The corrosion inhibitive properties of various kinds of potassium zinc phosphate pigments: Solution phase and coating phase studies

Progress in Organic Coatings 85 (2015) 109–122

Contents lists available at ScienceDirect

Progress in Organic Coatings journal homepage: www.elsevier.com/locate/porgcoat

The corrosion inhibitive properties of various kinds of potassium zinc phosphate pigments: Solution phase and coating phase studies F. Askari a,b , E. Ghasemi a , B. Ramezanzadeh b,∗ , M. Mahdavian b a b

Inorganic Pigment and Glazes Department, Institute for Color Science and Technology, Tehran, Iran Department of Surface Coatings and Corrosion, Institute for Color Science and Technology, Tehran, Iran

a r t i c l e

i n f o

Article history: Received 27 November 2014 Received in revised form 11 February 2015 Accepted 29 March 2015 Keywords: Carbon steel Coatings Electrochemical impedance spectroscopy (EIS) Epoxy Polarization

a b s t r a c t Potassium zinc phosphate (PZP) pigments were synthesized using different mole ratios of KOH/ZnCl2 , i.e. 1.5, 2, 2.5, 3, 3.5 and 4. The inhibition effects of the pigments were studied in the extract solution by polarization test and electrochemical impedance spectroscopy (EIS). The surface morphology was studied by scanning electron microscope (SEM). Surface analysis was performed by X-ray photoelectron spectroscopy. Pigments were also incorporated into the epoxy coating and salt spray and pull-off tests were implemented to investigate its corrosion protection properties. It was found that the mole ratio of KOH/ZnCl2 could significantly affect the corrosion inhibition properties of the PZP. Increasing the KOH/ZnCl2 mole ratio up to 2.5 caused significant improvement of the corrosion inhibition properties of the PZP both in the solution and coating phases. It was shown that PZP 2.5 could release Zn and P ions more than PZP 3.5 indicating its higher solubility in the 3.5 wt.% NaCl solution. PZP could significantly retard both anodic and cathodic reactions rates through releasing high amounts of Zn and P ions. PZP 2.5 enhanced the corrosion protection properties of the epoxy coating more than PZP 3.5 and decreased the adhesion loss significantly. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Organic coatings have been widely used to protect metals against corrosion. Anticorrosive pigments are added to the organic coatings in order to obtain long term corrosion protection properties [1–4]. The anticorrosive pigments are able to enhance the corrosion protection properties of the coatings through three main protection mechanisms including barrier, inhibitive and sacrificial [3–8]. Among these mechanisms, anticorrosive pigments with inhibitive action have been used to obtain coatings with longer corrosion protection service life. Zinc chromate is a well-known type of active corrosion inhibitive pigment which had been used in the organic coatings formulations. The mechanism of corrosion protection of this pigment is based on releasing solubilized inhibitive species which could form a protective layer on the metal surface. This pigment had been used for many years as an effective inhibitive pigment in the organic coatings. However, it has toxic and carcinogenic nature which contaminates the environment and represents a risk to human health [9–13]. Zinc phosphate and related substances are suggested as possible replacements for the

∗ Corresponding author. Tel.: +98 2122969771; fax: +98 2122947537. E-mail addresses: [email protected], [email protected] (B. Ramezanzadeh). http://dx.doi.org/10.1016/j.porgcoat.2015.03.018 0300-9440/© 2015 Elsevier B.V. All rights reserved.

chromates. However, the results reported in the literature showed less inhibitive performance of the zinc phosphate compared to chromate type pigments. This is attributed to its low solubility in water [12–15]. Therefore, attempts have been carried out to improve the corrosion inhibitive performances of the zinc phosphates through several physical and/or chemical modifications. In this way, the pigment solubility in water can be improved resulting in the increase of its inhibition properties. The second and third generations of the zinc phosphates have been developed in this way [16–20]. Naderi et al. investigated the inhibition effects of the zinc aluminum polyphosphate (ZAPP) and zinc aluminum phosphate (ZPA) pigments in the 3.5 wt.% NaCl solution on the mild steel specimens [19,20]. They found that polyphosphates showed better inhibitive action than orthophosphate. Compared to the conventional zinc phosphate (ZP) pigment, ZAPP and ZPA showed water solubility much greater than ZP. This resulted in more inhibitive properties of these two pigments than ZP. The corrosion inhibition mechanism of these two pigments has been referred to a protective layer precipitation on the metal surface. Polyphosphates include higher phosphate content than orthophosphate. As a result, it can produce stronger chelates with multivalent metal cations. These show that modification or replacement of the cationic and/or anionic parts of the pigment can lead to the pigment inhibition performance improvement. Iron, aluminum, potassium, sodium and lithium have been used for this purpose

