Amino acids have a number of desirable properties for use in formulation of MWFs. If used as alkali salts, they serve as mild bases and can thus be used to buffer the pH of common acidic CIs to values around 8–9 (just as the above-mentioned amino alcohols). Several derivatives are readily available in large tonnages at low cost. In addition, most amino acids are non-toxic, non-hazardous, non-volatile and chemically stable. Following these criteria, we selected a set of simple amino acids for evaluation as alkaline additives to aqueous MWFs. Within this set of amino acids, glycine (Gly) was chosen as the simplest, alanine (Ala) as an alkyl-substituted, arginine (Arg) and lysine (Lys) as basic, glutamate (Glu) as acidic, serine (Ser) and cysteine (Cys) as donor-substituted and histidine (His) as heteroaromatic proteinogenic amino acid. In addition, 6-amino hexanoic acid (AHX) as a non-proteinogenic higher homologue and taurine (Tau) as a α-amino sulfonic acid were also included. All amino acids were used in the form of their sodium salts.
Selection of amino acids via chip filter test
For a fast selection of the most promising combinations of amino acids with CIs, we performed a standard chip-filter-test on grey cast iron. The chip filter test according to DIN 51360-2 is a standard method used in industry for evaluation of anticorrosive properties of MWFs for iron. Briefly, a defined amount of sieved grey cast iron chips is placed on a round paper filter and submerged in a hard water solution of an acidic CI and an alkaline additive. After incubation for 2h at room temperature and subsequent washing of the filter, corrosion marks are visually detected and scored between 0 (no corrosion marks) and 4 (strong staining of the filter due to corrosion). This assay is operationally very simple and can be easily used for medium throughput screening. In addition, it is quite sensitive, due the large surface area and the high corrosivity of grey cast iron chips. However, the optical detection of staining is slightly subjective and misinterpretation according to staining that is not caused by corrosion is possible. The chip filter test was therefore used for a preselection of promising mixtures only. These were afterwards evaluated in more detail using electrochemical measurements and gravimetry. Tables with all corrosion scores obtained by chip filter assay are available in the supporting information (Tables S2 and S3). As a benchmark system we used an aqueous solution of 1,3,5-triazine-2,4,6-triaminocaproic acid (2, 64 mmol/L corresponding to 3 wt%) with TEA (3.6 or 4.5 molar equivalents respectively) as an alkaline additive. This mixture is used successfully in several commercial MWFs for applications involving steel tools. It has a relatively low pH value of 7.4–7.8 and shows complete corrosion protection (score = 0) in the chip filter assay. In brief, the following observations were made: 1. Organic additives such as TEA are essential for the performance of acidic CIs. When TEA was substituted by NaOH as alkaline additive to CI 2, a corrosion score of 4 was obtained although the pH of this mixture (pH = 10.3, suggesting improved alkaline corrosion protection) was significantly higher than that of TEA/2 (pH = 7.4). 2. Among the amino acids tested sodium salts of Tau, Glu and AHX were as effective as TEA (scores = 0 at 3.6 molar equivalents) all others were less effective (scores = 1–4). 3. Corrosion scores for all amino acids tested were improved to 0–1 with a slight increase in concentration from 3.6 to 4.5 molar equivalents. 4. Amino acids are also compatible with acidic CIs other than 2. We have tested three commercial CIs 1, 2 and 5 containing 1–3 carbocyclic acid groups as well as two recently described carboxyphosphonic acids 3 and 4 (Ruf et al. 2022) and obtained perfect corrosion scores for combinations with sodium salts of Tau, Glu and AHX in all cases. 5. All amino acids gave perfectly clear and colorless solutions upon formulation with acidic CIs. However, after storage of 1 month at room temperature, the mixtures containing Met showed a yellow color, indicating oxidative degradation processes. 6. Several mixtures with Gly gave an untypical uniform color reaction on filter paper after incubation with grey cast chips lacking the normal contoured staining pattern of corroded iron chips (see supporting information, Fig. S4). We attributed this finding to the formation of colored water soluble Gly-complexes with alloying metals, which might have been formed as a consequence of Gly-mediated metal leaching from grey cast iron chips. Of all amino acids tested Tau, Glu and AHX showed the most promising properties and were selected for further evaluation by electrochemical measurements.
