Stabilization Effect of Interfacial Solute Segregation on θ′ Precipitates in Al-Cu Alloys

Abstract

The effects of Sc, Mg and Si elements in an Al-Cu alloy have been studied by means of hardness tests and transmission electron microscopy analysis. The experimental results show that additions of Sc, Mg and Si can improve the heat resistance of the Al-Cu alloy. The Sc/Mg/Si segregation-sandwiched structure is the most stable, when compared with Sc segregation or Si/Sc co-segregation at the interface of θ′/Al. The additions of Si and Mg promote the aging–hardening response of the Al-Cu alloy. Mg is a micro-alloying element with great potential in stabilizing the size of θ′ phases, which further promotes the number density greatly. Consequently, the Al-Cu alloy achieves a high strength, matched with excellent thermal stability, due to the microalloying of Sc/Mg/Si solutes.

1. Introduction

Age-hardening Al-Cu alloys are used extensively in the automotive industry and aerospace industry [1,2,3,4]. As the significant strengthening precipitates in Al-Cu alloys, θ′-Al₂Cu phases often coarsen rapidly, resulting in the softening of Al-Cu alloys due to high-temperature conditions (above 150 °C) [5], which limits the engineering applications for Al-Cu alloys. Therefore, improving the stability of θ′ phases has become a topical issue in research and an urgent problem to be solved. Element segregation is an effective strategy to improve the thermal stability of the precipitated phase [6,7,8]. Results in the literature have shown that the segregation of Ag or Zr at the interface between precipitated phases and the Al matrix makes the structure of the Ω or θ′ phase more stable [9,10]. Adding Ag atoms to Al-Cu alloys forms a kind of Ω phase, which precipitates along the <111>_Al. Its precipitation increases the heat resistance of the Al-Cu alloy to more than 200 °C, which is mainly due to the segregation of Ag atoms at the interface of Ω/Al, improving the thermal stability of the Ω phase [11]. Similarly, Zr atoms tend to segregate at the θ′/Al interface in Al-Cu alloys; the segregation of Zr atoms prevents Cu atoms from diffusing into θ′ phases. Therefore, the thermal stability of θ′ phases is greatly enhanced [10].

It is reported that the segregation of Sc at the interface of θ′/Al also improves the thermal stability of θ′ phases and, sometimes, dispersed nano-scale Al₃Sc will form, which contributes to the thermal stability of Al-Cu alloys [6,10]. Microalloying with Mg in an Al-Cu alloy can promote high-temperature strength [12], but the underlying reason is not yet clear. Si can effectively accelerate the aging response [13]; at the same time, the addition of Si may form some atomic clusters that serve as sites for nucleation [14], which is helpful for increasing the number density of precipitates. Due to the addition of trace Si, the aging precipitation behavior of Al-Cu-Mg alloys is significantly altered [15], and some precipitates with better thermal stability will be formed. For example, the morphology and structure of the σ (Al₅Cu₆Mg₂) phase remains stable above 265 °C [15] and some Q-type phases (although its composition is still controversial, Al₃Cu₂Mg₉Si₇ is widely accepted) may be formed by changing the Si content, which enhances the thermal stability of Al-Cu-Mg alloys at elevated temperature conditions [16]. Whether the co-addition of Sc/Mg/Si elements in Al-Cu alloys can further improve the thermal stability synergistically is worth studying. This paper systematically reports the synergistic effects of Sc, Mg and Si elements on the heat resistance of the Al-4.0Cu alloy herein.

2. Material and Methods

The experimental alloys were prepared using traditional casting techniques. High-purity Al (99.999%) and Mg (99.999%), Al-50%Cu, Al-27%Si, and Al-2%Sc master alloys were used to product the designed alloys. The raw material was melted at 780 °C, then poured into an iron mold, and finally slowly cooled to room temperature to obtain alloy ingots. The actual chemical compositions of all studied alloys were analyzed.

Table 1. The compositions of all studied alloys.

