Surface & Coatings Technology 368 (2019) 138–146 Contents lists available at ScienceDirect Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat Monolayered versus multilayered electroless NiP coatings: Impact of the T plating approach on the microstructure, mechanical and corrosion properties of the coatings A. Salicio-Paza, H. Grandea, E. Pellicerb, J. Sortb,c, J. Fornellb, R. Offoiachd, M. Lekkad, E. García-Lecinaa, ⁎ a CIDETEC, Paseo Miramón 196, 20014 Donostia - San Sebastián, Spain b Departament de Física, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Spain c Institució Catalana de Recerca i Estudis Avançats (ICREA), Pg. Lluís Companys 23, E-08180 Barcelona, Spain d Polytechnic Department of Engineering and Architecture, University of Udine, Udine, Italy ARTICLE INFO ABSTRACT Keywords: Electroless nickel-phosphorous (NiP) coatings were produced on low carbon steel substrates for a total plating Electroless nickel plating time of 3 h. Different preparation modalities were pursued. Multilayered coatings were produced by stacking Multilayers three layers of the same composition by successive electroless plating with rinsing steps in between. On the other Corrosion resistance hand, coatings termed ‘monolayered’ for the sake of comparison were deposited by one step electroless process, Interface with and without undergoing bath replenishment of the electrolyte during plating. All the samples were sub- Annealing jected to thermal annealing at 400 °C for 1 h under argon atmosphere. The results show that the multilayer approach prevents crack propagation in the as-deposited coatings be- cause the interfaces between layers block the advance of defects. Bath replenishment during monolayered coatings production creates pseudo-interfaces similar to those of the multilayered case but they are ineffective in terms of corrosion protection. Un-replenishment of the electrolyte promotes a change in the coating's micro- structure from lamellar to columnar which severely worsens their performance. Upon annealing, the presence of interfaces, along with the recrystallization of the metallic matrix, promotes an upgrading of the corrosion performance of the multi-layered coatings. The corrosion products spread laterally at the interface where they stockpile. At a certain point, the accumulation of these by-products provokes the exfoliation of the outermost layer exposing the layer underneath to the corrosive media, thereby delaying the advancement of the corrosion attack. The results of this study highlight the importance of the plating approach selection, as well as the need for proper electrolyte maintenance during the production of high-performance electroless coatings. 1. Introduction over the last few years as a realistic alternative to hexavalent chromium [6]. Electroless nickel plating was firstly reported in the seminal paper In electroless processes, the electrons required to reduce the me- by Brenner and Riddell, published in 1946 [1]. Since then, and based on tallic cations (e.g. Ni2+) are provided by a reducing agent, present in their excellent mechanical properties and corrosion resistance, it has the electrolyte, without the assistance of electrical current [7]. The become one of the preferred engineering solutions for high demanding resulting deposit is an alloy because the reducing agent gets in- applications [2]. Recently, the increasingly stringent environmental corporated into the growing layer (e.g., P), thereby enhancing the regulations on the use of hexavalent chromium has pushed, both in properties of the bare metal (e.g. Ni) [8]. Among all the metals that can academia and plating industries, the need for more sustainable alter- be electrolessly plated, nickel and its alloys represent 95% of all the natives to hexavalent chromium [3,4]. Hexavalent hard chromium industrial applications, hypophosphite being the most used reducing deposits stand out for their excellent mechanical properties, including agent [9]. The properties of nickel-phosphorous (NiP) coatings greatly hardness and wear, as well as their superior corrosion resistance in depend on the phosphorous content in the alloy. Accordingly, three functional applications [5]. Electroless nickel coatings have emerged main classes of NiP coatings exist. Low phosphorous coatings (1–5 wt% ⁎ Corresponding author. E-mail address: egarcia@cidetec.es (E. García-Lecina). https://doi.org/10.1016/j.surfcoat.2019.04.013 Received 15 December 2018; Received in revised form 8 March 2019; Accepted 3 April 2019 Available online 04 April 2019 0257-8972/ © 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 P) exhibit very good mechanical properties at the expense of limited agents, buffers and stabiliser present in the electrolyte remain undi- corrosion resistance. High phosphorous coatings (10–14 wt% P) present sclosed due to confidentiality issues. It is worth mentioning that the superior corrosion resistance but poor mechanical properties. Medium electrolyte was organically stabilised, thus avoiding the use of heavy P content coatings (6–9 wt% P) offer a good compromise over the for- metals as stabilisers to meet the criteria established in the RoHS and mers in terms of mechanical properties and corrosion resistance [10]. ELV directives [21]. The pH of the solution was maintained at 6.5 by This versatility in terms of composition derived properties has NH4OH or 10% H2SO4 correction. The working temperature was set at broadened the application fields of electroless nickel, which is indeed 75 °C by means of a PT-1000 probe attached to the magnetic stirrer at the preferred solution in the automotive industry, electronics, petro- ω = 250 rpm (IKA, RTC Basic). All the