Leaching properties of amino acids for Co, Ni and Cu
Leaching of alloying metals is an undesired property of MWFs. Cobalt leaching, for example, causes increased wear of tools and is a serious environmental and health hazard (Barceloux 1999). Given the observations made for Gly in the chip filter assay, we analyzed the leaching ability of several amino acids at slightly alkaline pH values. In addition, we evaluated selected mixtures with acidic CIs. As benchmarks we compared the leaching abilities of amino acids with TEA, AMP and MIPA. The latter two are marketed as low leaching additives for Co-containing alloys. In a first set of experiments, the appropriate metal powder was suspended with an aqueous solution of the alkaline additive and heated to reflux for 24h. Dissolved metal concentrations were measured via FAAS and are shown in Table 1.
The results confirm the low leaching capability of AMP and MIPA for Co and Ni (Table 1, entries 3–4). TEA in contrast leads to moderate concentrations of dissolved Co and Ni (Table 1, entry 2), which has been noted by other researchers before and attributed to the formation of water soluble TEA-complexes at least for Co (Zhang et al. 2010). For Cu, the trend is inverted and TEA leads to lower amounts of dissolved Cu than AMP and MIPA. Sodium glycinate on the other hand leads to exceptionally high levels of Co and Ni in solution and can thus be classified as a high leaching additive for these metals (Table 1, entry 5). Leaching of Cu is moderate in this case. These findings are in accordance with the application of Gly as a Co-chelator for recycling of Co from lithium ion batteries (Nayaka et al. 2016). For comparison, we measured the leaching properties of N-methylglycine (Sarcosine, Sar, Table 1, entry 6) and N-dimethylglycine (DMG, Table 1, entry 7) and observed decreasing concentrations of dissolved Co and Ni with increasing steric demand of the Gly derivatives. Similar trends were observed with other proteinogenic amino acids (Table 1, entries 8–11): All amino acids tested with either increased steric hindrance through side chains at the α-carbon or the amino group are less corrosive to Co and Ni compared to Gly. It is of note, that a similar effect of sterically demanding sidechains is not observed for Cu-leaching. The leaching of Co and Ni is significantly increased for all amino carboxylic acids tested compared to amino alcohols like TEA, MIPA or AMP. The observed difference in leaching of Co and Ni versus Cu is reflected by the corresponding stability constants of amino carboxylic acid complexes with Co(II), Ni(II) and Cu(II) (Gergely et al. 1972). The reported complex stability constants reveal a small difference in stability for Gly complexes of Cu(II) compared to complexes with sterically more demanding α-amino acids. In contrast, significant differences in complex stability have been reported for Ni(II) and Co(II)-complexes of amino acids with different steric demand. For both metals, Gly-complexes were found to be the most stable. The observed dependence of complex stability from steric factors is most likely a consequence of a different complex geometry for Co(II), Ni(II) and Cu(II)-complexes. Amino acid complexes of Co(II) and Ni(II) have a 3:1-stoichiometry of ligand:metal with an octahedral coordination of the metal by three bidentate α-amino carboxylate ligands (Gu et al. 2007). Whereas amino acid complexes of Cu(II) have a 2:1-stoichiometry of ligand:Cu with a planar coordination of the metal by two bidentate α-amino carboxylate ligands (Casari et al. 2004). The latter planar arrangement is less crowded and therefore less sensitive to the steric impact of bulky side chains. Of all amino acids tested, Tau and AHX stand out with respect to their low corrosivity towards Co, Ni and Cu. While the α-amino sulfonic acid Tau (Table 1, entry 13) lead to concentrations of dissolved metal in the same range as the benchmark amino alcohol TEA, AHX as a mid-chain ω-amino carboxylic acid (Table 1, entry 12) leads to extremely low dissolved metal concentrations which were even lower than values obtained for the low leaching additives AMP and MIPA. Both non-proteinogenic amino acids show thus a drastically reduced corrosivity towards all three metals tested. Again, this decrease in corrosivity (and consequently improved leaching properties) can be explained with the corresponding complex chemistry as both Tau and AHX are less powerful chelate ligands compared to the proteinogenic α-amino acids mentioned before. We have also tested mixtures of various acidic CIs with AHX for leaching properties. All AHX combinations tested resulted in low concentrations of dissolved Co and Ni qualifying AHX as an excellent low leaching additive particularly for applications of MWFs with Co-hardened tools. Moderate corrosivity and accordingly higher concentrations of dissolved metal was observed for Cu only (Table 1, entries 14–18).