Alloys	Actual Composition (wt.%)
	Cu	Mg	Si	Sc	Al
1#—Al-4.0Cu	3.93	/	/	/	Bal.
2#—Al-4.0Cu-0.1Sc	3.90	/	/	0.10	Bal.
3#—Al-4.0Cu-0.15Si-0.1Sc	4.21	/	0.18	0.10	Bal.
4#—Al-4.0Cu-0.5Mg-0.15Si-0.1Sc	4.01	0.51	0.17	0.11	Bal.

shows the nominal and actual chemical compositions of all studied alloys and, for convenience, the alloys are designated as alloys 1#, 2#, 3# and 4#, respectively.

The two-step solution treatment process for all the samples was carried out as follows. Firstly, the samples were kept at 505 °C for 12 h and then heated to 540 °C and kept at that temperature for 5 h; finally, they were quickly cooled using water cooling. The samples were isothermally aged at 175 °C. Thermal stability was tested using thermal exposure at 225 °C for samples that had undergone peak aging at 175 °C. The age–hardening responses (mechanical properties) were characterized using a microhardness tester with a 200 g load and 10 s dwell time. The average of the five hardness values was calculated as the hardness level of samples in a certain state.

The microstructure of samples was characterized using transmission electron microscopy (TEM) (model FEI-G2), with a working voltage of 300 kV. The samples for TEM observation were prepared using twin-jet electrochemical equipment (model RL-1 type twin-jet instrument). The working current of the twin-jet electrochemical equipment was ~80 mA. Electrolyte was prepared from nitric acid/methanol in a 1/3 volume ratio, then cooled to ~15 °C using liquid nitrogen.

A detailed statistical analysis was performed on the θ′ phase (microstructure images recorded in <001>_Al orientations). The average diameter, thickness and aspect ratio (ratio of diameter to thickness) of θ′ phases were calculated. No fewer than 50 θ′-phase sizes were counted for each state of the alloys.

3. Results

Figure 1 displays microhardness–time curves of all experimental alloys during isothermal aging at 175 °C. As is shown in Figure 1, the hardness value of the alloy1# increases slowly to the peak hardness value within 64 h, and then the hardness value decreases rapidly. The hardness growth rate of alloy 2# in 1–64 h is greater than that of alloy 1#. The hardness of the alloy 2# decreases almost within 1000 h after reaching the peak hardness. The peak hardness of alloy 3# is obviously improved, compared with those of the alloys 1# and 2#. Alloys 3# and 4# have similar aging response rates within 4 h, which are much faster than those of alloys 1# and 2#. Because both the alloys 3# and 4# contain Si, it is reported that Si can promote the aging response in Al alloys [13]. The hardness of alloy 4# is still increasing rapidly after alloy 3# reaches peak hardness, which may be attributed to the addition of Mg. After aging at 175 °C for nearly 1000 h, alloy 4# is more stable than alloy 3#.

Figure 2 presents microhardness–time curves of all experimental alloys during thermal exposure at 225 °C. As can be seen from Figure 2, the hardness value of alloy 1# drops sharply from ~103 HV in peak-aged condition to ~62 HV, which is almost close to the hardness value of its solution state. The hardness value of alloy 2# decreases slowly in the early stages of thermal exposure at 225 °C, and then remains relatively stable from 3 h to 96 h. Finally, the hardness of alloy 2# decreases significantly until the thermal exposure time reaches ~1000 h, but it is obvious that the hardness of alloy 2# is still higher than the final hardness value of alloy 1# at 225 °C for ~800 h. During the initial period of thermal exposure at 225 °C, alloy 3# exhibits a slow decline which is similar to that of the alloy 2#, and then the hardness of alloy 3# remains basically stable. It is not until ~800 h that a negligible decline occurs. When isothermal aging occurs at 175 °C, the peak-aged hardness of alloy 4# is far larger than that of the alloy 1#, by ~40 HV, and the hardness of alloy 4# does not decrease significantly at 225 °C (thermal exposure process) until ~800 h. Taken together, the hardness of alloy 4# is the highest among those of alloys 1#–4# at 225 °C (thermal exposure). Alloys 2# and 3# exhibit similar thermal stability, and their hardness remains stable or decreases slightly between 3 h and 800 h at 225 °C (thermal exposure). Consequently, alloy 1# is worst and alloy 4# is the most outstanding, in terms of thermal stability and mechanical properties.