chemicals used in the prepara- chemical and aeronautics sectors [2,11–14]. These application fields tion of electroless baths were of analytical grade and dissolved in match those in which hard chromium coatings were historically em- deionized (DI) water. The total plating time was 3 h for all coatings. ployed for the enhancement of properties of base materials commonly Three different plating conditions were considered in this work. The employed in engineering applications. Additionally, the performance multilayer approach involved removing the sample from the electrolyte offered for as-deposited electroless NiP coatings can be further en- following the first electroless cycle (1 h), rinsing it in DI water and hanced by subjecting the coatings to thermal annealing. Generally, immersing the coating in H2SO4 10% solution to avoid surface oxida- annealing at 400 °C during 1 h under inert atmosphere offers the max- tion before the second and third electroless steps were attempted. In the imum increase in the properties due to nickel matrix re-crystallization interim, the exhausted components of the electrolyte were replenished, and Ni3P phase precipitation [15]. Meanwhile, crack development and the pH readjusted to its initial value. Electrolyte's replenishment during hard chromium plating has a detrimental effect on their corro- consisted in Ni2+ and H2PO2− addition to the electrolyte from freshly sion resistance, which is even magnified upon annealing [16]. This prepared stock solutions. Complexing agents, stabilisers and buffers underpins the role of electroless nickel as an alternative to hexavalent were also added accordingly. Higher stirring rate was used during re- chromium plating. plenishment events to assure proper solution homogenization. On the In traditional one-step electrodeposition, the properties of the NiP other hand, electroless plating was done only once and the resulting coatings are dictated by the P content, excluding the possibility to bring coatings are generically labelled as ‘monolayer’ for the sake of com- together the properties of NiP layers with dissimilar P content in the parison with the trilayered coatings fabricated on purpose. Chemistry same structure [17]. In recent years, especially in the plating industry, replenishment of the electrolyte while plating was performed hourly in multilayered coatings have gained much attention due to the ad- some cases without withdrawing the coating from the electrolyte. The vantages offered by these coatings in terms of mechanical and elec- resulting coatings are termed ‘rebalanced monolayer (r-monolayer).’ trochemical behaviour [18]. Narayanan et al. reported the production Alternatively, one step plating was also performed without readjust- of electroless multilayered coatings and showed that their corrosion ment of the chemistry of the solution; this type of coatings is named ‘un- resistance improved if properly selecting the outer layer [17]. Similar rebalanced monolayer (ur-monolayer)’. results were found by Gu et al. in multilayered systems by incorporating In order to study the effect of the thermal treatment on the prop- an intermediate electrolytic nickel layer [19]. Recently, Vitry et al. erties of the coatings, all the samples were subjected to a thermal an- showed the beneficial effect of combining NiP and NiB monolayers in nealing process. A tubular furnace (Hobersal, ST1611580) was em- the same deposit. The resulting coatings showed superior wear re- ployed for the annealing of the samples at 400 °C for 1 h time at a sistance while maintaining a good corrosion resistance when compared heating rate of 10° min−1.The process was carried out under argon to single layered coatings [20]. As can be inferred from the above, the atmosphere in order to prevent from surface oxidation. The system was combination of different layers into a coating offers the possibility to properly purged with argon three times before starting the heating enhance the performance, and thus broaden the functionalities, over routine to assure the protective environment inside the SiC tube. Once single layered materials. Despite the progress in the field, little atten- the annealing process was completed, the samples were kept inside so tion has been paid to the effect of the interfaces present in multilayered allowing natural cooling to room temperature. systems. Therefore, the purpose of this investigation is to deepen into the 2.2. Substrate preparation understanding of the impact of the multilayer plating approach, parti- cularly on the role of the interfaces, on the properties of electroless NiP Low carbon steel (AISI-1010) flat foils (75 × 100 × 0.6 mm3) were coatings. For the study, an optimised electroless low phosphorus for- used as working electrodes. The surface/volume ratio was fixed to mulation, combining the best performance in terms of corrosion re- 1.5 dm2 L−1 for all studied conditions. Prior to plating, the substrates sistance while maintaining its excellent mechanical properties, has been were chemically degreased and rinsed thoroughly afterwards. Then, employed (unpublished results). The low phosphorous electroless electrolytic degreasing was carried out with a cathodic/anodic cycle for nickel layers were sequentially stacked to give rise to the multilayered 2 and 1 min, respectively. After rinsing, the samples were etched in an coating. Their mechanical and corrosion resistance properties were HCl 40%: H2SO4 5% (v/v) mixture for 3 min before electroless nickel compared to those of monolayered deposits of the same thickness plating. grown by one step electroless deposition. The effect of bath replenish- ment during electroless nickel plating of the monolayers