Table 1
Co, Ni and Cu-concentrations after extraction of the appropriate metal powder into aqueous test solutions after 24 h at reflux. Metal concentrations were measured as triplicate from the aqueous extracts via flame atomic absorption spectroscopy (FAAS) and are given as mean values
|
additive[a]
|
pH
|
Co (mg/L)
|
Cu (mg/L)
|
Ni (mg/L)
|
1
|
NaOH
|
9.6
|
0.67
|
≤ 0.20
|
≤ 0.20
|
2
|
TEA
|
10.2
|
34.3
|
1.57
|
1.44
|
3
|
AMP
|
12.8
|
7.03
|
43.2
|
0.33
|
4
|
MIPA
|
12.6
|
0.82
|
26.4
|
1.32
|
5
|
Gly
|
9.6
|
2486
|
24.8
|
963
|
6
|
Sar
|
9.6
|
419
|
70.6
|
328
|
7
|
N,N-Dimethyl-Gly
|
9.6
|
207
|
154
|
75
|
8
|
Lys
|
10.8
|
501
|
109
|
166
|
9
|
Ser
|
9.6
|
592
|
0.33
|
231
|
10
|
Glu
|
9.4
|
454
|
86.0
|
192
|
11
|
Pro
|
9.6
|
438
|
239
|
318
|
12
|
AHX
|
10.1
|
≤ 0.20
|
0.81
|
≤ 0.20
|
13
|
Tau
|
8.7
|
34.4
|
5.76
|
8.92
|
14
|
1/AHX[d]
|
9.8
|
4.04
|
4.66
|
84.3
|
15
|
2/AHX[b]
|
9.8
|
3.76
|
5.53
|
53.3
|
16
|
3/AHX[b]
|
10.0
|
0.63
|
0.82
|
13.4
|
17
|
4/AHX[c]
|
9.9
|
0.72
|
0.70
|
152
|
18
|
5/AHX[b]
|
9.9
|
1.19
|
1.15
|
135
|
[a] Performed with 1wt% of the alkaline additive. Amino acids were used as sodium salts. [b] Performed with 3wt.% acidic CI and 4.5 molar equivalents AHX sodium salt. [c] Performed with 3wt.% acidic CI and 1.5 molar equivalents AHX sodium salt. [d] Performed with 3wt.% acidic CI and 3 molar equivalents AHX sodium salt. |
Electrochemical evaluation of amino acids as alkaline additives
We started the electrochemical evaluation of amino acids in hard water (20 dGH) under the same conditions used for the chip filter assay mentioned above. Tricarboxylic acid 2 was chosen as acidic CI. The combination with TEA as an alkaline additive is an industrial benchmark system for corrosion inhibition of iron and steel with high efficacy in hard water. Accordingly, we have not even been able to determine the limiting effective concentration of 2/TEA in hard water by electrochemical methods and observed perfect protection even at low millimolar concentrations (Table 2, entry 3). Lower concentrations of 2 lead to too low conductivity of hardwater solution to obtain reproducible potentiodynamic results. Under the experimental settings chosen (hard water and S235JR-steel), the electrochemical evaluation is less sensitive to corrosion compared to the chip filter assay. An important difference is the relatively small and more homogenous surface area exposed to the corrosive media in the electrochemical setup, where a polished bulk metal serves as a working electrode. Even the lowest concentrations of the acidic CI 2 tested, were thus likely far above the limiting effective concentration for complete corrosion protection. The choice of alkaline additive has therefore almost no influence on the protective effect of CI 2 (at least during the measurement time span of 24 h). Even neutralization of CI 2 with