Figure 3 demonstrates the microstructures of all studied alloys after isothermal aging at 175 °C for ~1000 h. Microstructural images are recorded from <100>_Al. The θ′ phases are the sparsest in distribution and the largest in size compared to those in other alloys after isothermal aging at 175 °C in Figure 3a. The size of θ′ phases in alloys 2# and 3# decreases obviously and the number density increases distinctly, as shown in Figure 3b,c. The θ′ phase of the alloy 4# is smallest and the most numerous in Figure 3d. In addition, plenty of dispersively distributed Q′ phases (marked with white arrows) and some cubic σ (Al₅Cu₆Mg₂) phases (marked with red arrows) can be seen in Figure 3d.

Figure 4 reports the microstructures of all studied alloys after being thermally exposed at 225 °C for ~800 h, and microstructural images are also recorded from <100>_Al. After being thermally exposed at 225 °C, the coarsening of θ′ phases in alloy 1# is the most severe, and the number density of θ′ phases is the lowest in alloy 1#. It can be seen from Figure 4b that there are many finer θ′ phases in alloy 2# and the number density of θ′ phases is significantly greater than that in alloy 1#. The coarsening degree of θ′ phases in alloy 3# is lower than that in alloy 2#. It is surprising that almost no coarsening occurs in alloy 4#. Simultaneously, Q′ phases and some cubic σ phases are clearly visible and without coarsening, especially for σ phases in Figure 4d.

The average size of θ′ phases is exhibited in detail in Figure 5 after isothermal aging (at 175 °C/~1000 h) and thermal exposure (at 225 °C/~800 h), respectively. The average diameter and thickness sizes of θ′ phases in alloy 1# are 513 nm and 16 nm, respectively, after aging at 175 °C for ~1000 h. After being thermally exposed at 225 °C for ~800 h, the sharp coarsening of θ′ phases occurs and their diameter and thickness almost doubles (969 nm and 35 nm, respectively) in alloy 1#. The sizes of diameter and thickness in alloy 2# are reduced by 104 nm and 4 nm, compared with those in alloy 1# after isothermal ageing (175 °C/~1000 h), respectively. After thermal exposure (225 °C/~800 h), the diameter and thickness sizes in alloy 2# are reduced by up to 220 nm and 13 nm, respectively. The addition of Sc makes the average sizes of θ′ phases significantly refined in alloy 2#, especially after thermal exposure; the average diameter and thickness of θ′ phases decrease by 22.7% and 37.1%, respectively, compared with those of the alloy 1#. The results show that the addition of Sc has a much more obvious inhibition effect on thickness than on diameter. The co-addition of Si and Sc into Al-4.0Cu alloy results in the average diameter of the θ′ phases being reduced by 132 nm and their thickness being decreased by 43.8% in alloy 3#, compared with those in alloy 1#, after isothermal aging at 175 °C for ~1000 h. After being thermally exposed at 225 °C for ~800 h, the size of the θ′ phases in alloy 3# is still smaller than that in alloy 2#. The addition of Mg refines the size of θ′ phases in alloy 4# compared with alloy 3#; the diameter is reduced from 381 nm to 55 nm and the thickness from 9 nm to 3 nm after aging at 175 °C for ~1000 h. The average size of θ′ phases in alloy 4# hardly changes after thermal exposure at 225 °C. The above average size quantitative analysis shows that the θ′ precipitates in alloy 4# are the most stable. These results are consistent with the hardness curves (Figure 1 and Figure 2) and the evolution of microstructures (Figure 3 and Figure 4), as analyzed above.

The aspect ratio (the ratio of diameter/thickness) of the θ′ phases in all four studied alloys under various heat treatment conditions is exhibited in

Table 2. Aspect ratio of θ′ phases in all four studied alloys at different states.

Alloys	175 °C/~1000 h	225 °C/~800 h
1#—Al-4.0Cu	32	28
2#—Al-4.0Cu-0.1Sc	34	34
3#—Al-4.0Cu-0.15Si-0.1Sc	42	35
4#—Al-4.0Cu-0.5Mg-0.15Si-0.1Sc	18	17

. As can be seen in Table 2, the aspect ratios of θ′ phases in alloys 1#, 2# and 3# gradually increase from 32 to 42 after isothermal aging (175 °C/~1000 h). Unexpectedly, the aspect ratio of the θ′ phases in alloy 4# suddenly drops to only 18. A similar phenomenon occurs after thermal exposure (225 °C/~800 h).