was also in- 2.3. Characterization vestigated. The impact of the plating approach on the microstructure of the coatings as well as on their mechanical properties and corrosion Surface topography and cross-section morphology of the specimens resistance is assessed. were examined by field-emission scanning electron microscopy (FESEM) (CARL ZEISS ULTRA PLUS). The weight percentage of phos- 2. Materials and methods phorous incorporated into the deposit was determined using an energy dispersive X-ray (Ametek® EDAX, APOLLO X EDX) detector coupled to 2.1. Electroless nickel bath and plating approach the electron microscope. For cross-section analysis, representative coated areas of the substrate were cut and embedded in hot mounting A proprietary low phosphorous electroless nickel electrolyte was epoxy resin. Samples were polished with SiC paper (120–4000 grit) and employed in this investigation. The volume of the electrolyte was 1 l in diamond paste down to 1 μm to achieve a mirror-like finishing of the all the cases. The nickel concentration in the electrolyte was 6.5 g L−1 embedded sample. Chemical etching was carried out in all samples and sodium hypophosphite was used as reducing agent at a con- using an HNO3:CH3COOH 1:1 (v/v) mixture for microstructural ob- centration of 20 g L−1. Nature and concentration range of complexing servation. 139 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Composition of the coatings was analysed by means of glow dis- visible on the surface of the sample. In the case of the r-monolayer charge optical emission spectroscopy (GDOES) using a JY RF-GD coatings, the more vigorous stirring applied during the chemistry re- PROFILER HR instrument (Horiba Jobin-Yvon). During GDOES ex- balance for faster homogenization of the electrolyte, might trigger H2 periments the samples were sputtered at an Ar pressure of 650 Pa under bubbles desorption, giving rise to a defect free surface. Multilayered an applied power of 35 W. Confocal microscopy (Leica Microsystems, coatings are removed from the plating bath and rinsed several times DCM 3D) was used to evaluate surface roughness of the as-deposited between replenishment events, so there is a fresh surface free from coatings as well as surface topography of both as-deposited and an- defects associated to adsorbed species ready to be plated each time. nealed coatings. A blue light source (λ = 470 nm) was employed for Multilayer and r-monolayer conditions showed thickness values of better resolution with objective magnification of 20× and 50× for 22.7 ± 1.7 and 23.4 ± 0.3 μm and a phosphorous content of high resolution images. The obtained images were analysed with the 2.0 ± 0.3 and 2.5 ± 0.2 wt% respectively. In the case of the ur- LeicaMap software (Leica Microsystems). The microstructure was in- monolayered coatings, the thickness was slightly lower showing a vestigated by means of X-ray diffraction (Bruker, D8) using CuKα ra- higher standard deviation standing at 19.2 ± 3.2 μm and having a diation in the Bragg Brentano geometry. Crystallite size was evaluated phosphorous content of 3.0 ± 0.5 wt%. Chemically etched cross sec- using the Scherrer's equation which is implemented in the EVA soft- tions of the coatings were also evaluated as shown in Figs. 1d–f and ware® (Bruker) of the diffractometer. 2d–f. Multilayered coatings showed the characteristic lamellar structure The hardness and reduced Young's modulus of the obtained coatings in each of the three stacks, which is typical of NiP coatings [23]. This were evaluated using a nanoindenter (Anton-Paar, NHT2 model) lamellar structure is assumed to be caused by local fluctuations of the equipped with a Berkovich pyramidal-shaped diamond tip operating phosphorous content due to pH changes in the diffusion layer at the under load control mode. To determine the variation of hardness along vicinity of the catalytic surface. Apart from the lamellar structure, two the coatings' thickness, 45 indentations were made on their cross-sec- interfaces can be clearly distinguished as a result of the interruption of tions, distributed into three separate rows parallel to the substrate the electroless process. As depicted in Fig. 1d, the three different layers surface (i.e., in a region close to the substrate, in the middle, and in a have similar thickness (about 7 μm each). After annealing, the lamellar region close to the air surface of the coating). L1 (line 1), L2 (line 2) and structure becomes a bit ill-defined due to metallic matrix re- L3 (line 3) were located at ~3.5, ~11 and ~17 μm from the substrate, crystallization (Fig. 2d). The interfaces resulting from the on-off elec- respectively. The maximum applied load was 10 mN with a holding troless, though, remain more distinguishable. GDOES analyses clearly time of 10 s, and a loading/unloading rate of 20 mN min−1. Prior to demonstrate the occurrence of these interfaces created throughout the nanoindentation, cross-sections of the samples were embedded in hot coatings' thickness (estimated from sputtering time), as can be seen in resin and were mechanically polished with 1 μm diamond suspension as Fig. 3, in concordance with the phosphorous content profile measured the last step. by EDX and depicted in Table 1. Interestingly, no phosphorous diffusion The corrosion resistance of the coatings was evaluated by electro- occurred during annealing since the stepped profile remained the same chemical techniques by means of a potentiostat/galvanostat (Biologic, as in the as-deposited state (Fig. 3). VMP3). All measurements were conducted in an aerated 3.5 wt% NaCl The cross-section of the NiP monolayered coatings produced from solution at room temperature in a typical