Germany) and H₂O₂ (35 wt.% solution in water, Acros Organics, Molinions, France).

Nitinol substrates were cleaned in denatured ethanol under ultrasonication for 15 min, blow dried with nitrogen, and stored under nitrogen for further use. Such samples are referred to here as Bare. Those samples submitted to a 1 h treatment in boiling water are called HT samples.

2.3. Conventional Grafting (CG) of the Organophosphonic Acid Derivatives on Nitinol

The chemical grafting (CG) on HT NiTi was achieved by its immersion (1 h at 90 °C) in 20 mL of 1 mM of the organophosphonic derivative (C₁P, MIP₂, or HEP₂) aqueous solution [40]. The substrates are then ultrasonically cleaned in denatured ethanol for 15 min before being blow dried under nitrogen.

2.4. Electro-Assisted Grafting (EG) of the Organophosphonic Acid Derivatives on Nitinol

Solutions (20.0 mL) composed of 1 mM of the organophosphonic derivative (C₁P, MIP₂ or HEP₂) and 10 mM KNO₃ are prepared in ultra-pure milli-Q water.

Electro-assisted grafting (EG) on the HT surface was obtained by immersion of the substrate in the solution for 10 min, under a voltage of 1.2 V vs. a saturated calomel electrode (SCE) [40] using a Princeton Applied Research VersaSTAT3 potentiostat/galvanostat. For comparison, a similar experiment (EG₀) has been carried out on the HT surface in absence of the organophosphonic derivatives. The substrates are then ultrasonically cleaned in denatured ethanol for 15 min, before being blow dried under nitrogen.

2.5. Substrate Characterization

The modified substrates are characterized by X-ray photoelectron spectroscopy (XPS), static water contact angle measurements (WCA), cyclic voltammetry (CV), linear sweep (LSV) voltammetry, electrochemical impedance spectroscopy (EIS), and roughness measurements. To assess the reproducibility of experiments, all analyses were performed in triplicate. In the case of XPS characterizations, each sample was also analyzed at three different locations.

XPS spectra were recorded on a K-Alpha spectrometer (Thermo Scientific, Waltham, United States) with a monochromatized X-ray K_α radiation (1486.6 eV), then the photoelectrons were collected at 0° with respect to the surface normal and detected with a hemispherical analyzer. The X-ray source spot size on the sample is 200 μm. The analyzer was operated with 200 eV and 50 eV as pass energy for the survey and the high-resolution core levels spectra, respectively. The core levels binding energy (BE) were calibrated with respect to the C1s BE set at 285.0 eV. Spectra were analyzed using a linear combination of Gaussian and Lorentzian curves in a 70/30 ratio. The peak areas were measured on core levels spectra. Quantitative XPS analyses were carried out by calculating the relevant abundance ratios on the basis of the core levels spectra and taking into account the corresponding Scofield sensitivity factors: [45] C1s 1.000, O1s 2.930, P2p 1.920, N1s 1.800, Ti2p 7.910 and Ni2p 22.180.

Static water contact angles were measured using a DIGIDROP (GBX Surface Technology, France) goniometer at room temperature and ambient atmosphere, with a syringe to deliver 2 µL milli-Q water droplets.

Electrochemical experiments were carried out on a Princeton Applied Research, Potentiostat/Galvanostat Model Versastat 3-LC using a three-electrode electrochemical cell with a controlled analysis spot surface (0.28 cm²). NiTi substrates, bare and modified, were used as a working electrode, a platinum foil as a counter electrode, and a saturated calomel electrode (SCE) as a reference. Cyclic voltammograms were recorded from 0.10 to +0.65 V at a scan rate of 20 mV s⁻¹ in 0.1 M NaOH. The blocking factors (BF) of the coatings on NiTi were determined using the formula [46]:

$$\mathrm{BF}=\frac{{\mathrm{a}}_{\mathrm{an},0}-{\mathrm{a}}_{\mathrm{an}}}{{\mathrm{a}}_{\mathrm{an},0}}\times 100$$

(1)

where a_an,0 and a_an are the area of the anodic peaks for the first cycle of bare and modified NiTi substrates, respectively. The corrosion inhibition efficiency (IE) was determined from the corrosion current densities (j_corr) obtained by the formula: [46]

$$\mathrm{IE}=\frac{{\mathrm{j}}_{\mathrm{corr},0}-{\mathrm{j}}_{\mathrm{corr}}}{{\mathrm{j}}_{\mathrm{corr},0}}\times 100$$

(2)

where j_corr,0 and j_corr are the corrosion current densities of bare and modified NiTi substrates, respectively.

Linear sweep voltammograms were recorded from 0.2 V below OCP (measured for 1 h) to + 1.0 V at a scan rate of 1 mV/s in 0.5 M NaCl.