It has been pointed out in the literature that θ′ phases are metastable, plate-shaped intermetallic compounds and that the aspect ratio of θ′ phases decreases after coarsening at elevated temperatures, thus leading to the strengthening effect of θ′ phases being weakened [17]. In our present experiment’s results, the evolution of the aspect ratio of the θ′ phases in alloys 1#–3# is completely consistent with the results of previous literature. However, the aspect ratio in alloy 4# is different from the conclusions in the literature. The size of the θ′ phases is obviously refined, but its aspect ratio is greatly reduced in alloy 4#. We believe this is because the addition of Mg reduces the diameter of the θ′ phases more distinctly. The thickness of the θ′ phases in alloy 4# is one third of that of alloy 3#, but the diameter is only one eighth of that of alloy 3# after 175 °C/~1000 h, and the results are analogous after thermal exposure (225 °C/~800 h), as shown in Figure 5. The size of the θ′ phases decreases sharply due to the addition of Mg, so the hardness value of alloy 4# is much higher than that of alloy 3#.

In order to further reveal the reason of the size evolution of the θ′ precipitates in all studied alloys, the distribution of elements in the θ′ phases are characterized (element mappings are also acquired from <100>_Al and the Al element is not shown here). Figure 6 shows element distribution of the θ′ precipitates in alloy 3# after being thermally exposed (225 °C/~800 h). As shown in Figure 6b,d, in addition to detecting the segregation of Cu in θ′ precipitates, which is well known, the Sc atoms are conspicuously segregated at the interface of θ′/Al and the segregation of Si is faintly visible in Figure 6c. Alloy 1# does not contain Sc; thus, Sc segregation is the most fundamental reason for enhancing the thermostability of the θ′ phases in alloys 2# and 3# (Sc segregation has also been observed in alloy 2# and not shown here). In the literature [6,18,19], Si segregation like Sc was also observed by means of three-dimensional atomic probe (3DAP) and transmission electron microscopy (TEM) in Al-Cu-Si-Fe-Sc alloys; however, the researchers believe that the segregation of Si with low interfacial excess has only a little effect on the inhibition of θ′ phase coarsening. Because the Si distribution tends to spread out spatially around the interface [6], the main role of Si is to work as a facilitation element to attract Sc into the θ′ precipitates. The size difference of the θ′ phase in alloys 2# and 3# is small. In this paper, our experimental results also agree well with their conjecture.

Figure 7 reveals the element distribution of the θ′ precipitates in alloy 4#, aged at 175 ℃ for ~1000 h (recorded along the <001>_Al). It is quite obvious that the segregations of Sc/Mg/Si occur simultaneously at the interface of θ′/Al in alloy 4#. The Sc/Mg/Si layers are similar to a sandwich structure. The θ′ phase is wrapped inside the sandwich structure.

4. Discussion

4.1. Stabilization Effect of Solute Segregation

According to the results in Figure 3, Figure 4 and Figure 5, the addition of Sc (in alloys 2# and 3#) inhibits the coarsening of θ′ phases, compared with alloy 1#. There is only a small difference in the average sizes of θ′ precipitates between alloys 2# and 3#, due to the segregation of Sc, as shown in Figure 6d, while the striking difference lies in the number density of the θ′ precipitates (Figure 3b vs. Figure 3c) because of the addition of Si. The co-addition of Si and Sc (in alloy 3#) realizes both the refinement of the θ′ phases and an increase in the number density of the θ′ phases, compared to alloy 1#. This is because Si-Sc or Si-v (vacancy) clusters may be formed [20,21], which contribute to the nucleation of the θ′ phases on them, due to the addition of Si and Sc. The refinement effect of the θ′ phases is the most significant and the average size is smallest in alloy 4# after co-adding Sc/Mg/Si together into the Al-4.0Cu alloy (Figure 3d and Figure 4d). The most important reason is the formation of the sandwich structure (Figure 7). The structure has a higher stability than the Si/Sc or Sc segregation structure. The sandwich structure can further inhibit the diffusion of Cu atoms into the θ′ phase and, thus, inhibit their coarsening. Simultaneously, the sandwich structure may further reduce the interface energy.