three electrode cell config- the un-rebalanced solution (ur-monolayer) was drastically different uration (flat cell; EG&G Princeton Applied Research, Oak Ridge, TN, from the multilayered coatings, evolving from a lamellar to a columnar USA) and with a geometrical working area of 1 cm2. A Pt mesh and Ag/ structure as the deposit builds up (Fig. 1e). This behaviour is assumed AgCl/NaCl (3 M) electrode were used as counter and reference elec- to be associated with the unbalanced chemistry of non-replenished trodes, respectively. At immersion, the samples' corrosion potential, electrolytes at long plating times. The occurrence of defects, some of Ecorr, was measured and recorded until no further changes were ob- them being “V” shaped, and promoted by substrate's roughness, are also served (less than 10 mV h−1). In order to obtain the anodic and visible. Annealing of these coatings had a similar effect on coating's cathodic Tafel slopes (ba, bc), partial potentiodynamic scans microstructure, blurring the original microstructure (Fig. 2e). No ap- (0.1667 mV s−1) on different areas of the same sample were performed, parent changes in phosphorous distribution upon annealing can be starting from the open circuit potential in the anodic or cathodic di- gleaned from GDOES analysis (Fig. 3). rection. The corrosion current density (jcorr) was calculated by means of Fig. 1f shows the cross-section microstructure of the as-deposited r- Stern-Geary relation [22]. monolayer coating obtained from the balanced electrolyte. The char- acteristic lamellar structure is clearly visible as well as two pseudo- 3. Results and discussion interfaces matching the replenishment events performed during plating. Such interfaces are analogous to those observed in the multilayered 3.1. Morphology and composition of obtained coatings coating and are similar to those reported by Vitry et al. [24]. The oc- currence of these interfaces is thought to be associated with unbalanced After the plating process, the surface of all the NiP coatings pre- solution chemistry during replenishment, which leads to a different sented a lustrous silvery-white appearance for the unaided eye. Surface grain growth until the equilibrium is reached again. Defects of the evaluation by FESEM showed that all samples present the typical coatings related to substrate roughness were also clearly identified. globular topography of electroless coatings before (Fig. 1a–c) and after Thus, flat areas of the coatings showed defect free regions whereas annealing (Fig. 2a–c). In all cases they were dense, homogeneous and rough areas were more prone to promote defects. “V” shaped defects relatively smooth as can be inferred from the average surface roughness were also detected, which crossed the coating's thickness almost en- (Sa) values listed in Table 1. The smoothness is maintained after an- tirely, reaching the substrate in many cases, and indicating a decreased nealing. However, a close-up look of the surface topography showed ability for an effective protection of the substrate. After annealing the that the multilayered and r-monolayered exhibit a very similar defect- lamellar structure was almost lost but the pseudo-interfaces originated free surface morphology, whereas small pores and/or defects can be from bath replenishment were still visible (Fig. 2f). Similar to the clearly observed on the surface of the ur-monoloyered ones. A priori, multilayered coatings, cracks associated to substrate roughness one would expect to find more defects on the ur-monolayered coating smoothed and could not be further straightforwardly identified. In produced without electrolyte replenishment because the working piece terms of chemical composition, these coatings showed a similar trend as remains in the electrolyte for longer plating time compared to the the multilayer coatings. The location of the interfaces can be gleaned multilayer case. Long plating times can promote the adsorption of hy- from the profile and the overall phosphorus content is slightly higher. drogen bubbles on the catalytic surface, thus blocking the growth of Again, the distribution of phosphorous inside the coating did not new nickel crystals and ultimately causing the formation of the defects change upon annealing (Fig. 3). 140 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Fig. 1. Surface morphology and corresponding etched cross-sections of as-deposited electroless NiP coatings, respectively, for (a, d) multilayered, (b, e) ur-mono- layered and (c, f) r-monolayered cases. 3.2. XRD analysis Table 1 Coating characterization (thickness, P content and average surface roughness) Fig. 4a shows the diffractogram of the coatings obtained from the for the different electroless nickel coatings in the as-deposited state. three different conditions. The deposits are all nanocrystalline as ex- Condition Thickness P content Surface roughness (Sa) pected from the relatively low phosphorous content in the coatings [7]. (μm) (%wt.) (μm) Diffraction pattern analysis revealed that the peaks can be indexed to Multilayer 22.7 ± 1.7 2.0 ± 0.3 1.8 the face-centred cubic (fcc) phase of Ni. The strongest reflection ob- r-monolayer 23.4 ± 0.3 2.5 ± 0.2 1.7 served at 2θ = 44.5° in all diffractograms corresponds to the (111) ur-monolayer 19.2 ± 3.2 3.0 ± 0.5 1.6 plane. Besides, a weaker reflection at 2θ = 51.8° matching the (220) plane and other characteristic less intense reflections of the fcc Ni phase are observed, in agreement with PDF 065-2865. Only as-deposited ur- coatings are shown in Fig. 5, both in the as-prepared and annealed monolayer coatings showed some slight differences in the (200) plane states. Table 3 lists the corresponding hardness (H) and reduced Young's probably due to a different crystal growth resulted from unbalanced modulus (Er) values. Results indicate that Er and H values are very si- bath chemistry at long plating times. The crystallite size determined milar among the three samples, in agreement with the similar crystal from the Scherrer's equation is around 5–6 nm for all coatings (Table 2). structure and crystallite size measured from XRD diffraction. These Annealing of the deposits promote the re-crystallization of the results also indicates that the mechanical properties are not sensitive to nickel matrix as seen from the narrowing of the diffraction peaks the change from lamellar to columnar growth revealed by the cross- (Fig. 4b). Again, the (111) reflection is by far the most intense. The section FESEM images in the coatings fabricated using the different relative intensity of the (200) plane is lowered for the ur-monolayered synthetic approaches. Typically, low P amounts are added to Ni coat- coating compared to the other two. The precipitation of the tetragonal ings to increase its hardness [25]. The enhanced mechanical perfor- (BCT) Ni3P phase is detected in all cases. Cristal size analysis confirms mance has been attributed to the supersaturated solid solution the re-crystallization of the metallic matrix. As a result, the crystal size strengthening and to grain size refinement caused by P addition fol- increased by almost 10 fold (44–51 nm) after annealing for all studied lowing the Hall-Petch relationship [25]. Upon annealing, an increase in conditions except for the ur-monolayer coating, for which the crystal H and Er can be observed for the three different coatings. This increase size is slightly lower. Yet, the crystal sizes remain in the nanoregime. can be mainly attributed to the precipitation of hard Ni3P intermetallic phase (detected by XRD, Fig. 4b) leading to precipitation hardening. 3.3. Mechanical properties Commonly, formation of Ni3P phase in heat-treated NiP coatings has been reported to effectively increase its hardness [15,25]. The mono- Representative load-unload displacement curves for the different layered coating prepared from the unadjusted solution shows, on Fig. 2. Surface morphology and related chemically etched cross-section of annealed electroless NiP coatings: (a, d) multi-layered, (b, e) ur-monolayered and (c, f) r- monolayered cases. 141 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Fig. 3. Compositional profiles obtained by GDOES analysis for a) as-deposited and b) annealed NiP coatings where a) multilayer, b) ur-monolayer and c) r- monolayer. Fig. 4. XRD patterns of a) as-deposited and b) annealed NiP coatings where a) multilayer, b) ur-monolayer and c) r-monolayer. Table 2 vs Ag/AgCl, the current increases rapidly with the applied potential, Crystallite size obtained using the Scherrer's equation on the (111) peak width. indicating a breakdown of the passivity and the on-set of localized Crystallite size corrosion [28]. No significant differences were observed among the (nm) cathodic branches of the polarization curves. To further understand the corrosion behaviour of the nickel coat- As-deposited Annealed ings, both surface and cross-section observations were carried out by Multilayer 5.6 51.1 FESEM after the potentiodynamic polarizations. In all cases, surface r-monolayer 4.7 44.4 micrographs of as-deposited coatings (Fig. 7a–c) indicate that the cor- ur-monolayer 4.9 32.4 roded areas show intergranular corrosion located at the boundaries of the nodules. A closer examination of the ur-monolayer coating reveals the presence of some debris on the surface of the coatings, not present average, a slightly higher H. Note that hardness values in excess of on the multilayered and r-monolayered coatings. Observation of the 10 GPa are obtained after annealing, which is comparable to the surface at higher magnifications (Fig. 8) indicated that the debris cor- hardness of electrodeposited chromium [26] and the recently published responded to NiP fibres that form the columnar microstructure, de- electrodeposited Fe-25at%W [27]. tached during the electrochemical polarization from the matrix by se- lective dissolution attack at the fibre boundaries. As a result, a fray-like effect along the thickness up to 5 μm depth into the coating's was ob- 3.4. Corrosion resistance served. This feature must be directly linked to the columnar micro- structure shown by the ur-monolayer coating. The Cl− anion pre- Table 4 summarizes the electrochemical corrosion parameters ob- ferentially attacks the coating through the column boundaries, creating tained for the as-deposited and annealed coatings in a chloride-con- multiple attack paths toward the substrate. This phenomenon, not ob- taining electrolyte. As observed in the table, the corrosion potential served on the multi-layered and r-monolayered coatings, points to dif- (Ecorr) was very similar for all the as-deposited electroless nickel sur- ferences in the coating's performance as a function of the plating mode faces (−300 ± 25 mV vs Ag/AgCl), in agreement with their analogous that were not evident from the potentiodynamic tests. chemical composition. Among the as-deposited coatings, multilayered Cross-section evaluation of chemically etched corroded samples was coatings showed a slightly lower corrosion current density although no helpful to understand the corrosion mechanisms operating in the dif- significant differences on polarization resistance neither on the Tafel ferent as-deposited electroless nickel coatings. As shown in Fig. 9a, the slopes were observed. multilayered coating did not fully protect the substrate, but the mi- Potentiodynamic behaviour (Fig. 6a) of all as-deposited nickel crograph shows that the interfaces created in the synthetic approach coatings present similar characteristics. The anodic potentiodynamic hinder the progress of the corrosion attack; thus, some of the cracks curves show two regions: in the first region from the OCP to ca. promoted by the corrosive media become effectively blocked and do −150 mV vs Ag/AgCl, the current density shows a pseudo-passive not reach the substrate. In the case of the ur-monolayer coatings, the corrosion mechanism with an anodic Tafel slope of about 80 ± 5 mV corrosive attack consists of a microselective dissolution at the fibre dec−1. In the second region, at potentials more positive than −200 mV 142 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Fig. 5. Load vs depth curves obtained from nanoindentation tests for as-deposited and annealed NiP coatings where a) multilayer b) ur-monolayer and c) r- monolayer. Table 3 regardless the followed plating approach (Table 4). Thus, after an- Hardness and Young's modulus values of the electroless NiP coatings de- nealing there is a shift of about 150 mV toward more positive values of termined by nanoindentation. the Ecorr, an increase of the polarization resistances and anodic Tafel Condition As-deposited Annealed slopes and a decrease of corrosion current densities, indicating changes in composition, better passivity and a decreased corrosion suscept- H Er H Er ibility. A possible explanation for the generalized improvement of the (GPa) (GPa) (GPa) (GPa) corrosion resistance upon annealing for all coatings relies on the dif- Multilayer L1 9.48 ± 0.17 174.5 ± 2.2 10.42 ± 0.21 214.5 ± 6.4 ferences in crystallinity before and after thermal annealing. As-de- L2 9.40 ± 0.15 166.4 ± 2.2 10.36 ± 0.17 213.3 ± 4.9 posited coatings are nanocrystalline, with crystallite sizes in the range L3 9.15 ± 0.32 140.3 ± 8.6 10.15 ± 0.23 197.0 ± 6.4 of 4.7–5.6 nm, and therefore characterized by a great amount of grain r-monolayer L1 9.80 ± 0.15 171.6 ± 2.3 10.98 ± 0.39 220.6 ± 4.8 boundaries. These grain boundaries trigger the development of micro- L2 9.52 ± 0.23 168.2 ± 2.6 11.11 ± 0.25 218.7 ± 3.8 cells which can pave the way for the advancement of the corrosive L3 9.47 ± 0.09 156.8 ± 3.2 10.40 ± 0.27 201.5 ± 8.3 ur-monolayer L1 9.57 ± 0.17 168.5 ± 2.3 11.15 ± 0.23 215.3 ± 3.8 attack [29]. Annealing of the coatings brings about an increase of al- L2 9.54 ± 0.12 166.6 ± 3.0 11.34 ± 0.17 210.9 ± 4.0 most ten-fold in the crystallite size which diminishes the density of L3 9.44 ± 0.15 151.7 ± 5.2 10.68 ± 0.08 188.6 ± 9.7 grain boundaries and, in turn, improves the corrosion resistance. This reduction in the grain boundaries due to grain coarsening dominates over the effect imparted by Ni3P phase precipitation, which could act in boundaries (Fig. 9b). Finally, in the r-monolayer coatings, the corrosive parallel as active sites for corrosion attack [30]. attack proceeds along the intrinsic defectivity of the coatings. Thus, Corrosion performance of annealed coatings was found to be no- when the Cl− anions reach a defect, the attack progresses directly to the tably different from that of the as-deposited coatings (Fig. 9). Surface substrate, indicating that the pseudo-observed-interfaces of the r- observation revealed a change in the corrosion mechanism from loca- monolayer do not efficiently block the defect advancement. lized intergranular corrosion initiated at the nodule boundaries for the Annealing brings about an enhancement of the corrosion resistance as-deposited coatings to localized pitting corrosion when annealed Table 4 Electrochemical parameters obtained after electrochemical corrosion tests. Condition −Ecorr Rp ba −bc jcorr (mV vs Ag/AgCl) (kΩ cm2) (mV dec−1) (mV dec−1) (μA cm−2) As-deposited coatings Multilayer 282 ± 5 29 ± 1 76 ± 5 92 ± 1 0.63 ± 0.05 r-monolayer 322 ± 8 23 ± 2 84 ± 9 96 ± 2 0.85 ± 0.11 ur-monolayer 332 ± 11 16 ± 1 78 ± 9 90 ± 1 1.14 ± 0.15 Annealed coatings Multilayer 154 ± 9 41 ± 4 109 ± 3 65 ± 3 0.43 ± 0.09 r-monolayer 161 ± 15 31 ± 7 124 ± 9 68 ± 4 0.61 ± 0.15 ur-monolayer 186 ± 13 25 ± 9 129 ± 6 44 ± 4 0.57 ± 0.19 143 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Fig. 6. Polarization curves of a) as-deposited and b) annealed NiP electroless coatings where a) multilayer, b) ur-monolayer and c) r-monolayer. Fig. 7. Surface morphology of (a–c) as-deposited and (d–f) annealed electroless NiP coatings after polarization studies where (a, d) multilayered, (b, e) ur-mono- layered and (c, f) r-monolayered coatings. Fig. 8. Zoomed details of surface morphology of as-deposited ur-monolayered NiP coatings after polarization studies. coatings were tested. Pitting corrosion phenomena was exacerbated in multilayered case. Fig. 9e shows a more homogeneous coating in which the monolayered coatings (Fig. 7d–f). Cross-sections analysis of the the lamellar structure is imperceptible due to metallic matrix re- samples after corrosion tests, also indicated additional differences crystallization. Yet, some defects are still visible across the coating's among the coatings. Annealed multilayered coatings were still pro- surface which, when exposed to the corrosive media, facilitates that tecting the substrate and no sign of red corrosion was observed. The chloride anions progress through these defects ultimately reaching the multilayered coating dissolves layer by layer underlining the layer-type substrate and leading to a complete detachment of V-shaped fragments morphology of the coatings. It is hypothesized that once the corrosive of the coating. On the other hand, Fig. 9f shows the cross-section of an attack proceeds through the first layer, it stops at the interface