NaOH lead to perfect corrosion inhibition (Table 2, entry 2). The same perfect anticorrosive effect was observed for the addition of sodium glutamate (Table 2, entry 4) confirming, that the use of glutamate as a substitute of TEA has at least no negative effect on the corrosion protection of CI 2. More interesting was the influence of the alkaline additive on the carboxyphosphonic acid 3. We have previously reported, that 3 is a less powerful CI in hard water solution than 2 and attributed this loss of performance to the low hard water compatibility of carboxyphosphonic acid 3 leading to precipitation of hardly soluble salts with di- and trivalent cations (Ruf et al. 2022). At concentrations significantly higher than carboxylic acid 2, carboxyphosphonic acid 3 is thus less effective in corrosion protection of steel at least when neutralized with TEA (Table 2, entry 5). The time dependent OCP in Fig. 2A reveals a slow immobilization of 3 on steel accompanied by a steady decrease in potential, which is characteristic for the self-assembly of protective nanolayers by most amphiphilic phosphonic acids (Felhősi et al. 2002). However, after 16 hours, a rapid increase in value of potential is observed accounting for the partial degradation of the protecting layer on the metal surface. In parallel a milky suspension is formed indicating the precipitation of phosphonate salts. If TEA is substituted by sodium glutamate for neutralization of CI 3, we observed a decrease in value of OCP over several hours down to a constant value indicating the formation of stable protective layers (Fig. 2A). The potentiodynamic curve reveals a pronounced anodic shift of the corrosion potential (Fig. 2B) corresponding to almost complete corrosion protection of steel in this case according to Tafel analysis (Table 2, entry 4). Electrochemical impedance measurements confirm these findings. Bode plots of electrochemical impedance measurements show the high anticorrosive efficiency of tricarboxylic acid 2 which is independent from the additives NaOH, TEA or Glu under the conditions employed. For carboxyphosphonic acid 3, in contrast, the alkaline additive has an influence on the formation of protective layers and thus anticorrosive efficacy. The values of the Bode modules in the low frequency region are higher for 3/Glu compared to 3/TEA (Fig. 2C). The corresponding capacitive loop in the Nyquist plots is also larger (Fig. 2D). Both, high Bode moduli and large capacitive loops in Nyquist plots correlate with high corrosion resistance (Mansfeld 1990). We observed still a slightly milky suspension for 3/Glu, however significantly less precipitate was formed (compared to 3/TEA) suggesting that Glu improves the hard water compatibility of carboxyphosphonic acid 3 and serves as an antiscalant, similarly to other poly(carboxylic) acids (Jafar Mazumder 2020).