110

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

[21–29]. To the best of our knowledge there is no systematic study on investigation the effects of KOH/ZnCl2 mole ratio on the phase composition and its correlation with corrosion inhibition properties of the PZP both in solution and coating phases. In the present work, a series of PZPs, having different chemical compositions and phase structures, were synthesized by changing KOH/ZnCl2 mole ratio. The inhibition properties of the pigments were studied preparing pigments extracts in the 3.5 wt.% NaCl solution and introducing them in the epoxy coating matrix. Electrochemical techniques including potentiodynamic polarization and EIS were used to investigate the inhibition properties of the pigments on the mild steel surface. SEM/EDS and XPS analyses were utilized to investigate the morphology and chemical composition of the films precipitated on the steel surface. Salt spray and pull-off tests were conducted in order to investigate the corrosion resistance and adhesion properties of the epoxy coatings containing PZPs on the steel substrate. 2. Experimental 2.1. Materials and sample preparation A series of potassium zinc phosphate pigments (PZP) were synthesized using different mole ratios of KOH/ZnCl2 i.e. 1.5, 2, 2.5, 3, 3.5 and 4. Zinc chloride (Merck, 0.06 mol) was added to 85 wt.% phosphoric acid solution (Merck, 0.06 mol) and distilled water (100 ml). This mixture was stirred until complete dissolution was achieved. In the next step, a solution of KOH (Merck) in distilled water (80 ml) was added to the above solution. The mixture was stirred for 3 h and heated up to 100 ◦ C for 12 h. The resulted residue was recovered by filtration, washed by distilled water and dried at 100 ◦ C. Samples obtained were denoted according to Table 1. The inhibition effects of the pigments prepared were studied on the surface of mild steel specimens (St-37). The mild steel panels have the following chemical composition: Al: 0.04, S:0.05, P:0.05, Mn:0.32, Si:0.34, C:0.19 and Fe:99.01 (wt.%). Samples were abraded by emery papers of 600, 800, 1200 and 2400 grades followed by acetone degreasing. Finally, the samples were rinsed with distilled water and dried in an oven at 40 ◦ C. 2.1.1. PZP extracts preparation To prepare the extracts, 1 g of the PZP was stirred in 1 L of 3.5 wt.% NaCl solution for 24 h. The solutions were then filtered for further analysis. The concentration of dissolved species in the pigment extracts was measured by an inductively coupled plasmaoptical emission spectrometer [Varian Vista Pro ICP-OES]. The pH values of the solutions containing PZPs extracts were assessed by a Metrohm model 827 pH lab. 2.1.2. Epoxy coating preparation Epoxy coatings containing 15 wt.% of PZP and ZP were prepared. For this purpose, the epoxy resin of Araldite GZ7 7071X75 (based on bisphenol-A in a xylene solution) was prepared from Saba Shimi Co. The epoxy value, density and solid content of the resin were 0.1492–0.1666 eq/100 g, 1.08 g cm−3 and 74–76%, respectively. Pigments were then added to the epoxy resin and dispersed by a perl-mill for 8 h in order to obtain an average particle size up to 10 ␮m. EFKA-2025, a silicone based defoamer in cyclohexanone Table 1 Sample coding for the pigments synthesized by changing KOH/ZnCl2 mole ratio. Name of pigments

PZP 1.5

PZP 2

PZP 2.5

PZP 3

PZP 3.5

PZP 4

KOH/ZnCl2 mole ratio

1.5

2

2.5

3

3.5

4

solvent, was used at maximum consumption of 0.1 wt.%. BYK-306, a polyether modified polydimethylsiloxane in xylene and 2phenoxyethanol solvents, was locally provided and used as surface modifier at maximum consumption of 0.5 wt.%. In the next step, the mill base was mixed with a stoichiometric amount of a polyamide curing agent (epoxy/hardener (w/w) = 2.3/1). The hardener used in this study was based on an amido polyamide, CRAYAMID 115, from Arkema Co. The solid content, density and viscosity at 40 ◦ C of the hardener are 50%, 0.97 g/cm3 and 50,000 cps, respectively. At the end, the coatings prepared were applied on the cleaned and abraded steel samples using a film applicator. All of the coatings were cured at 120 ◦ C for 30 min. The dry film thickness of 52 ± 3 ␮m was measured for the samples by Defelsko Posi Tector 6000. 2.2. Techniques 2.2.1. Electrochemical measurements The corrosion performance of the steel specimens was studied in 100 cc of test solutions. Polarization curves were obtained employing AUTOLAB G1. The test was carried out in a conventional three electrode cell including steel specimen (1 cm2 area) as working electrode, platinum as counter electrode and saturated Ag/AgCl as reference electrode. The polarization curves were obtained at sweep rate of 1 mV/s in the range of ±100 mV from open circuit potential (OCP). The electrochemical impedance spectroscopy (EIS) measurements were conducted in a three electrode cells (like the one used in the polarization test) at the frequency range and peak to zero amplitude of 10 kHz–100 mHz and ±10 mV, respectively. The impedance data was analyzed by NOVA 1.8 software. The EIS and polarization measurements were carried out 3 times to ensure the repeatability of the measurements. All experiments were conducted at different immersion times of 2, 4, 24 and 48 h on 1 cm2 of the samples. 2.2.2. Surface analysis The surface morphology and composition of the steel specimens exposed to the solutions with and without pigment extract were studied by a scanning electron microscope (SEM) model Philips XL30 (equipped with energy dispersive spectroscopy: EDS). The SEM analysis was done after 70 h immersion at 25 ◦ C. Also, the composition of the film deposited on the metal surface was studied by a Specs EA 10 Plus energy dispersive X-ray photoelectron spectroscopy (XPS) equipped with a concentric hemispherical analyzer (CHA). In this experiment, the radiation source (at pressure of 10−9 mbar) was Al K␣ . The shift of binding energies (BE) was calibrated with respect to reference peak of carbon at binding energy of 285 eV. 2.2.3. Corrosion protection studies of epoxy coatings The epoxy coated samples were exposed to salt spray for evaluation of the corrosion protection performances of the samples. For this purpose, the x-scribed coatings were exposed to salt spray test cabin according to ASTM B117 (NaCl 5 wt.% solution) for 100 h. The adhesion strength values of the coatings applied on the steel substrates were also measured after 100 h exposure to salt spray by a Posi Test pull-off adhesion tester (DeFelsko). For this purpose, the aluminum dollies were glued on the surface of the epoxy coating using a two-part Araldite 2015 (Huntsman advanced materials, Germany) adhesive. Samples were then kept at ambient temperature for 24 h to insure that the glue fully cured. Finally, a slot was made around dollies and they were pulled at a speed of 10 mm/min normal to the coating surface until the epoxy coating was detached from the steel substrate. Measurements were performed on three replicates to ensure the repeatability of data.

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122 Table 2 pH values of the 3.5 wt.% NaCl solutions without and PZPs extracts. Name of pigments

PZP 1.5

PZP 2

PZP 2.5

PZP 3

PZP 3.5

PZP 4

Blank

pH

5.1

5.1

5.1

6.0

6.0

6.1

6.1

3. Results and discussion

111

Table 3 The results of ICP-OES analysis of 3.5 wt.% NaCl solutions containing PZP 2.5 and PZP 3.5 extracts. Sample

