1 Introduction

Mechanical plating is commonly used to deposit metallic coatings onto small workpieces such as fasteners, nails, and washers at room temperature (Ref 1,2,3). The coating process involves placing the metallic parts, glass beads, metal powders, water, and chemical promoters into a tumbling barrel, and the rotation of the barrel causes the metal powders to peen onto the surface of the parts.

Mechanical plating has been used as an alternative to electroplating, especially on high-strength steel parts, because it does not induce hydrogen embrittlement (Ref 4,5,6). Post-heat treatment is required on electroplated parts, especially hardened steels, to eliminate the risk of hydrogen embrittlement (Ref 7, 8), but it is not mandatory on mechanically plated parts. Nevertheless, heat treatment after the mechanical plating process has remarkably enhanced the corrosion resistance of Zn and ZnSn coatings on steel fasteners (Ref 9,10,11,12). Post-heat treatment resulted in the formation of an iron-zinc (Fe-Zn) intermetallic layer between the substrate and the coating and an increased porosity at the coating. Energy-dispersive X-ray spectroscopy (EDS) results showed the presence of ζ-FeZn13, δ1p-FeZn10, δ1k-FeZn7 and Γ-Fe3Zn10 phases within the intermetallic layer; however, the x-ray diffractometry (XRD) results only confirmed the presence of ζ-FeZn13, δ1p-FeZn10, δ1k-FeZn7 phases but not Γ1-Fe5Zn21 and Γ-Fe3Zn10 (Ref 12), which were expected to occur according to the equilibrium phase diagram (Ref 13). This may be related to the sample preparation, the low volume fraction of Γ1 and Γ phases below the detection limit of XRD, or the peaks of those phases that overlapped with those of the ζ and δ phases.

Cyclic salt spray tests showed that the heat-treated mechanically plated coatings exhibited better corrosion resistance in terms of red rust formation despite the increased porosity, suggesting that the intermetallic layer acted as an effective barrier to the corrosion of the underlying steel substrate (Ref 10, 12). Potentiodynamic polarization tests showed the dissolution of the Fe-Zn intermetallic layer at a corrosion potential higher than that of Zn but lower than that of Fe, hence providing an ultimate barrier to the underlying steel substrate (Ref 11).

Previous works (Ref 10,11,12) that characterized the coatings using standard techniques such as scanning electron microscope (SEM), energy-dispersive x-ray spectroscopy (EDS), and x-ray diffractometry (XRD) have identified that the use of nondestructive advanced diffraction technique is necessary to resolve the metallurgical structure, which ranges from nanoscales to microscales for the pores and intermetallic phases formed, of the new mechanically plated coatings.

1.1 Advanced Diffraction Technique

Small/ultra-small-angle neutron scattering (SANS/USANS) can be used to obtain quantitative data on the inhomogeneities in the sample due to the contrast between the scattering body and the surrounding matrix (Ref 14). When a monochromatic neutron beam hits the sample, the scattering intensity caused by the sample fluctuations or inhomogeneities at the nanoscale will be measured (Fig. 1) (Ref 15). The incident beam is scattered when there is a difference in the scattering length density between the scattering body and the matrix. The scattering length density, ρ, of a scattering body consisting of \(i\) different elements can be calculated by Eq 1.

$$\rho = \frac{{\rho_{{\text{b}}} N_{{\text{A}}} }}{{M_{{\text{w}}} }}\sum\nolimits_{i} {n_{i} b_{i} }$$
(1)

where \({\rho }_{{\text{b}}}\) is the bulk density (g/cm3), \({N}_{{\text{A}}}\) is the Avogadro constant (6.022 × 1023 mol−1), \({M}_{{\text{w}}}\) is the molecular weight or atomic mass (g/mol), \({n}_{i}\) is the number of atoms, and \({b}_{i}\) is the scattering length of atoms.