Table 2
Results of the potentiodynamic polarization measurements of S235JR-Steel after 24 h exposure against hardwater-solution (20 dGH, pH 8.3 ± 0.5) containing tricarbocxylic acid 2 or carboxyphosphonic acid 3 and various alkaline additives at room temperature
entry
|
CI, c (mmol/L)
|
additive
|
icorr (nA/cm2)
|
IEIcorr (%)
|
Ecorr (mV)
|
βa (mV/dec)
|
βc (mV/dec)
|
CR (µm/y)
|
1
|
-
|
NaOH[a]
|
37500
|
-
|
-584
|
544
|
1540
|
436
|
2
|
2, 2.2
|
NaOH[b]
|
43.8
|
99.89
|
-123
|
156
|
79.7
|
0.35
|
3
|
2, 2.2
|
TEA[b]
|
56.2
|
99.85
|
-141
|
184
|
76.9
|
0.65
|
4
|
2, 2.2
|
Glu[b]
|
12.5
|
99.97
|
-153
|
171
|
76.5
|
0.14
|
5
|
3, 18.8
|
TEA[b]
|
1790
|
95.23
|
-368
|
402
|
244
|
20.8
|
6
|
3, 18.8
|
Glu[b]
|
5.72
|
99.98
|
-142
|
186
|
61.5
|
0.15
|
[a] pH was adjusted with 1 M NaOH to 8.0; [b] 4.5 molar equivalents were used. Glu was used as disodium salt. |
Although the observed antiscaling effect of Glu with carboxyphosphonic acid 3 in hard water was interesting to note, it was difficult to confirm the synergistic additive effects observed in the chip filter test on the corrosion protection of more powerful CIs in hard water. We changed the media therefore to more corrosive aqueous 0.5% NaCl solution. Carboxyphosphonic acids like 3 or 4 have been shown to have good corrosion protective properties in chloride containing media. Both compounds were therefore selected as acidic CIs and evaluated in combination with NaOH, TEA and different amino acid salts as alkaline additives. We determined the limiting effective concentrations of these combinations to achieve complete corrosion protection by potentiodynamic measurements and a selection of data is presented in Tables 3 and 4. The limiting effective concentration for 4/NaOH was found to be around 30–35 mmol/L (Table 3, entries 2–4) for 3/NaOH this value is slightly higher at 35–40 mmol/L (Table 4, entries 2–3). For carboxyphosponic acid 4, the substitution of NaOH by TEA has a positive effect on the corrosion rate (Table 3 entries 5–6) lowering the effective limiting concentration to 20 mmol/L. No measurable effect has been seen for carboxyphosponic acid 3 and the limiting effective concentration remained at 35–40 mmol/L for 3/TEA (Table 4, entries 4–5). The same trend was observed when amino acid salts were used as alkaline additives: For carboxyphosponic acid 3, only Tau led to the same effective limiting concentration as NaOH. All other amino acid salts lead to a less effective corrosion protection with CI 3 and effective limiting concentrations between 60–70 mmol/L. The typical layer forming protective mechanism of amphiphilic phosphonic acids on steel is reflected by a steady decrease in OCP value over several hours as observed for 3/NaOH (Fig. 4A). Similar time dependent OCPs were observed for 3/Tau and with slightly higher values also for 3/TEA. For 3/AHX however, OCPs are significantly higher in value suggesting a less perfect layer formation. It is interesting to note, that the mixture of 3/AHX has a significantly lower critical micelle concentration (cmc) of 1.7 mmol/L (measured by 1H-NMR, see supporting information, Table S1) compared to 3/NaOH (30 mmol/L). Although not strictly comparable with the formation of protective monolayers on steel, this decreased cmc suggests a higher surface activity for 3/AHX compared to 3/NaOH which should come with better corrosion protection. The opposite effect was