PZP 2.5 extract PZP 3.5 extract

Ion released from the pigments [mg l−1 ] Zn

K

P

3.1409 2.9905

63.044 9.0187

25.778 2.1388

3.1. Evaluation of PZP properties in the extract solution 3.1.1. pH-metry and ICP-OES analyses of the pigments extracts The pH values of the 3.5 wt.% NaCl solutions without and with PZPs extracts were measured and the results are listed in Table 2. It can be seen from Table 2 that the pH of the blank solution is approximately neutral with a pH around 6. Results show the acidification of 3.5 wt.% NaCl solution by incorporation of PZPs 1.5, 2 and 2.5 (pH 5.1). However, PZPs 3, 3.5 and 4 extracts did not change the pH compared to the blank solution. Also, the ICP-OES analysis was performed on the solutions containing PZPs 2.5 and 3.5 as typical extracts having different pHs. This experiment was done in order to

evaluate the concentration of the components released from these two pigments in the 3.5 wt.% NaCl solution (Table 3). From Table 3 it is clear that PZPs 2.5 and 3.5 released zinc, phosphorus and potassium containing species into the 3.5 wt.% NaCl. According to the ICP and XRD results [30] it can be seen that the PZPs released K+ , Zn2+ and PO4 3− ions. In addition to these ions it seems that the PZPs containing KZn2 PO4 (HPO4 ) phase may also release HPO4 2− ion that can change the extract pH to acidic due to the release of H+ cations by dissociation of HPO4 2− . A slight difference between the released Zn2+ ion contents of the PZPs 2.5 and 3.5 can be observed in Table 3. However, PZP 2.5

Fig. 1. Polarization curves of the steel samples dipped in the solutions with PZP 2.5, PZP 3.5 extracts and Blank solution after (a) 2 h, (b) 4 h and (c) 24 h immersion times at 25 ◦ C.

112

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

released phosphate and potassium ions much more than PZP 3.5. All of these results show that the PZPs 1.5, 2 and 2.5 may exhibit different electrochemical behaviors compared to PZPs 3, 3.5 and 4. 3.1.2. Potentiodynamic polarization measurements The inhibition properties of the synthesized PZPs were studied by polarization measurements. For this purpose, the steel specimens were dipped in the solutions containing PZP extracts and the test was carried out at different immersion times. Polarization curves are presented in Fig. 1 for the steel specimens tested in the 3.5 wt.% NaCl solution containing PZPs 2.5 and 3.5 extracts as typical extract solutions having different pHs. The values of corrosion current density (icorr ) and polarization resistance (Rp ) were extracted from the polarization curves using Tafel extrapolation method. The icorr and Rp values of different samples are compared in Fig. 2. From Fig. 2 it can be clearly seen that introducing PZP 1.5, 2 and 2.5 extracts into the 3.5 wt.% NaCl solutions caused significant decrease of the icorr and increase of Rp compared to the blank sample. This indicates that the extracts include corrosion inhibitive species prevented the steel against corrosion through formation of protective film on the active sites of the steel. The higher icorr and the lower Rp were obtained in the extract solutions containing PZPs 3, 3.5 and 4 compared to PZPs 1.5, 2 and 2.5 extracts. It can be also seen that PZP 4 extract did not show good corrosion inhibition behavior compared to the blank sample. All of these observations show that the ratio of KOH/ZnCl2 (as a parameter considered during PZP synthesis) is an effective parameter influencing the corrosion inhibition

properties of the PZPs. Considering comparative results given in Fig. 2, it is evident that the increase of immersion time resulted in the decrease of icorr and the increase of Rp values for the samples immersed in the solutions containing PZPs 1.5, 2 and 2.5 extracts. It can be demonstrated that the pigments extracts formed protective films on the metal surface as the immersion time increased. This means that increasing immersion time resulted in the increase of film thickness and surface coverage for the samples immersed in the solutions containing PZPs 1.5, 2 and 2.5 extracts. These observations show that PZPs 1.5, 2 and 2.5 showed proper inhibitive properties but PZPs 3, 3.5 and 4 showed poor inhibitive action. From Fig. 1a it can be clearly seen that the shape of anodic branch affected in the solution containing PZP 2.5 greater than cathodic one at the beginning of immersion (2 h). The anodic current density decreased in the presence of PZP 2.5 but the cathodic current density changes were not significant compared to the blank sample. Also, it can be seen that increasing immersion time from 2 to 4 h (Fig. 1b) caused the decrease of anodic current density without significant changes of cathodic current density compared to the blank solution. Moreover, the Ecorr value shifted toward more positive values. These observations indicate that the anodic inhibition properties of the PZP 2.5 were more dominant than its cathodic inhibition at the beginning of immersion. In fact, PZP 2.5 reduced the anodic dissolution rate of the steel specimen without significant effects on the cathodic reaction mechanism at the preliminary stage. However, different results were obtained at longer immersion times (Fig. 1c). It can be seen that the Ecorr shifted toward negative direction and both the cathodic and anodic current

Fig. 2. (a) Corrosion current density (icorr ) and (b) polarization resistance (Rp ) values of the steel samples immersed in PZPs extracts and blank solution after 2, 4 and 24 h, at 25 ◦ C.

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

113

Fig. 3. The SEM micrographs of the precipitated layers on the surface of steel samples after 70 h of immersion in 3.5 wt.% NaCl solution containing (a) PZP1.5, (b) PZP 2, (c) PZP 2.5, (d) PZP 3, (e) PZP 3.5, (f) PZP 4 extract and (g) no pigment at 25 ◦ C.

densities decreased significantly after 24 h immersion compared to the blank sample. These are indicative of cathodic and anodic reactions suppression through adsorption of the released inhibitive species on the cathodic and anodic sites of the metal surface at prolonged immersion times. Different results were obtained when the steel panels were dipped in the solution containing PZP 3.5 extract.

From the obtained results it can be seen that increasing immersion time resulted in the shift of corrosion potential toward negative values and also the increase of anodic and cathodic current densities. Comparing the polarization curves of the PZP 3.5 and blank sample may imply that this pigment did not show inhibitive properties as the anodic and cathodic current densities remained almost

114

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

Fig. 4. (a) Nyquist (Z versus −Z ) and (b and c) Bode plots (log |Z| versus log(f) and −phase angle versus log(f)) for the steel samples immersed in 3.5 wt.% NaCl solutions without and with PZP 2.5, PZP 3.5 and ZP extracts for 4 h at 25 ◦ C.

unchanged. This can be related to the very low solubility of this pigment as shown in ICP-OES results.