Fig. 1
figure 1

Schematic representation of small-angle scattering (SAS) experiment (Ref 15)

Full size image

When the neutron scattering is elastic (such that the energy of the incident beam does not change), scattering vector (\(\overrightarrow{q}\)) can be defined from the difference between the incident beam (\(\overrightarrow{{k}_{i}}\)) and scattered beam (\(\overrightarrow{{k}_{{\text{s}}}}\)) at angle 2θ (Fig. 1), i.e., \(\overrightarrow{q}=\overrightarrow{{k}_{{\text{s}}}}-\overrightarrow{{k}_{i}}\). The magnitude of the scattering vector, \(q\), is given as:

$$q = \frac{4\pi \sin \theta }{\lambda }({\text{nm}}^{ - 1} {\text{ or {\AA}}}^{ - 1} )$$
(2)

where λ is the neutron wavelength.

The inhomogeneities in the sample will cause the scattering pattern to show a correlation peak (Ref 15). The dimension \(d\) of the inhomogeneities scattering at the scattering angle \(2\theta\) at peak maximum can be related to Bragg’s law, \(2d{\text{sin}}\theta =\lambda\). By combining the magnitude of the scattering vector and Bragg’s law, the scattering vector can be estimated by Eq 3.

$$q = \frac{2\pi }{d}$$
(3)

In general, q is the inverse of a characteristic length in direct space. As a result, large particles will scatter at small q values and vice versa. The scattered intensity, I(q), as a function of q can be generally described as in Eq 4 (Ref 14).

$$I(q) = N_{p} \cdot V_{p}^{2} \cdot \left( {\Delta p} \right)^{2} \cdot P(q) \cdot S(q) + B_{inc}$$
(4)

where Np is the number of particles, Vp is the volume of particle, Δρ is the scattering length density difference, P(q) is the form factor of particle, S(q) is the structure factor of particle, and Binc is the incoherent background.

Form factor, P(q), is the scattering from within one particle and describes the particle shape, while structure factor, S(q), is the scattering from different particles and contains the information about the positions of the particles with respect to each other (Ref 16). By plotting the curve of the scattering intensity, \(I\left(q\right)\), against q, the distribution of the inhomogeneities (e.g., the intermetallic phases and pore size in this study) can then be determined.

SANS/USANS have been primarily used for research on soft matter; however, their application to metals research is expanding. SANS has been successfully used to gain quantitative information about the size and volume fraction of nanoscale precipitates in Maraging, low-carbon, and high-carbon steels (Ref 17,18,19,20,21,22,23) and pores in plasma-sprayed zirconia coatings (Ref 24,25,26), and enable in situ studies of plasma-sprayed yttria-stabilized zirconia (YSZ) coatings at high temperatures (Ref 25). The technique has also been used to determine the surface roughness of mesoporous silica (Ref 27) and the surface area/roughness of porous magnesium hydroxide [Mg(OH)2/Mg(OD)2] resulting from corrosion (Ref 28, 29). SANS is a potential technique used in corrosion studies to obtain in situ scattering data as a function of exposure time during dealloying corrosion of silver-gold (Ag-Au) (Ref 30, 31) and ex situ data during dealloying of copper–platinum (Cu-Pt) (Ref 32). Ex situ corrosion studies of magnesium alloys showed that SANS has been a valuable technique to observe the evolution of porous magnesium hydroxide and the scattering of the corrosion products enhanced with the incorporation of heavy water (D2O) into NaCl solution (Ref 28, 29). Literature also shows that by combining SANS/USANS (Ref 33,34,35), the information obtained gave a more comprehensive coverage from nanoscales to microscales.

This research work aims to identify the phases in the microstructures of new mechanically plated coatings, their relative proportions, and distribution using SANS and USANS. Resolving the nature and volume fraction of the phases in the intermetallic layer, as well as the distribution and relative proportion of porosity, Zn, and Sn phases in the outer coating, is required to develop a better understanding of the corrosion behavior of the coated fasteners. However, due to the irregular shape of the fastener, which is challenging for existing sample holders, it is therefore not feasible to study the coatings applied on fasteners. In addition, the complexity of the fastener’s shape will cause multiple scattering, leading to challenges in data interpretation. As a result, this study was performed on coated steel disks instead of coated steel fasteners, assuming both shapes underwent the same heat treatment mechanisms.

