MATERIALS TRANSACTIONS
Online ISSN : 1347-5320
Print ISSN : 1345-9678
ISSN-L : 1345-9678
Vacancy Behavior during Aging at 50 and 100°C in Al–Mg–Si Alloys with Excess Si Studied by Positron Annihilation Spectroscopy
Koji InoueKen TakataKoji IchitaniYasuharu Shirai
著者情報
ジャーナル フリー HTML

2019 年 60 巻 11 号 p. 2255-2259

詳細
Abstract

We have investigated the aging behavior of quenched-in vacancies in excess Si type Al–Mg–Si-alloys by coincidence Doppler broadening of positron annihilation radiation and positron lifetime spectroscopy. The chemical composition around the quenched-in vacancies is initially rich in Si. For aging at both 50°C and 100°C, the Mg/Si composition ratio around the vacancies increases with aging time. The final Mg/Si composition ratio around the vacancies was found to be almost the same for aging at both 50 and 100°C. The difference of the aging time dependence of the chemical composition around the vacancies at 50 and 100°C was observed in the initial stage, i.e., the vacancy-Si-rich solute complexes was formed at 50°C, while the formation of the vacancy-Si-rich solute complexes was avoided at 100°C. Therefore, the avoidance of the formation of the vacancy-Si-rich solute complexes by pre-aging around 100°C before storage at room temperature may be a key to avoid negative effect of artificial aging in Al–Mg–Si alloys.

1. Introduction

6000-series aluminum alloys, based on Al–Mg–Si alloys, are widely used for automotive applications because of their high age-hardening response in the paint baking process. The baking process is typically performed at 180°C after solution heat-treatment and annealing, so-called artificial aging. However, the age-hardening response decreases when the alloys undergo aging (storage) at room temperature before the artificial aging. Aging at room temperature, so-called natural aging, has negative interactions with the subsequent artificial aging step. This is recognized negative effect of two-step aging.1) It has been reported that this effect is caused by nano-clusters that form from the solid solution during the aging at room temperature.2) The nano-clusters formed at room temperature have a detrimental effect on the subsequent precipitation of the strengthening β′′ phase (the most important hardening phase in 6000-series) during the artificial aging, reducing the amount of β′′ phase, which leads to a lower peak hardness than if the natural aging could have been avoided.35)

In contrast to the natural aging, a pre-aging treatment at temperatures of 70–120°C before storage at room temperature can effectively eliminate the detrimental effect of the natural aging on the subsequent precipitation of the β′′ phases during the artificial aging.6) This is believed to be because the nano-clusters formed at temperatures higher than 70°C differ from those formed at lower temperatures.5,7) The nano-clusters formed at temperatures below and above 70°C are called low-temperature clusters and high-temperature clusters,8) respectively, or cluster (1) and cluster (2).7) Only cluster (2) can easily transform into the β′′ phase during artificial aging. The nature of these nano-clusters has not been completely realized because they are difficult to be observed. Therefore, there is still considerable debate regarding the complicated aging behaviors of Al–Mg–Si alloys.

The formation of the Mg–Si nano-clusters in the initial stage of the aging are observed by three-dimensional atom probe (3D-AP).8) In the case of small nano-clusters, spatial resolution of 3D-AP blurs the nano-clusters and the chemical composition of the nano-clusters strongly depends on the parameters employed in the cluster analysis. Moreover, 3D-AP cannot reveal that the nano-clusters include vacancy or not because of the limited detection efficiency of the 3D-AP apparatus.

Vacancies in Al alloys play a significant role in the formation of nano-clusters because the supersaturated solute atoms diffuse by interchanging positions with the vacancies and form the nano-clusters. Therefore, to understand the nature of the nano-clusters, it is important to understand the vacancy behaviors in the sequence from the formation of vacancy-solute complexes to the generation of nano-clusters, i.e., complementary information on the nano-clusters are obtained by investigating the vacancy behaviors, which lead to better understanding of the nano-clusters.

Positron annihilation spectroscopy is a highly sensitive method to detect vacancy-type defects.9,10) Existence of the vacancies and chemical information around vacancies can be obtained by positron annihilation spectroscopy. The detailed behavior of vacancy in the aging of Al–Mg–Si alloys was investigated by several researchers.9,11,12) Especially, Banhart et al. measured the behavior of vacancy in the very early stage of the aging by the mean positron lifetime systematically.9) They also measured momentum distribution by a single Doppler broadening technique because high count rate is necessary to investigate the very early stage of the aging.11) They did not measure it by coincidence Doppler broadening (CDB) which can measure chemical environment around vacancy more precisely but the data acquisition time needs longer. Therefore, change of chemical environment around vacancies during long time aging was not investigated by CDB.