observed and other factors besides surface activity must be operating. Bode plots of electrochemical impedance measurements confirm the high efficiency of 3/NaOH, 3/TEA and 3/Tau as layer-forming CIs, because the values of the Bode modules in the low-frequency region are highest for these three mixtures (Fig. 4C). Their capacitive loops in the Nyquist plots are also the largest among the mixtures tested (Fig. 4D). As mentioned above, both high Bode moduli and large capacitive loops in Nyquist plots correlate with high corrosion resistance (Mansfeld 1990). Bode plots of all mixtures including carboxyphosphonic acid 3 reveal only one time constant, suggesting charge-transfer-controlled corrosion processes (Macdonald 2005). A comparison with the second carboxyphosphonic acid 4 reveals a different influence of the alkaline additive. All amino acid additives lead to a significant decrease of limiting effective concentrations compared to 4/NaOH. Again, AHX and also Gly were least effective and gave the highest limiting concentrations for the corresponding mixtures with 4, whereas Tau and Glu had the best effect on corrosion protection of 4 and gave the lowest limiting concentrations (Table 3, entries 7–10). The corresponding time dependent OCPs (all measured at a concentration of 20 mmol/L for CI 4) confirm the formation of protective layers with 4/Tau, 4/Glu and 4/TEA with a regular decrease of OCP values within a few hours to a low constant value around − 100 to -200 mV. For 4/AHX and 4/NaOH significantly higher OCP values (-300 mV to -400 mV) were observed suggesting the formation of less perfect protective layers on steel. The potentiodynamic curves shown in Fig. 3B confirm the trend in efficacy for the alkaline additives tested and their pronounced influence on the anodic corrosion reaction. The noblest potentials and lowest current densities were observed for 4/Tau and 4/Glu followed by 4/TEA, 4/AHX and 4/NaOH. The curves are significantly shifted in this series, suggesting a pronounced influence of the alkaline additive on the anticorrosive properties of 4. We have again measured the influence of the additive on the cmc of 4 in water by 1H-NMR, but observed only small differences (cmc4/NaOH = 4.5 mmol/L, cmc4/TEA = 4.7 mmol/L, cmc4/AHX= 7.8 mmol/L) which can again not account for the observed large differences in anticorrosive properties.
Table 3
Results of the potentiodynamic polarization measurements of S235JR steel after 24 h exposure to aqueous NaCl-solution (0.5wt.%, pH 8.3 ± 0.5) containing carboxyphosphonic acid 4 and various alkaline additives at room temperature
entry
|
4, (mmol/L)
|
additive
|
icorr (nA/cm2)
|
IEIcorr (%)
|
Ecorr (mV)
|
βa (mV/dec)
|
βc (mV/dec)
|
CR (µm /y)
|
1
|
-
|
NaOH[a]
|
26900
|
-
|
-511
|
136
|
144600
|
313
|
2
|
20
|
NaOH[a]
|
19200
|
28.62
|
-430
|
235
|
1020
|
223
|
3
|
30
|
NaOH[a]
|
536
|
98.01
|
-182
|
260
|
102
|
62.3
|
4
|
35
|
NaOH[a]
|
9.51
|
99.96
|
-153
|
251
|
79.1
|
0.111
|
5
|
15
|
TEA[b]
|
-[c]
|
-[c]
|
-[c]
|
-[c]
|
-[c]
|
-[c]
|
6
|
20
|
TEA[b]
|
11.4
|
99.96
|
-190
|
382
|
37.9
|
0.134
|
7
|
15[a]
|
Glu[b]
|
14800
|
44.98
|
-458
|
193
|
460
|
167
|
8
|
20[a]
|
Glu[b]
|
7.59
|
99.97
|
-116
|
237
|
69.5
|
0.092
|
9
|
10[a]
|
Tau[b]
|
10800
|
59.85
|
-314
|
269
|
272
|
125
|
10
|
15[a]
|
Tau[b]
|
3.80
|
99.97
|
-115
|
173
|
68.0
|
0.044
|
11
|
25[a]
|
AHX[b]
|
3030
|
88.74
|
-413
|
344
|
166
|
35.2
|
12
|
30[a]