Nevertheless, SANS/USANS have yet to be used so far to investigate mechanically plated coatings; hence, the findings from this study will provide new insight into the evolution of intermetallic phases in mechanically plated coatings.

2 Materials and Method

2.1 Materials

Zn and ZnSn coatings were mechanically plated on 0.95-mm-thick CA3SN-G steel disk substrates (Table 1) of 18 and 35 mm diameter, specifically for the SANS and USANS experiments, respectively. To reduce multiple scattering, thinner samples are favorable for SANS/USANS measurements, especially for samples that scatter massively (Ref 15, 36). To properly assess the q-range that multiple scattering shows small or no effect, it is beneficial to perform systematic measurements using various combinations of neutron wavelengths and samples with different thicknesses to observe at which q value the scattering stops changing (Ref 36). However, due to the tumbling nature during the mechanical plating process, the substrates must be thick enough to avoid sample damage. As a result, only a 0.95-mm-thick steel disk was used as the substrate in this study.

Table 1 Nominal chemical composition of CA3SN-G grade steel
Full size table

The standard mechanical plating procedure is shown in Fig. 2. Detailed plating procedures were explained in previous work (Ref 11). The thickness of the coating was checked at approximately 50-60 μm before the steel disks were rinsed with water and discharged from the barrel. After the mechanical plating process, the coated steel disks were heat treated to 340 ± 10 °C for 30 and 60 min in an air atmosphere and then cooled naturally to room temperature. Table 2 lists the samples prepared.

Fig. 2
figure 2

Mechanical plating procedure

Full size image
Table 2 Zinc and zinc-tin-coated samples
Full size table

2.2 Metallographic Preparation and Characterization

The coated steel disks were cold-mounted in EpoFix resin and cross sectioned. The cross sections were ground using 1200-grit SiC abrasive papers and polished using 9 μm, 6 μm, 3 μm, and 1 μm diamond suspensions, with ultrasonic cleaning in ethanol in between the polishing. The microstructures of coatings were examined under a ZEISS SUPRA 40VP field emission scanning electron microscope (FESEM).

2.3 Small- and Ultra-Small-Angle Neutron Scattering (SANS and USANS)

The SANS and USANS experiments were performed at Quokka (Ref 37) and Kookaburra (Ref 38) facilities of the Australian Nuclear Science Technology Organisation (ANSTO) at Lucas Heights, respectively. SANS collected Q ranges from 7 × 10−4 to 0.5 Å−1 while USANS collected Q ranges from 1.8 × 10−5 to 10−3 Å−1.

Specimens with 18- and 35-mm diameters were placed into demountable cells with quartz windows for SANS and USANS experiments, respectively (Fig. 3). Quartz windows were used in the demountable cells due to their high transmission and low level of scattering (Ref 15). The thickness of each specimen was measured before being placed into the cell.

Fig. 3
figure 3

Demountable cell with quartz window holding a specimen of (a) 18 mm diameter for SANS and (b) 35 mm diameter for USANS

Full size image

The SANS facility is based on pin-hole collimation geometry (Ref 37). To obtain a full Q-range, 3 detector distances (1.3 m, 12 m, and 20 m with lens optics) were used in this study. The settings were: L1 = 12 m, L2 = 1.3 m, L1 = L2 = 12 m, and L1 = L2 = 20 m, at which L1 is the distance between the monochromatic beam source (velocity selector) and sample, while L2 is the distance between sample and detector (Fig. S1). The size of the beam was 10 mm, while the wavelengths of the beam were 8.1 and 5 Å, respectively, at L1 = 20 m, and 12 and 1.3 m. Scattering data and transmission data for all distances were collected except for L2 = 1.3 m since the transmission does not change with respect to detector distance (however, it changes as the wavelength of neutron changes). Data collection also included the empty cell and blocked (Cd) beam (electronic background) for further subtraction during data analysis.

USANS facility is operated using perfect silicon single crystals (Ref 38) (Fig. S2). A neutron of 2.37 Å wavelength was used with two Si(311) crystals at a Bragg angle of 46.4°. 2.37 Å wavelength was used due to the concern of multiple scattering in the sample. The background threshold was set at 0.26 counts/second.