Compared with alloy 3#, the segregation of Mg/Si in alloy 4# should be another crucial reason for the sharp drop in the size of the θ′ phases. The Mg element promotes size reduction of the diameter far more than that of the thickness. Therefore, the addition of Mg is responsible for the sharp decrease in the aspect ratio of the θ′ phases for alloy 4#. Although the deeper action mechanism of Mg in alloy 4# remains to be further studied, the conclusion can be drawn that Mg is a micro-alloying element with great potential in the stabilization of θ′ phases.

4.2. Effect of Precipitate Evolution

The difference in composition between alloys 4# and 3# is only 0.5%Mg, but the types of precipitated phases in alloy 4# change obviously. The addition of Mg promotes the formation of Q′ and σ phases in alloy 4# (see Figure 3d and Figure 4d). On the one hand, some Q′ phases form in the early stage [16,22] of isothermal aging at 175 °C and dispersively distribute in the Al matrix, serving as nucleating sites for the θ′ phases. So, the number density of the θ′ phases is further greatly increased in alloy 4#. As the most effective strengthening phases in the Al-Cu alloy, θ′ phases improve the strength of alloy 4#. On the other hand, some fine Q′ phases dispersed in the Al matrix play a role in precipitation strengthening, which is also an important reason that the strength of alloy 4# can be greatly improved. In addition, the heat resistance temperature of σ precipitates is up to 265 °C [15,23]; thus, σ precipitates make a certain contribution to the heat resistance of alloy 4# as well. As a result, the alloy 4# achieves an optimal match of strength and heat resistance.

It is worth noting that there are some research studies [24,25,26,27] indicating that the nanoscale Al₃Sc phases will precipitate in Al-Cu-Sc and Al-Sc alloys, which enhance the mechanical strength of these aluminum alloys. However, no Al₃Sc nanoparticles are observed in alloys 2#, 3# and 4#, as shown in Figure 3c,d and Figure 4c,d. Our experimental results also show that there are no Sc-containing precipitates (Al₃Sc) in the grain- and sub-grain boundaries, except some primary Sc-containing particles. In our current work, Sc elements dissolved in the Al matrix are preferentially distributed around the θ’ phases, which reduces interfacial energy of θ’/Al [18]. In some studies in the literature [18,28,29], some dissolved Sc atoms in Al-Cu-Sc alloys will be precipitated in the form of Al₃Sc particles, due to the high content of Sc (>0.2wt.%). The content of Sc in our studied alloy is only 0.1 wt.%, so there are not enough Sc atoms to form Al₃Sc precipitates. The more important reason is that the precipitation of Al₃Sc requires a temperature of more than ~300 °C [30]. Our previous work [31] revealed the competition between precipitation of Sc (formed Al₃Sc) and Sc segregation at the interface of θ’/Al. Therefore, the formation of Al₃Sc is related to the amount of Sc added and the temperature of heat treatment. The heat treatment temperature in our experiment is only 225 °C or below, so no Al₃Sc precipitate contributed to the mechanical behavior.

5. Conclusions

The aging behavior of the Al-Cu alloy and the effects of solute elements (Sc, Mg and Si) on the thermal stability of θ′ phases have been studied systematically with the methods of hardness test and TEM herein. The main conclusions are summarized as follows:

The addition of Si boosts the aging–hardening effect of the Al-4.0Cu alloy, and the segregation of Sc at the interface of θ′/Al improves the structural stability of the θ′ phases. The sandwich structure of Sc/Mg/Si elements facilitates the thermal stability of the θ′ phases. The Sc/Mg/Si co-addition enhances the strength and thermostability of the Al-4.0Cu alloy. The formation of Q′ and σ precipitates due to adding Mg is partly responsible for the excellent properties of the Al-4.0Cu alloy, when micro-alloyed with Sc/Mg/Si. Mg is a potential element in the stabilization of θ′ phases. This paper provides scientific and reasonable ideas for the design of heat-resistant Al alloys.