where annealed r-monolayer coating after polarization. The image shows that the corrosion products spread laterally, leading to exfoliation of the the coating gets progressively dissolved by the corrosive media until it outermost layer and causing the exposure of the layer beneath to the finds a weak point which allows the attack of the substrate by pit for- corrosive media (Fig. 9d). A similar mechanism has been previously mation. proposed for NiP electrodeposited multilayers [31]. Annealing of the These results highlight the importance of chemical solution main- ur-monolayered coatings has a similar behaviour compared to the tenance at long plating times as well as the effect of the plating 144 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 Fig. 9. Chemically etched cross-section analysis after polarization tests of as-deposited (a–c) and annealed (d–f) electroless NiP coatings where (a, d) multilayered, (b, e) ur-monolayered and (c, f) r-monolayered coatings. approach used for the production of high-quality electroless coatings monolayered coatings, while maintaining their mechanical properties. for demanding applications. Acknowledgments 4. Conclusions The authors would like to acknowledge the EU funded COST Action Electroless nickel multilayered and monolayered coatings were MP-1407 “e-minds” for their networking support. One of the authors produced from a proprietary low phosphorous nickel electrolyte. As- also acknowledge COST Action MP-1407 “e-minds” for their financial deposited coatings were nano-crystalline, dense and homogeneous in support for carrying out a STSM at Udine University which greatly all the cases. At higher magnification, though, ur-monolayered coat- contributes to this article. This work was also partially supported by the ings, showed some defects like small pores. The multilayered coatings Spanish Government (MAT2017-86357-C3-1-R, MAT2017-86357-C3-2- were characterized by well-defined interfaces stemming from process R projects), by the Departamento de Desarrollo Económico y interruption, which effectively prevent crack propagation along the Competitividad of the Basque Government (ELKARTEK, KK-2017/ coating thickness, thereby offering better protection ability compared 00096, KK-2018/00108) and Generalitat de Catalunya (2017-SGR- with the monolayered coatings. r-monolayered coatings showed inter- 292). AS acknowledges the PhD programme in Materials Science from faces analogous to the multilayer case, matching the replenishment Universitat Autònoma de Barcelona in which he is enrolled. events performed during plating. Such interfaces do not completely protect the surface and defects are able to traverse the coating's thick- References ness, reaching more easily the substrate without getting blocked at the interfacial region. As result of the unbalanced chemistry of the elec- [1] A. Brenner, G.E. Riddell, Nickel plating on steel by chemical reduction, J. Res. Natl. Bur. Stand. 37 (1946) (1934) 31–34, https://doi.org/10.6028/jres.037.019. trolyte at long plating times, the growth of the ur-monolayered coatings [2] E.J. O'Sullivan, Fundamental and practical aspects of the electroless deposition evolved from lamellar to columnar type. Annealing of the coatings reaction, in: D.M.K. Richard C. Alkire (Ed.), Adv. Electrochem. Sci. Eng, Wiley-VCH, promotes the recrystallization of the metallic matrix, thereby increasing 2001, pp. 225–273. [3] D. Ahmadkhaniha, F. Eriksson, P. Leisner, C. Zanella, Effect of SiC particle size and the crystallite size by tenfold and giving rise to coatings with fewer heat-treatment on microhardness and corrosion resistance of NiP electrodeposited defects or weak points. Interfaces were still visible for both multilayer coatings, J. Alloys Compd. 769 (2018) 1080–1087, https://doi.org/10.1016/j. and r-monolayered coatings but the fine lamellar microstructure was jallcom.2018.08.013. [4] N. Imaz, J.A. Díez, M. Ostra, M. Sarret, E. García-Lecina, Advanced Ni–W coatings almost lost. obtained by the combination of pulse plating and chemometric techniques, Trans. Nanoindentation experiments performed on the as-deposited coat- IMF. 92 (2014) 305–315, https://doi.org/10.1179/0020296714Z.000000000214. ings' cross-sections revealed that hardness and reduced Young's mod- [5] N. Imaz, M. Ostra, M. Vidal, J.A. Díez, M. Sarret, E. García-Lecina, Corrosion be- haviour of chromium coatings obtained by direct and reverse pulse plating elec- ulus were very similar despite the differences observed by FESEM, in trodeposition in NaCl aqueous solution, Corros. Sci. 78 (2014) 251–259, https:// agreement with the similar crystalline structure revealed by XRD ana- doi.org/10.1016/j.corsci.2013.10.005. lysis. Upon annealing, both parameters increased, presumably due to [6] J.N. Balaraju, N. Raman, N.T. Manikandanath, Nanocrystalline electroless nickel poly-alloy deposition: incorporation of W and Mo, Trans. IMF. 92 (2014) 169–176, precipitation of Ni3P. https://doi.org/10.1179/0020296713Z.000000000123. Corrosion analysis of as-deposited coatings indicated intergranular [7] G.G.O. Mallory, J.B. Hajdu, Electroless Plating: Fundamentals and Applications, corrosion located at the surface nodules boundaries. Microstructural (1990), https://doi.org/10.1155/APEC.3.103. characterization revealed significant changes among the coatings. Thus, [8] W. Riedel, Electroless Nickel Plating, ASM International, 1991, https://books. google.es/books?id=XX9xQgAACAAJ. while the interfaces of the multilayered coatings delay the progress of [9] J. Sudagar, J. Lian, W. Sha, Electroless nickel, alloy, composite and nano coatings - the corrosive attack by blocking the advance of the defects produced by a critical review, J. Alloys Compd. 