3 Results and Discussion

3.1 Metallurgical Characterization

Before heat treatment, Zn and ZnSn coatings showed a distinct interface between the coating and the substrate (Fig. 4a and 5a), respectively], but after heat treated for 30 and 60 min, an intermetallic layer formed between the steel substrate and the outer coating (Fig. 4b, c, 5b and c). Before heat treatment, Zn and Sn particles were distributed randomly in ZnSn coating (Fig. 5a); however, after heat treatment, the intermetallic layer formed was Fe-Zn-rich while the outer coating was Sn-rich (Ref 10, 12). The intermetallic layer formed in heat-treated Zn coatings was also Fe-Zn-rich (Ref 11).

Fig. 4
figure 4

Scanning electron micrographs of the cross section of Zn coating (a) before heat treatment, (b) after 30-min heat treatment, and (c) after 60-min heat treatment.

Full size image
Fig. 5
figure 5

Scanning electron micrographs of the cross section of ZnSn coating (a) before heat treatment, (b) after 30-min heat treatment, and (c) after 60-min heat treatment

Full size image

3.2 Small- and Ultra-Small-Angle Neutron Scattering (SANS/USANS)

3.2.1 Data Analysis

The intention of performing SANS and USANS in this study was to obtain a full Q-range by combining both collected data. However, multiple scattering (explained in Supplementary Information) was observed from the data collected, and hence, SANS and USANS data were not combined for further discussion.

Data reduction was performed using macros in IGOR Pro software from NIST (Ref 39), but SANS data reduction was adapted with Quokka-specific parameters (Ref 37, 40). The scattering intensity is rescaled by considering the data collected from transmissions, empty cells, and blocked beams (Ref 37). The scattering intensity data were then reduced to absolute scattering cross sections in a unit of cm-1. The cross section contains information about the inhomogeneities scattering and how they are distributed in the sample (Ref 41). The scattering curves from USANS measurements that covered the low-Q-range from 1.8 × 10−5 to 10−3 Å−1 (Fig. S3 and S4 for Zn- and ZnSn-coated steels, respectively, in Supplementary Information) exhibited multiple scattering at very low-Q and low-Q Guinier regions (< 7 × 10−4 Å−1) and hence will not be discussed here. Only the Porod region (Q-range > 7 × 10−4 Å−1) will be analyzed and discussed. The same goes for the SANS measurements, in which only the high Q-range data collected from the 1.3-m detector distance will be discussed. The full Q-range data of the combined low, medium, and high Q data collected from 3 detector distances, which are 20, 12, and 1.3 m, respectively, are shown in Supplementary Information.

3.2.2 Ultra-small-Angle Neutron Scattering (USANS)

Figure 6 shows the Porod region of the scattering curves of Zn-coated steels from USANS measurements. Based on Porod’s law, a slope value of − 3 represents a rough surface, while − 4 represents a sharp and smooth surface (Ref 41). Hence, the slope values of − 3.8 and − 3.7 [Fig. S4] for Zn and ZnSn coatings (with and without heat treatment), respectively, indicate that the surfaces of the particles from all the interfaces were generally rough. The scattering in the Porod region was reliable compared to that in the low-Q Guinier region (explained in Supplementary Information) because the scattering intensity was very low in the Porod region. The low neutron scattering intensity from the beginning indicates less possibility for the neutron to scatter again and cause multiple scattering.

Fig. 6
figure 6

Porod region of the scattering curves of uncoated steel and Zn-coated steels (without heat treatment and with 30- and 60-min heat treatment) obtained from USANS measurements

Full size image

3.2.3 Small-Angle Neutron Scattering (SANS)

Figure 7 shows the Porod region of high Q-range data collected from the 1.3-m detector distance in SANS measurement for Zn-coated steels, while the one for ZnSn-coated steels is shown in Supplementary Information Fig. S8. The characteristics of the particles can be observed in the Porod region because the scattering intensity was low, and hence, there is less possibility of causing multiple scattering. In addition, the background at the high Q-range was flat, and the scattering showed minor differences among the samples.