3.1.3. SEM analysis of the samples exposed to PZPs extracts The surface morphology of the steel samples exposed to the PZPs extracts for 70 h was studied by SEM. The SEM micrographs of different samples are given in Fig. 3. According to Fig. 3, a layer with the morphology presented in Fig. 3 precipitated on the surface of the samples exposed to the solutions containing PZPs 1.5, 2, 2.5 and 3 extracts. It is clear from this figure that a more compact film with less corrosion products was formed on the surface of the sample immersed in the solution containing PZP 2.5 extract compared to other samples. However, corrosion products were formed on the surface of the samples immersed in the blank solution and also the solutions containing PZPs 3.5 and 4 extracts. These observations reveal that the corrosion inhibition properties of the PZPs strongly depend on the chemical composition of the pigments. In fact, PZPs 1.5, 2 and 2.5 showed corrosion inhibition efficiencies much greater than PZPs 3, 3.5 and 4. These observations are in accordance with the results shown in Fig. 2.

3.1.4. Electrochemical impedance measurements (EIS) The inhibition properties of the PZPs 2.5 and 3.5 were compared with conventional ZP by EIS. Figs. 4–6 show Nyquist and Bode plots of the specimens immersed in 3.5 wt.% NaCl solution containing ZP, PZPs 2.5 and 3.5 extracts after 4, 24 and 48 h immersion. Electrochemical equivalent circuits with one and two relaxation times were utilized in order to model the corrosion process. The typical fitting process is shown in Fig. 7. It is clear from Figs. 4 to 6 that all plots of the samples dipped in the solutions without and with ZP and PZP 3.5 extracts include one relaxation time after 4 h immersion. However, two relaxation times were observed for the solution including PZP 2.5 after 4 h immersion. This can be attributed to the film growth on the surface of the steel specimen immersed in the solution containing PZP 2.5 extract. In Fig. 7, Rs , Rc , Rct , CPEc and CPEdl are solution resistance, film resistance, charge transfer resistance, constant phase element of the film and constant phase element of double layer. The effective double layer capacitance (Cdl ) and film capacitance (Cc ) values were calculated according to Eq. (1) [31]. C=

(Y0 R)1/n R

(1)

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

115

Fig. 5. (a) Nyquist (Z versus −Z ) and (b, c) Bode plots (log |Z| versus log(f) and −phase angle versus log(f)) for the steel samples immersed in 3.5 wt.% NaCl solutions without and with PZP 2.5, PZP 3.5 and ZP extracts for 24 h at 25 ◦ C.

where C, Y, R and n represent capacitance of double layer or film, admittance of CPE elements of double layer or film, charge transfer resistance or film resistance, and the empirical exponent of CPE elements of double layer or film, respectively. The electrochemical parameters extracted from impedance data are displayed in Table 4. Table 4 shows higher Rct and lower Cdl values of the sample immersed in the solutions containing PZPs than blank solution and the one with ZP extract. This indicates that PZPs are inhibitive pigments with higher corrosion inhibition properties than ZP. Results show that the increase in immersion time caused a decrease in Rct and an increase of Cdl values of the samples immersed in the blank solution and ZP and PZP 3.5 extract solutions. This means that the corrosion inhibition efficiency of the ZP and PZP 3.5 pigments decreased as the immersion time elapsed. Different results were obtained for the samples immersed in the solution containing PZP 2.5 extract. Results show higher Rct and lower Cdl values in

the extracts of the PZP 2.5 even after 48 h immersion compared to ZP and PZP 3.5 samples. This means that PZP 2.5 is a stronger corrosion inhibitive pigment than others. In fact, the observation of two relaxation times for this sample indicates that this pigment is an effective inhibitive pigment which developed a protective film on the steel surface. The film growth continued at longer immersion times resulting in a better surface coverage together with higher thickness. As a result, it reduced corrosion rate of the steel more than other pigments especially at prolonged immersion periods. The impedance magnitude at 100 mHz and also phase angle values at 10 kHz were obtained from the Bode plots and the results are given in Fig. 8. From Fig. 8 it is clear that PZP 2.5 showed higher impedance values at all immersion times than ZP and PZP 3.5. Phase angle is another useful parameter obtained from the Bode diagrams.

116

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

Fig. 6. (a) Nyquist (Z versus −Z ) and (b, c) Bode plots (log |Z| versus log(f) and −phase angle versus log(f)) of the steel samples immersed in 3.5 wt.% NaCl solutions without and with PZP 2.5, PZP 3.5 and ZP extracts for 48 h at 25 ◦ C.

The more negative phase angle of PZP 2.5 than ZP and PZP 3.5 samples can show the capacitive behavior improvement of the electrode/electrolyte system. More positive phase angles were observed for the samples immersed in the blank solution after 24 and 48 h immersion. However, increasing the immersion times up to 48 h caused a negative shift of phase angle of the samples immersed in the solution containing PZP 2.5 extract. This is another prove showing the film growth on the metal surface in the presence of PZP 2.5 [32]. These observations reveal that PZP 2.5 has better inhibitive properties than other pigments. It may be attributed to the better film precipitation capability of this pigment compared to the PZP 3.5 and ZP.

3.1.5. Visual observation of the samples immersed in the test solutions Steel panels were dipped in the solutions without and with ZP and PZP 2.5 extracts. The visual performances of the samples were studied at different immersion times (Fig. 9). From Fig. 9 it can be obviously seen that brown rust appeared on the steel sample immersed in the solution without pigment extract after 1 h immersion. However, corrosion of the steel was not seen until 24 and 120 h immersion in the solutions containing ZP and PZP 2.5 extracts, respectively. These observations can clearly show that the PZP 2.5 extract showed inhibitive behavior much greater than ZP one. It seems that PZP 2.5 can form a protective layer on the steel surface.