EIS measurements are carried out in the frequency range of 10 kHz to 0.1 Hz at an amplitude of 10 mV peak to peak. The obtained EIS data were presented in Nyquist representation and were fitted to the electrical equivalent circuit shown in Figure 2, where Rel, Rct and Rlayer are the resistances of the electrolyte, of charge transfer, and of the surface layer (oxide layer or oxide/SAM layer), respectively, and where Qdl and Qlayer are the constant phase element (CPE) of the electrochemical double layer and of the surface layer.

The impedance of CPE, used to replace the capacitance element in non-ideal cases, is defined as [46]:

$$Z_{CPE}=\frac{1}{Q{\left(j\omega \right)}^n}$$

(3)

where Q is a frequency-independent parameter of the CPE. Deviation of the exponent $n$ from idealvalue $n$ = 1 indicates the presence of inhomogeneity at the solid/electrolyte interface caused by roughness [47].

2.6. Molecular Modeling

The optimized molecular geometries of C₁P, HEP₂, and MIP₂ were obtained by the molecular editor Avogadro using the molecular mechanics force field MMFF94.

3. Results and Discussion

The organophosphonic grafting through time-saving processes such as the electro-assisted approach was highly desirable either to activate the reaction and to improve the quality of the so-formed organic layer. This process is highly dependent on the number of partial charges carried by the molecule, thus influencing the way it will be attracted to the polarized electrode. For this purpose, the use of bi-phosphonic acid derivatives (MIP₂ and HEP₂) is compared to 1-methylphosphonic acid (C₁P).

Grafting of Organophosphonic Derivatives on HT Nitinol

C₁P, HIP₂ and MIP₂ were used to assess the grafting efficiency on the hydrothermally treated NiTi (HT) via the conventional (CG) and the electro-assisted (EG) processes with a particular attention on the anchoring group size impact on NiTi corrosion resistance. The footprint size, estimated from molecular modeling (Figure 3), of the mono-phosphonic C₁P and the bi-phosphonic derivatives, HIP₂ and MIP₂, are 2.5, 5.6 or 7.3 Å, respectively. These molecules are grafted on HT NiTi plates by the conventional (CG) and electro-assisted (EG) procedures described in Section 2.3 and Section 2.4, respectively.

Figure 4 provides the water contact angle values. It is observed that, comparatively to HT and irrespective of the modification process (CG and EG), these values are slightly higher in the case of C₁P and MIP₂ and similar (or slightly smaller) for HEP₂. This is consistent with the presence of the OH and of the CH₃ head group carried out by HEP₂ and by C₁P and MIP₂, respectively (Figure 1). Moreover, the similar contact angle between the oxidized surface (EG₀ – 22.2° ± 2.7°) and HEP₂-covered NiTi (22.2° ± 4.6°) can be explained by the presence of -OH group on HEP₂ and the potential hydroxylation of the oxide layer through EG₀.

XPS analyses performed on HT, C₁P, MIP₂, and HEP₂ confirm the successful modification of HT NiTi through the appearing signal of P2p and through the general shape of the obtained spectra for the C1s core level, as presented in Figure 5. These samples exhibit peaks at similar energies but with higher intensities in the case of modified NiTi with organophosphonic derivatives, indicating a successful organic modification of the surface alloy. The peaks at energies of 285, 286.5, 287.7, and 289.1 eV correspond respectively to carbon atoms involved in C-H/C-C, C-O/C-P/C-N, C=O, and O-C=O bonds present in the grafted molecules and contributions from atmospheric contaminations. The increase of the peak at 286.5 eV is particularly noticeable in the case of EG-MIP₂ and EG-HEP₂. XPS Ti2p and Ni2p core-levels for CG and EG, regardless of the type of grafted molecule, are consistent with the CG₀ and EG₀ results (Section 2.1), namely the metallic components of the Ti2p and Ni2p peaks are drastically reduced in the case of the EG treatment.

The amount of NiTi is evaluated considering the total concentration of detected nickel and titanium (metallic and oxidized). The P/NiTi ratios are used to evaluate the amount of organic molecules grafted on the NiTi surface. The evolution of these ratios (

Table 1. XPS atomic ratios for HT NiTi, C₁P, MIP₂, and HEP₂ samples.

Sample	P/NiTi	O/NiTi	Ni/NiTi	Ti/NiTi
HT	-	4.27 ± 0.70	0.15 ± 0.01	0.85 ± 0.01
C1P (CG)	0.02 ± 0.01	2.65 ± 0.56	0.18 ± 0.04	0.82 ± 0.04
C1P (EG)	0.05 ± 0.01	3.12 ± 0.21	0.13 ± 0.02	0.87 ± 0.02
MIP2 (CG)	0.09 ± 0.01	2.83 ± 0.23	0.21 ± 0.03	0.79 ± 0.03
MIP2 (EG)	0.14 ± 0.02	3.92 ± 0.36	0.15 ± 0.02	0.85 ± 0.02
HEP2 (CG)	0.08 ± 0.01	2.80 ± 0.23	0.30 ± 0.03	0.70 ± 0.03
HEP2 (EG)	0.12 ± 0.02	3.51 ± 0.90	0.15 ± 0.02	0.85 ± 0.02

) first attests to a better grafting efficiency (higher P/NiTi ratios) for the EG process. The presence of more phosphonic groups and thus more negative charges on the MIP₂ and HEP₂ molecules promotes their migration towards the anodically polarized sample.