Fig. 7
figure 7

Porod region of the scattering curves of uncoated steel and Zn-coated steels (without heat treatment and with 30- and 60-min heat treatment) obtained from SANS measurements

Full size image

In the neutron scattering technique, the differential scattering cross section consists of two possible contributions, which are coherent and incoherent scattering (Ref 41). Coherent scattering is Q-dependent and contains information about the structure of the material, while incoherent scattering is Q-independent and provides no structural information. The Q-independent background was due to the incoherent scattering of neutron that did not provide any structural information (Ref 15, 41). However, the background could also relate to the possible fluorescence of secondary elements in the material (Ref 42).

If a scattering curve contains an unknown amount of constant background, the value of the background, B, can be determined based on Porod’s law (Ref 41, 43):

$$I(Q) = \frac{A}{{Q^{n} }} + B$$
(5)

where A is the Porod’s constant, B is the background constant, and n is the Porod’s slope.

To verify if the flat background is related to incoherent scattering, the background, B, was determined by plotting the Porod plot, IQ4 vs. Q4 (Fig. 8 and S9) following Eq 6.

$$IQ^{4} = A + BQ^{4}$$
(6)
Fig. 8
figure 8

IQ4 vs. Q4 plot of (a) uncoated steel, (b) Zn-coated steel, (c) Zn-coated steel with 30-min heat treatment, and (d) Zn-coated steel with 60-min heat treatment. Slopes, B, from the plots are incoherent scattering background values

Full size image

The background constant, B, obtained from the slope of IQ4 vs. Q4, was then subtracted from the raw intensity data (Fig. 9 and S10) to determine Porod’s slope, n, according to Porod’s law (Eq 5), which is derived into Eq (7).

$$\log \left[ {I(Q) - B} \right] = \log A - n\log Q$$
(7)
Fig. 9
figure 9

Scattering curves with background values subtracted (corrected data) at Porod region for (a) uncoated steel, (b) Zn-coated steel, (c) Zn-coated steel with 30-min heat treatment, and (d) Zn-coated steel with 60-min heat treatment. Porod’s slope, n, was calculated from the corrected data

Full size image

Once the background constant, B, was subtracted, the scattering curve followed a linear slope of approximately − 4 down to the noise level (Fig. 9a), indicating that the background constant had caused the scattering curve leveling off at large Q-values (Ref 43). This, therefore, verified that the incoherent scattering added a constant to the scattering curve.

The background constants, B, obtained from all the samples are between 0.005 and 0.007 (Table 3). These values were compared with the theoretical incoherent scattering of Fe, Zn, Sn, and Fe-Zn intermetallic phases (Table 4) (Ref 44). The calculated background constant of uncoated steel was 0.00584, which did not match the theoretical incoherent scattering value of Fe (0.034 cm−1). This could be due to the presence of small impurities in the sample, e.g., carbon or hydrogen, that introduced additional incoherent scattering (Ref 14, 15, 41). For Zn- and ZnSn-coated steels (without heat treatment), the calculated background constants were slightly higher (0.006 from Table 3) than that of the incoherent scattering of Zn (0.005 cm−1 from Table 4), and this could be due to the trapped water or moist within the highly porous coating, hence leading to a higher incoherent scattering. However, with heat treatment, the background constants of the Zn-coated steels (0.00556 and 0.00511 for 30- and 60-min heat treatment, respectively) matched with the incoherent scattering of Zn (0.005 cm−1), most probably due to the removal of moist or trapped water during the heat treatment process. Hence, it can be assumed that the flat background in the scattering curves of Zn- and ZnSn-coated steels was due to the incoherent scattering of the Zn.