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

117

Table 4 The electrochemical parameters extracted from impedance data of different samples. Sample

Rs ( cm2 )

Rct a (k cm2 )

Y0 PZP 2.5 (4 h) PZP 2.5 (24 h) PZP 2.5 (48 h) PZP 3.5 (4 h) PZP 3.5 (24 h) PZP 3.5 (48 h) ZP (4 h) ZP (24 h) ZP (48 h) No pigment (4 h) No pigment (24 h) No pigment (48 h) a b c d

5.01 6.42 9.54 5.38 4.7 6.75 7.06 5.03 6.47 4.89 5.22 4.42

5.11 6.31 6.61 1.80 1.62 1.05 1.7 1.4 0.91 1.3 1.5 1.1

Cdl (nF cm−2 )

CPEdl b

−1

(

−2

cm

n

s )

0.00015 0.00016 0.00015 0.00025 0.00032 0.00045 0.0003 0.00036 0.00047 0.00034 0.00044 0.00035

Rc d (k cm2 )

c

n

0.75 0.74 0.74 0.82 0.74 0.71 0.82 0.74 0.70 0.77 0.77 0.79

Y0 0.137 0.160 0.149 0.210 0.264 0.340 0.258 0.285 0.334 0.268 0.391 0.277

10.81 24.06 36.98 – – – – – – – – –

Cc (nF cm−2 )

CPEc b

−1

(

0.0002 0.00023 0.000016 – – – – – – – – –

−2

cm

n

s )

c

n

0.72 0.75 0.72 – – – – – – – – –

0.20 0.041 0.023 – – – – – – – – –

The standard deviation range for Rct values was between 2% and 8.0%. The standard deviation range for Y0 values was between 0.5% and 8.0%. The standard deviation range for n values was between 0.2% and 1.5%. The standard deviation range for Rc values was between 1.2% and 8.4%.

Fig. 7. Representative diagrams of using the equivalent circuits to fit the experimental data for the samples dipped in the solutions (a) containing PZP 2.5 and (b) without extract after 24 h immersion, at 25 ◦ C.

118

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122 Table 5 The results of EDSanalysis of the precipitated layers on the surface of steel samples after 70 h of immersion in 3.5 wt.% NaCl solution containing PZP 2.5 and PZP 3.5. Sample

PZP 2.5 PZP 3.5

Weight percent (wt.%) Fe

Zn

K

P

84.9 92.35

1.47 2.62

0.47 0.31

13.16 4.72

PZP 2.5 than PZP 3.5 cannot be attributed to the Zn ions deposition on the cathodic sites of the surface. The amount of K detected on the surface of the samples exposed to the solutions containing PZPs extracts are the same and lower than Zn. Regarding to the results given in Table 3, the PZP 2.5 released K much more than PZP 3.5. However, the EDS analysis revealed that K did not precipitate on the metal surface considerably; therefore it could not participate in the corrosion inhibition reactions. Results show that the film precipitated on the metal surface exposed to the PZP 2.5 extract has more P content compared to PZP 3.5 extract. The phosphate ions act as anodic inhibitor and can retard the anodic reaction rate through forming protective layer over the anodic sites restricting the access of the corrosive species to the metal surface. Therefore, the higher inhibition performance of the PZP 2.5 than PZP 3.5 can be attributed to its greater capability of forming protective layer on the anodic sites. This observation is in accordance with the obtained results from polarization test. It was shown in the polarization test that PZPs affected anodic reaction more than cathodic one.

Fig. 8. (a) log |Z| at 10 mHz and (b) −phase angle at 10 kHz (degree) obtained from EIS results of steel samples immersed in 3.5 wt.% NaCl solution and the solutions containing PZP 2.5, PZP 3.5 and ZP extracts and 3.5 wt.% NaCl solution without pigment for4 h, 24 h and 48 h, at 25 ◦ C.

3.2. Surface analysis 3.2.1. SEM/EDS analysis SEM/EDS analysis was done to investigate the morphology and elemental composition of the films precipitated on the steel surface. The test was done on the steel samples immersed in 3.5 wt.% NaCl solutions containing PZPs 2.5 and 3.5 extracts after 70 h immersion. Obtained results from the SEM/EDS analysis are shown in Table 5. SEM/EDS analysis revealed the existence of elements including P, Zn, K and Fe on the steel surface. This indicates that the ions released by the pigments precipitated on the steel surface. Table 5 shows a greater Zn on the surface of the sample exposed to the solution containing PZP 3.5 extract than the one exposed to PZP 2.5. This finding reveals that the greater inhibition properties of the

3.2.2. XPS analysis The XPS spectrum of the sample immersed in the extract solution of PZP 2.5 is shown in Fig. 10. Based on the XPS spectrum, the peaks corresponding to different bonding energies are obtained and the results are given in Table 6 [33–41]. In order to better understand the composition of the film formed on the steel surface, the deconvolution of O 1s spectrum was performed on high resolution XPS spectrum. The O 1s spectrum (Fig. 11) was deconvoluted into three peaks. The results of deconvoluted spectrum are listed in Table 7 [42,43]. From Tables 6 and 7, it can be clearly seen that the protective layer formed on the steel surface was composed of oxide/hydroxides and phosphate components. 3.2.3. Corrosion inhibition mechanism of the PZP It has been shown that KZn2 PO4 (HPO4 ) is the main phase of PZP 2.5. Moreover, the obtained results from ICP-OES analysis revealed that PZP 2.5 released Zn2+ and K+ cations and phosphate components. The inhibition mechanism of PZP can be explained through considering the pH of PZP 2.5 extract and the corrosion potential of the steel sample immersed in the solution containing PZP extract.

Fig. 9. Visual performances of the samples immersed in the 3.5 wt.% NaCl solutions without and with ZP and PZP 2.5 extracts at different immersion times and 25 ◦ C.

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122 Table 7 Results obtained from deconvoluted XPS spectrum of O1s.

6000 O (1s)

5000

Element

Zn (2p) Zn (2p)

Count

Binding energy (eV)

Component

530.24

Hydroxide compositions, example: Zn(OH)2 , Fe(OH)2 , Fe(OH)3 and FeOOH Oxide form of metals, example: metal oxide, ZnO, FeOOH and Fe2 O3 Phosphate component, example: PO4 3− and HPO4 2− group

O 1s

4000 Fe (2p)

531

Zn

3000 C (1s)

531.96

2000 P (2s)

Reference

2H2 O + O2 + 4e  4OH−

0 1000

1200

800

600 400 Binding Energy (eV)

200

0

Fig. 10. XPS survey spectrum associated with the surface of steel sample immersed in PZP 2.5 extract for 70 h, at 25 ◦ C. Table 6 XPS survey spectrum results. Element

Binding energy (eV)

Component

Reference

Zn 2p

1045

[35,37]

Zn 2p

1023

The peaks assigned to a non-conductive form of Zn representing formation of solid particles of zinc phosphate or PZP The peaks assigned to a conductive form of Zn representing formation of zinc oxide and zinc hydroxide on the surface The states of the phosphorus (PO4 3− ) bonded in metal phosphates The conductive form of Fe representing formation of iron oxides

[36,40,42]

P 2p

133.8

P 2s Fe 2p

191 710

[35,37,38]

[35,39–41]

Relative intensity (a.u.)