Those results also attest for the formation of denser layers with bi-phosphonic than with mono-phosphonic molecules (Figure 6). For the modification with organophosphonic derivatives (Table 1), the Ni/NiTi ratios (0.13–0.15) are lower and O/NiTi (3.12–3.92) ratios are higher for EG comparatively to CG. Ni/NiTi ratios for the CG process are even higher (0.18–0.3) than for the HT treatment (0.15).

The difference in the oxide layer composition between CG and EG processes and the nature of the grafting organophosphonic molecule have a synergetic effect and impact on NiTi corrosion resistance.

Table 2. Values of E_corr, j_corr, IE, BF and R_layer for HT NiTi, C₁P, MIP₂, and HEP₂ samples (obtained by LSV, CV, and EIS).

Sample	Ecorr (mV vs. SCE)	jcorr (nA cm−2)	IE (%)	BF (%)	Rlayer (MΩ cm-2)
HT	−284 ± 88	103.6 ± 22.2	-	-	2.900 ± 0.231
C1P (CG)	−392 ± 20	8.5 ± 2.6	91.8 ± 2.5	12.5 ± 3.2	3.904 ± 0.514
C1P (EG)	−189 ± 10	5.4 ± 1.6	94.8 ± 1.5	97.2 ± 0.5	10.562 ± 0.237
MIP2 (CG)	−463 ± 29	16.6 ± 2.6	83.9 ± 2.5	2.4 ± 0.4	3.149 ± 0.371
MIP2 (EG)	−371 ± 56	6.1 ± 2.4	94.1 ± 2.5	95.9 ± 2.9	9.247 ± 0.216
HEP2 (CG)	−354 ± 11	15.5 ± 2.6	85.1 ± 2.5	4.9 ± 1.1	3.576 ± 0.420
HEP2 (EG)	−182 ± 3	15.3 ± 0.9	94.9 ± 0.9	97.5 ± 0.5	10.934 ± 0.206

lists the corrosion current densities, the corrosion potentials, and the inhibition efficiencies for all NiTi treatments. In particular, it shows the superiority of C₁P and MIP₂ made by EG, which confer to HT-NiTi a very good corrosion protection highlighted by an inhibition efficiency of around 95%.

As attested to by the EIS measurements (Figure 7 and Table 2), this protection results from the increased surface layer resistance (R_layer) from 2.900 MΩ cm⁻² in the case of the HT-NiTi oxide layer to values of around 3 MΩ cm⁻² and 10 MΩ cm⁻² for C₁P, MIP₂, and HEP₂ made by CG and EG, respectively.

The alteration of the oxide layer within the CG process, which results in particular in an increase of Ni content (Table 1), is confirmed by CV measurements (Figure 8). The first CV cycles obtained for the CG-(C₁P, MIP₂, and HEP₂) show a larger oxidation peak, corresponding to a higher oxidation amount of Ti comparatively to HT-NiTi and to the reappearance of the Ni²⁺ to Ni³⁺ oxidation peak, which is more detectable in the case of CG-(HEP₂). A higher blocking factor (BF) is obtained for EG up to 98.4%, resulting from a very low anodic current density and the absence of a nickel oxidation peak. CV, LSV and XPS results confirm the effectiveness of the electro-assisted grafting process (EG) regardless of the type of organophosphonic derivative.

4. Conclusions

Nitinol alloy modified by conventional (CG) and electro-assisted (EG) grafting of three organophosphonic derivatives (C₁P, MIP₂, and HEP₂) with one or two anchoring groups has been compared.

The electro-assisted process, achieved in 10 min and at room temperature by anodic polarization of HT-NiTi has been shown to impact the oxide layer composition as well as the organophosphonic derivatives grafting reaction.

XPS results obtained for EG (C₁P, MIP₂, and HEP₂) point to a successful surface modification by C₁P, MIP₂, and HEP_2, as well as a reinforcement of the oxide layer. The use of molecules with larger footprint size increases the grafting rate and leads to the formation of denser layers, thanks to the presence of more partially negative charge on the molecules.

The impact of the nature of the molecule and the grafting process assessed by electrochemical techniques (LSV and CV) confirm the superiority of the EG process over the CG process. It leads to an enhancement of the corrosion resistance and to the formation of a more blocked surface. The EG process has two synergetic impacts: the reinforcement of the oxide layer and a high grafting efficiency of the organophosphonic derivatives. By contrast, the conventional grafting (CG) leads to a decreased corrosion resistance, with j_corr becoming up to three times higher and up to double the Ni surface content being obtained.

The (electro)grafting of bisphopshonic acid derivatives thus appear to be a promising alternative for mixed monolayers, which were previously used for the grafting of bulky polymerization initiators [44]. In this context, a monolayer made of a polymerization initiating specie deriving from HEP₂ or MIP₂ needs to be investigated.