Table 3 Background constant, B, and Porod’s slope, n, calculated from the Porod region
Full size table
Table 4 Scattering length density and neutron incoherent scattering cross section of Fe, Zn, Sn, and Fe-Zn intermetallic phases (Ref 44)
Full size table

Besides proving the presence of a constant background in the scattering curve, the calculated Porod’s slope, n, also exhibits information about the surface of the particles (Ref 14, 41, 43). Table 3 summarizes the background constant, B, and the Porod’s slope, n, calculated from the Porod region. The n values for all the samples were between − 3 and − 4, indicating the fractal systems with a dense core and fractal structures on the surface (Ref 14). There were differences between the n values of uncoated steel and Zn- and ZnSn-coated steels (with and without heat treatment), revealing the occurrence of changes at the interfaces. The Porod’s slope for uncoated steel was − 3.87, corresponding to a near-smooth particle surface in steel (Ref 41). With Zn and ZnSn coating on the steels, the slope value changed to approximately − 3.6, indicating that the Zn and Sn particles contributed to the overall particles’ surface roughness. This was supported by the SEM analysis on the cross-sectional samples (Fig. 4a and 5a) that showed the rough surface coatings. After heat treatment, the slope values for Zn- and ZnSn-coated steels changed to approximately between − 3.7 and − 3.9, showing the contributions from the Fe-Zn intermetallic layer on the particles’ surface fractals to a nearly smooth one. This was supported by the dense and smooth Fe-Zn intermetallic layer observed under SEM (Fig. 4b, c and 5b, c).

To support the findings, in which the calculated Porod’s slope exhibited the contributions from other interfaces, scattering curves of Zn- and ZnSn-coated steels (with and without heat treatment) in the Porod region were re-plotted after the subtraction of the background constants, B (Fig. 10). The curves display differences between the scattering (as seen from Q-range 0.04 to 0.1 Å−1), hence validating the presence of other interfaces.

Fig. 10
figure 10

Scattering curves in the Porod region after the subtraction of background constants, B, for (a) uncoated steel and Zn-coated steels (with and without heat treatment), and (b) uncoated steel and ZnSn-coated steels (with and without heat treatment)

Full size image

In summary, the main challenge with the interpretation of data in this study is the presence of multiple contributions to the scattering, i.e., from the steel matrix, Fe-Zn intermetallic phases (possibly ζ-FeZn13, δ1p-Fe13Zn126, δ1k-FeZn7, Γ1-Fe11Zn40, and Γ-Fe3Zn10), and Zn and Sn particles. The scattering intensity from these contributions overlapped with each other, while the main contribution was from the steel matrix. Though there are limitations with the interpretation of data to quantify the specific Fe-Zn intermetallic phases, the data still provides valuable information about the presence of other interfaces based on the differences between the calculated Porod’s slope, n, and the scattering curves after the subtraction of background constant, B.

4 Conclusions

In this study, advanced characterization techniques, namely small-angle neutron scattering (SANS) and ultra-small-angle neutron scattering (USANS), were employed to investigate the behavior of mechanically plated Zn and ZnSn coatings both before and after heat treatment. The data collected presented inherent complexities due to the prevalent multiple scattering stemming from the substantial steel substrate (0.95 mm) in contrast to the relatively thin coating (~ 0.1 mm). Overcoming these challenges, we strategically focused on the Porod region at a high Q-range, obtained explicitly at a 1.3-m detector distance in SANS measurements. This region was deemed suitable for analysis, considering its low scattering intensity, which minimized the likelihood of subsequent scattering events.

To discern information pertaining to particle surface roughness, we calculated the background constant, B, and subtracted it from the scattering curve to derive the Porod’s slope, n. The subtle differences observed among various samples, post-background subtraction at high Q-range, hinted at potential contributions from interfaces beyond the steel substrate.

While the specific identification of Fe-Zn intermetallic phases or porosity formed post-heat treatment remained elusive due to the complexities of multiple scattering, this work lays a foundational framework for future investigations into coating behavior using SANS and USANS techniques. A critical insight gleaned from this study is the significant impact of steel matrix thickness on scattering behavior. An avenue for future research is to explore the influence of varying steel substrate thicknesses on multiple scattering across different q-ranges. Such investigations might necessitate the adoption of alternative coating techniques, given the limitations of mechanical plating on thinner substrates. Additionally, exploring alternative methodologies such as grazing incidence small-angle neutron scattering (GISANS) could prove beneficial in circumventing the challenges posed by multiple scattering, offering a novel perspective on the behavior of mechanically plated coatings.