O 1s

534.5

533.5

532.5

531.5 Binding energy (eV)

530.5

529.5

528.5

Fig. 11. High resolution XPS spectrum of O 1s peak and its deconvolution for the steel surface that immersed in PZP 2.5 extract solution after 70 h, at 25 ◦ C.

The pH of the solution containing PZP extract is acidic (5.1) before the steel immersion. Therefore, the corrosive electrolyte attacks to the steel surface causing anodic dissolution of iron (Eq. (2)) and cathodic reduction of oxygen (Eqs. (3) and (4)). Fe  Fe

[31,42,43]

P (2p)

1000

+

119

2+

+ 2e

4H + O2 + 4e  2H2 O

(2) (3)

(4)

According to Eq. (3) the consumption of H+ on the metal surface causes the increase of pH. As a result, the oxygen reduction happens in the next step on the metal surface. The pH of the solution containing PZP 2.5 extract was also measured after steel panel immersion for 70 h. It was found that the pH increased from 5.1 (before the immersion) to 5.4 (after the immersion), indicating H+ cations consumption. From the polarization test results it was found that the corrosion potential shifted to more positive values at short immersion times. This can be attributed to the acidic nature of the solution containing PZP 2.5 extract causing more PO4 3− or HPO4 2− anions complexes precipitation on the anodic sites [33]. Based on XPS results, it seems that the layer precipitated on the steel surface was composed of oxides/hydroxides and phosphate components. This means that PO4 3− anions may react with Fe2+ and Zn2+ cations producing Fe3 (PO4 )2 and Zn3 (PO4 )2 compounds [44,45]. However, the consumption of H+ cations on the metal surface caused the increase of local pH. Consequently, the Fe2+ and Zn2+ cations reach the cathodic sites and react with OH− ions to produce hydroxide compounds such as Fe(OH)2 and Zn(OH)2 [43]. The zinc oxide/hydroxide precipitation on the cathodic regions reduces the cathodic reaction rate more than anodic one, resulting in the shift of corrosion potential toward less positive values at longer immersion times as depicted in Section 3.1.2. The precipitation pH values of the metal oxide/hydroxide components, detected on the metal surface through XPS analysis, have been measured. For this purpose, the precipitation pHs of the Zn(OH)2 and Fe(OH)2 were measured by adding different values of NaOH as neutralizing agent to the solutions containing ZnCl2 and FeCl2 . Also, the precipitation pH values of the Zn3 (PO4 )2 and Fe3 (PO4 )2 components were obtained by adding stoichiometric amounts of phosphoric acid to the mixture of ZnCl2 and FeCl2 and using NaOH as neutralizing agent (Table 8). From the obtained results (Table 8), it can be seen that the precipitation pH of Zn3 (PO4 )2 (4.28) and Fe3 (PO4 )2 (4.95) are lower than the pH of the extract solution (pH 5.4). However, the precipitation pHs of Zn(OH)2 (6.72) and Fe(OH)2 (5.54) are higher than the pH of the solution containing PZP extract. In fact, precipitation of such components on the steel surface, confirmed by the XPS analysis, reveals that the pH on the surface of steel would be greater than the pH of the bulk of solution causing the Zn2+ and Fe2+ cations precipitation as Zn(OH)2 and Fe(OH)2 components. Table 8 Precipitation pH values obtained through neutralizing different solutions by NaOH in order to confirm the results of XPS analysis. Solution composition

Precipitated component

Precipitation pH

10 mM ZnCl2 + phosphoric acid 10 mM FeCl2 + phosphoric acid 10 mM ZnCl2 10 mM FeCl2

Zn3 (PO4 )2 Fe3 (PO4 )2 Zn(OH)2 Fe(OH)2

4.28 4.95 6.72 5.54

120

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

Fig. 12. Digital photographs of the x-scribed coated samples after 100 h exposure to salt spray condition and disbonding from x-scribe for samples containing (a1 and a2 ) PZP, (b1 and b2 ) ZP and (c1 and c2 ) sample without pigment.

3.3. Studying the effects of PZP on the protective performance of epoxy coating

100 h salt spray exposure. Then, the adhesion loss values were calculated by Eq. (5) [18]:

3.3.1. Salt spray exposure PZP was introduced into the epoxy/polyamide coating in order to enhance its corrosion resistance. Moreover, the blank sample and coating loaded with ZP were prepared as references. For this purpose, the epoxy coating loaded with 15 wt.% PZP was prepared and applied on the steel surface. Samples were x-scribed and then located in the salt spray chamber for 100 h. The visual appearances of the samples are shown in Fig. 12. According to Fig. 12, rust, blister and accumulation of corrosion products appeared near the scribed area and beneath the blank and ZP pigmented coating. It can be seen that addition of PZP to the coating significantly decreased the number of corrosion spots and blisters. Moreover, the corrosion products development near the scribed area and beneath the coating reduced significantly in the presence of PZP. Fig. 12 also shows that the coating disbondment area decreased significantly after addition of PZP. These show that PZP efficiently enhanced the anticorrosion properties of the epoxy coating. In fact, the PZP released inhibitive species inside the coating matrix and at the defected area restricting aggressive corrosion species access to the active sites on the metal surface. This can result in lower OH− ions creation beneath the coating. As a result, the coating delamination and corrosion products growth beneath the coating was limited in the presence of PZP. This observation can prove the great inhibitive action of this pigment.

Pull-off strength loss (%) =

3.3.2. Pull off adhesion test The adhesion strength values of the coatings without and with ZP or PZP pigments were measured by pull off test before and after

a−b × 100 a

(5)

where a and b are dry and recovery pull-off strength, respectively. The pull-off strength loss values are shown in Fig. 13. Electrochemical reactions occur at the coating/metal interface when the corrosive electrolyte reaches beneath the coating. As a result, hydroxyl ions can be produced during cathodic reaction, leading to the increase of pH at the coating/metal interface. This is responsible for the adhesion bonds breakdown and coating

Fig. 13. Pull-off adhesion strength loss of the coated sample containing PZP, ZP and without pigment before and after 100 h exposure to salt spray condition.