|
AHX[b]
|
27.4
|
99.90
|
-221
|
476
|
87.7
|
0.032
|
13
|
35[a]
|
AHX[b]
|
3.93
|
99.99
|
-113
|
177
|
67.7
|
0.046
|
[a] pH was adjusted with 1 M NaOH to 8.0; [b] 4.5 molar equivalents were used. Amino acids were used as sodium salts; [c] inconclusive Tafel-Plot due to extensive pitting corrosion. |
Table 4
Results of the potentiodynamic polarization measurements of S235JR steel after 24 h exposure to aqueous NaCl-solution (0.5wt.%, pH 8.3 ± 0.5) containing carboxyphosphonic acid 3 and various alkaline additives at room temperature
entry
|
3, (mmol/L)
|
additive
|
icorr (nA/cm2)
|
IEIcorr (%)
|
Ecorr (mV)
|
βa (mV/dec)
|
βc (mV/dec)
|
CR (µm /y)
|
1
|
30
|
NaOH[a]
|
365
|
98.64
|
-296
|
978
|
93.5
|
4.18
|
2
|
35
|
NaOH[a]
|
364
|
98.65
|
-258
|
518
|
87.5
|
4.23
|
3
|
40
|
NaOH[a]
|
14.1
|
99.95
|
-164
|
183
|
71.7
|
0.164
|
4
|
35
|
TEA[b]
|
506
|
98.12
|
-581
|
38.9
|
106
|
5.88
|
5
|
40
|
TEA[b]
|
37.5
|
99.86
|
-240
|
263
|
68.2
|
0.436
|
6
|
30
|
Tau[b]
|
2650
|
90.15
|
-690
|
123
|
72.2
|
308
|
7
|
35
|
Tau[b]
|
253
|
99.06
|
-198
|
658
|
96.5
|
0.296
|
8
|
40
|
AHX[b]
|
1.420
|
94.57
|
-322
|
757
|
129
|
16.4
|
9
|
60
|
AHX[b]
|
1.470
|
94.54
|
-305
|
1270
|
146
|
17.1
|
10
|
70
|
AHX[b]
|
253
|
99.06
|
-297
|
802
|
115
|
11.1
|
[a] pH was adjusted with 1 M NaOH to 8.0; [b] 4.5 molar equivalents were used. Amino acids were used as sodium salts. |
Gravimetric evaluation of anticorrosive properties
As noted previously, the performance of layer-forming CIs can be time-dependent. Since many industrial applications require efficient corrosion inhibition over longer periods than those measured with our electrochemical studies, we included a long-term gravimetric assay with rectangular steel slides (S235JR) of 30 × 10 × 3 mm size. 2% aqueous NaCl was used as media, sufficiently corrosive to provide a weight loss of 186 mg within 12 weeks of incubation at room temperature and constant pH ~ 8 (TEA/AcOH buffer) when no additional CI was added. This amounts to a corrosion rate of 121 µm/y (Fig. 5A). All CIs tested were compared to this value, and the corresponding reduction in corrosion is given in Fig. 5A (red bars, inhibitory efficiency, IE) next to the corrosion rates (blue bars). We have used the TEA/AcOH buffer as a reference system, because the pH is in the same range as in our test solutions, and AcOH is known to have almost no anticorrosive properties at neutral or slightly alkaline pH values on iron and steel (Hefter et al. 1997). We have previously reported, that 2/TEA does not show good corrosion protection in chloride containing media (IE = 29%) whereas 3/TEA is moderately effective (IE = 66%) and 4/TEA is highly effective under these conditions (IE = 88%). The use of amino acid salts such as Glu instead of TEA improves the anticorrosive properties of 2 and 3 in NaCl solution significantly and leads to almost perfect corrosion inhibition with 3/Glu (IE = 92%). With 2, corrosion is reduced to a less perfect (IE = 47%) but still improved value, when Glu was used as an alkaline additive instead of TEA. However, visual inspection of the test specimens after incubation in 2% aqueous NaCl for 72 weeks revealed almost polished metal surfaces for 2/Glu (Fig. 5C), whereas large quantities of corrosion products were deposited for 2/TEA (Fig. 5B). As noted above, Glu serves as an antiscalant preventing the deposition of corrosion products on the metal surface.