571 (2013) 183–204, https://doi.org/10.1016/j. the corrosive media, in the monolayered coatings the corrosion pro- jallcom.2013.03.107. [10] G.G. Gawrilov, Chemical (Electroless) Nickel Plating, Portcullis Press Ltd., 1979. gress through the coating reaching the substrate, aggravated in the case [11] D. Baudrand, B. Durkin, Automotive applications of electroless nickel, Met. Finish. of ur-monolayered coatings. After annealing, corrosion performance of 96 (1998) 20–23, https://doi.org/10.1016/S0026-0576(98)80080-9. the NiP coatings improved as a result of reduced grain boundary due to [12] F.B. Mainier, M.P.C. Fonseca, S.S.M. Tavares, J.M. Pardal, Quality of electroless Ni- P (nickel-phosphorus) coatings applied in oil production equipment with salinity, J. grain coarsening. Moreover, a change from intergranular to pitting Mater. Sci. Chem. Eng. 1 (2013) 1–8, https://doi.org/10.4236/msce.2013.16001. corrosion was observed. Cross-section evaluation indicated that the [13] G. Cui, S. Liu, K. Wang, Q. Li, G. Wu, Discovering p-doped mechanism in non- interfaces of the annealed multilayered coatings successfully prevent magnetic Ni–P films for HDD substrate: a combined experimental and theoretical study, RSC Adv. 4 (2014) 14663, , https://doi.org/10.1039/c3ra47217e. the progress of the corrosive attack, unlike monolayered coatings. [14] M.C.L. de Oliveira, O.V. Correa, B. Ett, I.J. Sayeg, N.B. de Lima, R.A. Antunes, According to the results, annealed electroless nickel multilayers allows Influence of the tungsten content on surface properties of electroless Ni-W-P coat- obtaining coatings with enhanced corrosion resistance, compared to ings, Mater. Res. 21 (2017) 1–13, https://doi.org/10.1590/1980-5373-MR-2017- 145 A. Salicio-Paz, et al. Surface & Coatings Technology 368 (2019) 138–146 0567. (1985) 7–15, https://doi.org/10.1016/0376-4583(85)90043-3. [15] M. Czagány, P. Baumli, G. Kaptay, The influence of the phosphorous content and [24] V. Vitry, A.F. Kanta, F. Delaunois, Evolution of reactive concentration during bor- heat treatment on the nano-micro-structure, thickness and micro-hardness of elec- ohydride-reduced electroless nickel-boron plating and design of a replenishment troless Ni-P coatings on steel, Appl. Surf. Sci. 423 (2017) 160–169, https://doi.org/ procedure, Ind. Eng. Chem. Res. 51 (2012) 9227–9234, https://doi.org/10.1021/ 10.1016/j.apsusc.2017.06.168. ie202687y. [16] P. De Lima-neto, A.N. Correia, R.A.C. Santana, R.P. Colares, E.B. Barros, [25] H.C. Huang, S.T. Chung, S.J. Pan, W.T. Tsai, C.S. Lin, Microstructure evolution and P.N.S. Casciano, G.L. Vaz, Morphological, structural, microhardness and electro- hardening mechanisms of Ni-P electrodeposits, Surf. Coat. Technol. 205 (2010) chemical characterisations of electrodeposited Cr and Ni–W coatings, Electrochim. 2097–2103, https://doi.org/10.1016/j.surfcoat.2010.08.115. Acta 55 (2010) 2078–2086, https://doi.org/10.1016/j.electacta.2009.11.037. [26] D. Wu, J. Zhang, J.C. Huang, H. Bei, T.G. Nieh, Grain-boundary strengthening in [17] T.S.N. Sankara Narayanan, I. Baskaran, K. Krishnaveni, S. Parthiban, Deposition of nanocrystalline chromium and the Hall-Petch coefficient of body-centered cubic electroless Ni-P graded coatings and evaluation of their corrosion resistance, Surf. metals, Scr. Mater. 68 (2013) 118–121, https://doi.org/10.1016/j.scriptamat.2012. Coat. Technol. 200 (2006) 3438–3445, https://doi.org/10.1016/j.surfcoat.2004. 09.025. 10.014. [27] A. Nicolenco, N. Tsyntsaru, J. Fornell, E. Pellicer, J. Reklaitis, D. Baltrunas, [18] S.R. Anvari, S.M. Monirvaghefi, M.H. Enayati, Novel investigation on nanos- H. Cesiulis, J. Sort, Mapping of magnetic and mechanical properties of Fe-W alloys tructured multilayer and functionally graded Ni-P electroless coatings on stainless electrodeposited from Fe(III)-based glycolate-citrate bath, Mater. Des. 139 (2018) steel, J. Mater. Eng. Perform. 24 (2015) 2373–2381, https://doi.org/10.1007/ 429–438, https://doi.org/10.1016/j.matdes.2017.11.011. s11665-015-1533-y. [28] J.N. Balaraju, V.E. Selvi, V.K.W. Grips, K.S. Rajam, Electrochemical studies on [19] C. Gu, J. Lian, G. Li, L. Niu, Z. Jiang, High corrosion-resistant Ni-P/Ni/Ni-P mul- electroless ternary and quaternary Ni-P based alloys, Electrochim. Acta 52 (2006) tilayer coatings on steel, Surf. Coat. Technol. 197 (2005) 61–67, https://doi.org/10. 1064–1074, https://doi.org/10.1016/j.electacta.2006.07.001. 1016/j.surfcoat.2004.11.004. [29] Y. Gao, Z.J. Zheng, M. Zhu, C.P. Luo, Corrosion resistance of electrolessly deposited [20] V. Vitry, L. Bonin, Formation and characterization of multilayers borohydride and Ni-P and Ni-W-P alloys with various structures, Mater. Sci. Eng. A 381 (2004) hypophosphite reduced electroless nickel deposits, Electrochim. Acta 243 (2017) 98–103, https://doi.org/10.1016/j.msea.2004.04.077. 7–17, https://doi.org/10.1016/j.electacta.2017.04.152. [30] T. Rabizadeh, S.R. Allahkaram, A. Zarebidaki, An investigation on effects of heat [21] B. Zhang, Amorphous and nano alloys electroless depositions, http://www. treatment on corrosion properties of Ni-P electroless nano-coatings, Mater. Des. 31 sciencedirect.com/science/article/pii/B9780128026854000042, (2016). (2010) 3174–3179, https://doi.org/10.1016/j.matdes.2010.02.027. [22] M.A.L.G. Stern, Electrochemical polarization, J. Electrochem. Soc. 104 (1957) 559, [31] L. Elias, K.U. Bhat, A.C. Hegde, Development of nanolaminated multilayer Ni–P https://doi.org/10.1149/1.2428653. alloy coatings for better corrosion protection, RSC Adv. 6 (2016) 34005–34013, [23] C.C. Nee, R. Weil, The banded structure of Ni-P electrodeposits, Surf. Technol. 25 https://doi.org/10.1039/C6RA01547F. 146