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

delamination. Results show that adhesion loss for the blank coating was considerably higher than the pigmented coatings. However, there was a slight difference between the adhesion loss values of the coatings loaded with ZP and PZP. In fact, PZP and ZP released Zn2+ cations which reacted with hydroxyl ions produced at the cathodic sites. As a result, a protective film composed of the zinc oxide/hydroxide covered the active sites of steel surface restricting the cathodic reactions.

4. Conclusion Results revealed that the mole ratio of KOH/ZnCl2 is an effective parameter influencing the PZP inhibition properties. It was shown that PZPs reduced corrosion current density and shifted corrosion potential toward positive values. PZP 2.5 resulted in the greatest inhibition properties among the pigments even at long immersion times. From the obtained results, it was found that PZP 2.5 affected anodic reaction more than cathodic one. Impedance measurements showed higher inhibition effects of the PZPs than ZP. An increase in inhibition properties of the PZP 2.5 was observed as the immersion time increased. The impedance data also revealed the progressive precipitation of a protective film on the surface of the sample during immersion in the PZP 2.5 extract solution. The obtained results from ICP-OES analysis of the solution containing extracts showed that PZP 2.5 released Zn, K and P ions more than PZP 3.5 indicating the higher water solubility of this pigment. SEM/EDS and XPS analyses revealed that a film composed of zinc hydroxide, iron phosphate and zinc phosphate precipitated on the steel surface. Obtained results from salt spray and pull off tests demonstrated that addition of PZP 2.5 to the epoxy coating improved its corrosion protection properties significantly.

Acknowledgment The authors are grateful to Ms. Ashari for experimental contributions, and to the supplier of the paint resins (Saba Shimi Arya Co.).

References [1] C.-H. Chang, T.-C. Huang, C.-W. Peng, T.-C. Yeh, H.I. Lu, W.I. Hung, C.-J. Weng, T.-I. Yang, J.-M. Yeh, Novel anticorrosion coatings prepared from polyaniline/graphene composites, Carbon 50 (14) (2012) 5044–5051. [2] C.-J. Weng, J.-Y. Huang, K.-Y. Huang, Y.-S. Jhuo, M.-H. Tsai, J.-M. Yeh, Advanced anticorrosive coatings prepared from electroactive polyimide–TiO2 hybrid nanocomposite materials, Electrochim. Acta 55 (28) (2010) 8430–8438. [3] S.K. Dhoke, A.S. Khanna, Electrochemical behavior of nano-iron oxide modified alkyd based waterborne coatings, Mater. Chem. Phys. 117 (2) (2009) 550–556. [4] N. Kouloumbi, G.M. Tsangaris, C. Vourvahi, F. Molnar, Corrosion resistance and dielectric properties of an iron oxide-filled epoxy coating, J. Coat. Technol. 69 (7) (1977) 53–59. [5] G. Williams, S. Geary, H.N. McMurray, Smart release corrosion inhibitor pigments based on organic ion-exchange resins, Corros. Sci. 57 (2012) 139–147. [6] S.K. Dhoke, A.S. Khanna, Effect of nano-Fe2 O3 particles on the corrosion behavior of alkyd based waterborne coatings, Corros. Sci. 51 (1) (2009) 6–20. [7] M. Deya, G. Bluslein, R. Romagnoli, B. del Amo, The influence of the anion type on the anticorrosive behaviour of inorganic phosphates, Surf. Coat. Technol. 150 (2002) 133–142. [8] A.C. Bastosa, M.L. Zheludkevich, I. Klüppel, G. Grundmeierc, M.G.S. Ferreira, Modification of zinc powder to improve the corrosion resistance of weldable primers, Prog. Org. Coat. 69 (2) (2010) 184–192. [9] J. Sinko, Challenges of chromate inhibitor pigments replacement in organic coatings, Prog. Org. Coat. 42 (3) (2001) 267–282. [10] G. Buxbaum, G. Pfaff, Cadmium Pigments, Industrial Inorganic Pigments, WileyVCH, 2005, pp. 121–123. [11] M. Hernandez, J. Genesca, Effect of an inhibitive pigment zinc-aluminumphosphate (ZAP) on the corrosion mechanisms of steel in waterborne coatings, Prog. Org. Coat. 56 (2) (2006) 199–206.

121

[12] F. Galliano, D. Landolt, Evaluation of corrosion protection properties of additives for waterborne epoxy coatings on steel, Prog. Org. Coat. 44 (3) (2002) 217–225. [13] P. Kalenda, Properties of anticorrosion pigments depending on their chemical composition and PVC value, Pigment Resin Technol. 35 (4) (2006) 188–199. [14] M. Mahdavian, M. Attar, Evaluation of zinc phosphate and zinc chromate effectiveness via AC and DC methods, Prog. Org. Coat. 53 (3) (2005) 191–194. [15] A.C. Bastos, M.G. Ferreira, A.M. Simoes, Corrosion inhibition by chromate and phosphate extracts for iron substrates studied by EIS and SVET, Corros. Sci. 48 (6) (2006) 1500–1512. [16] R. Naderi, M.M. Attar, Electrochemical study of protective behavior of organic coating pigmented with zinc aluminum polyphosphate as a modified zinc phosphate at different pigment volume concentrations, Prog. Org. Coat. 66 (3) (2009) 314–320. [17] S.M. Mousavifard, P.M.M. Nouri, M.M. Attar, B. Ramezanzadeh, The effects of zinc aluminum phosphate (ZPA) and zinc aluminum polyphosphate (ZAPP) mixtures on corrosion inhibition performance of epoxy/polyamide coating, J. Ind. Eng. Chem. 19 (3) (2013) 1031–1039. [18] M.R. Heydarpour, A. Zarrabi, M.M. Attar, B. Ramezanzadeh, Studying the corrosion protection properties of an epoxy coating containing different mixtures of strontium aluminum polyphosphate (SAPP) and zinc aluminum phosphate (ZPA) pigments, Prog. Org. Coat. 77 (1) (2014) 160–167. [19] R. Naderi, M.M. Attar, Electrochemical assessing corrosion inhibiting effects of zinc aluminum polyphosphate (ZAPP) as a modified zinc phosphate pigment, Electrochim. Acta 53 (18) (2008) 5692–5696. [20] R. Naderi, M.M. Attar, The inhibitive performance of polyphosphate-based anticorrosion pigments using electrochemical techniques, Dyes Pigments 80 (3) (2009) 349–354. [21] M. Bethencourt, F.J. Botana, M. Marcos, R.M. Osuna, J.M. Sanchez, Inhibitor properties of green pigments for paints, Prog. Org. Coat. 46 (4) (2003) 280–287. [22] E. Alibakhshi, E. Ghasemi, M. Mahdavian, A comparison study on corrosion behavior of zinc phosphate and potassium zinc phosphate anticorrosive pigments, Prog. Color Color. Coat. 5 (2012) 91–99. [23] E. Alibakhshi, E. Ghasemi, M. Mahdavian, Optimization of potassium zinc phosphate anticorrosion pigment by Taguchi experimental design, Prog. Org. Coat. 76 (1) (2013) 224–230. [24] S. Boontima, C. Danvirutai, T. Srithanratana, Thermal decompositions kinetics and reversible hydration study of the Li2 Zn(HPO4 )2 ·H2 O, Solid State Sci. 12 (7) (2010) 1226–1230. [25] G.S. Goplakrishna, S.P. Madhu, M. Mahendra, B.H. Doreswamy, M.J. Mahesh, M.A. Sridhar, J. Shashidhara Prasad, Hydrothermal synthesis, crystal structure and characterization of 2(LiZnHP2 O7 ), Mater. Lett. 60 (5) (2006) 613–617. [26] E. Alibakhshi, E. Ghasemi, M. Mahdavian, Corrosion inhibition by lithium zinc phosphate pigment, Corros. Sci. 77 (2013) 222–229. [27] E. Alibakhshi, E. Ghasemi, M. Mahdavian, Sodium zinc phosphate as a corrosion inhibitive pigment, Prog. Org. Coat. 77 (7) (2014) 1155–1162. [28] M.A. Abd El-Ghaffar, N.A. Abdel-Wahab, M.A. Sanad, M.W. Sabaa, High performance anti-corrosive powder coatings based on phosphate pigments containing poly(o-aminophenol), Prog. Org. Coat. 78 (2015) 42–48. [29] M. Puig, L. Cabedo, J.J. Gracenea, J.J. Suay, The combined role of inhibitive pigment and organo-modified silica particles on powder coatings: mechanical and electrochemical investigation, Prog. Org. Coat. 80 (2015) 11–19. [30] F. Askari, E. Ghasemi, B. Ramezanzadeh, M. Mahdavian, Effects of KOH:ZnCl2 mole ratio on the phase formation, morphological and inhibitive properties of potassium zinc phosphate (PZP) pigments, J. Alloys Compd. 631 (15) (2015) 138–145. [31] B. Hirschorna, M.E. Orazema, B. Tribollet, V. Vivier, I. Frateurc, M. Musianid, Determination of effective capacitance and film thickness from constantphase-element parameters, Electrochim. Acta 55 (2010) 6218–6227. [32] M. Mahdavian, M.M. Attar, Another approach in analysis of paint coatings with EIS measurement: phase angle at high frequencies, Corros. Sci. 48 (12) (2006) 4152–4157. [33] J.P.G. Farr, M. Saremi, Molybdate in aqueous corrosion inhibition I: effects of molybdate on the potentiodynamic behaviour of steel and some other metals, Surf. Coat. Technol. 19 (2) (1983) 137–144. [34] R. Naderi, M.M. Attar, Application of the electrochemical noise method to evaluate the effectiveness of modification of zinc phosphate anticorrosion pigment, Corros. Sci. 51 (8) (2009) 1671–1674. [35] Y. Hao, F. Liu, E.-H. Han, S. Anjum, G. Xu, The mechanism of inhibition by zinc phosphate in an epoxy coating, Corros. Sci. 69 (2013) 77–86. [36] D.M. Lenz, C.A. Ferreira, M. Delamar, Distribution analysis of TiO2 and commercial zinc phosphate in polypyrrole matrix by XPS, Synth. Met. 126 (2) (2002) 179–182. [37] K. Aramaki, The inhibition effects of cation inhibitors on corrosion of zinc in aerated 0.5 M NaCl, Corros. Sci. 43 (8) (2001) 1573–1588. [38] C.-Y. Tsai, J.-S. Liu, P.-L. Chen, C.-S. Lin, A two-step roll coating phosphate/molybdate passivation treatment for hot-dip galvanized steel sheet, Corros. Sci. 52 (10) (2010) 3385–3393. [39] M.V. Vinnichenko, M.T. Pham, T. Chevolleau, L.V. Poperenko, M.F. Maitz, In situ ellipsometric investigation of stainless steel corrosion behavior in buffered solutions with amino acids, Appl. Surf. Sci. 207 (1) (2003) 176–182. [40] A. Ghicov, H. Tsuchiya, J.M. Macak, P. Schmuki, Titanium oxide nanotubes prepared in phosphate electrolytes, Electrochem. Commun. 7 (5) (2005) 505–509. [41] T. Yamashita, P. Hayes, Analysis of XPS spectra of Fe2+ and Fe3+ ions in oxide materials, Appl. Surf. Sci. 254 (8) (2008) 2441–2449.

122

F. Askari et al. / Progress in Organic Coatings 85 (2015) 109–122

[42] R. Naderi, S.Y. Arman, S.H. Fouladvand, Investigation on the inhibition synergism of new generations of phosphate-based anticorrosion pigments, Dyes Pigments 105 (2014) 23–33. [43] Y. Shao, C. Jia, G. Meng, T. Zhang, F. Wang, The role of a zinc phosphate pigment in the corrosion of scratched epoxy-coated steel, Corros. Sci. 51 (2) (2009) 371–379.

[44] S. Mandel, A.C. Tas, Brushite (CaHPO4 ·2H2 O) to octacalcium phosphate (Ca8 (HPO4 )2 ·(PO4 )4 ·5H2 O) transformation in DMEM solutions at 36.5 ◦ C, Mater. Sci. Eng. C 30 (2010) 245–254. [45] D. Chen, S. Ma, H. Yu, A. Yuan, A. Liao, J. Wu, Thermal decomposition kinetic and electrochemical properties of KZn2 (PO4 )(HPO4 ), Adv. Mater. Res. 152 (2011) 184–191.