In the aging at 50 and 100°C, strength is increased largely at long aging time. The change of strength is related with the change of chemical information on the nano-clusters such as chemical composition and chemical bonding because it affects the interaction between dislocations and the nano-clusters even if the nano-clusters are sheared or bypassed by dislocation. The investigation on the systematical change of the chemical information on the nano-clusters during the longtime aging is important to understand the origin of the strength changing.

Therefore, this study is not focused on the very initial stage of the aging but is focused on the relatively longer aging time dependence of behaviors of vacancies and their chemical environment during aging at temperatures below and above 70°C corresponding to the formation of cluster (1) and cluster (2).7) In this study, the behaviors of vacancies and their chemical environment during aging at 50 and 100°C are investigated by positron lifetime and CDB measurements. These measurements reveal that the nano-clusters contains vacancies. Therefore, systematical change of the chemical information on the nano-clusters during the aging are investigated from the change of the chemical environment around vacancies.

2. Experimental Procedure

The samples used in this study were Si-excess type Al–Mg–Si (AA-6022) alloys of 1 mm thick. The chemical compositions of the alloys are listed in Table 1. The alloys were solution-treated at 550°C for 30 min in an electrical furnace and subsequently quenched in ice-containing water. Isothermal aging at 50 and 100°C was performed in an oil bath immediately after the quenching.

Table 1 Chemical compositions of the excess Si type Al–Mg–Si alloy (mass%).

A positron source of 22NaCl deposited on a thin capton film was sandwiched between two identical samples for both positron lifetime and coincidence Doppler broadening (CDB) measurements. The strength of the 22Na positron source was about 1 MBq.

The positron lifetime measurements13) were carried out using a fast digital oscilloscope and BaF2 scintillators with a time resolution of about 190 ps at full width at half maximum (FWHM). A total of at least 1 × 106 events were accumulated for each measurement. All the positron lifetime and CDB measurements described in the following were performed at room temperature (20°C).

The CDB measurements14) were carried out using two Ge-detectors. In the present experiments, the overall energy resolution was about 0.9 keV at FWHM, which corresponds to a momentum resolution of about 3.5 × 10−3mc (FWHM), where m is the electron/positron rest mass and c is the speed of light. A total of at least 5 × 106 at least coincidence events were accumulated for each measurement. The low momentum component fraction (LMCF) and high momentum component fraction (HMCF) are defined as the ratios of the counts in the low momentum (|pL| < 4 × 10−3mc) and high momentum (12 × 10−3mc < |pL| < 20 × 10−3mc) regions in the CDB spectrum to the total counts, respectively. It is well known that the LMCF and HMCF are sensitive to vacancy-type defects. If the vacancy concentration is changed, positron trapping into these vacancies will result in a change in the LMCF and HMCF that lie on a straight line in a LMCF–HMCF plot. Normally, the LMCF in a vacancy is larger than one in the bulk aluminum and the HMCF in a vacancy is smaller than one in the bulk aluminum because positrons in vacancies preferentially annihilate with conduction electrons rather than core electrons. Hence, the (LMCF, HMCF) points in the LMCF–HMCF plot move from the lower right to the upper left with decreasing vacancy concentration. In this alloy, as shown in Table 1, the element of the greatest concentration is Al. Therefore, the (LMCF, HMCF) point moves on a straight line to the point representing Al with decreasing vacancy concentration. Deviations from a straight line in the LMCF–HMCF plot show the change in chemical environment around the vacancies detected by the positrons.

The Vickers hardness was measured at room temperature using an Akashi MVK-G3 model with a 1 kg load.

3. Results and Discussions

Figures 1(a) and 1(b) show the isothermal aging behavior of Vickers hardness and mean positron lifetime, respectively, for aging at both 50 and 100°C in quenched samples. The mean positron lifetime in the as-quenched samples, as shown in Fig. 1(b), is 227 ps, which is much larger than the positron lifetime in bulk pure Al (160 ps) and is similar to that in mono-vacancy in pure Al (230–245 ps).1518) It means that many quenched-in vacancies are present in the as-quenched sample. Vacancies in pure Al are mobile at around −50°C,18) and therefore they can only survive by binding with solute atoms, i.e., the quenched-in vacancies exist as vacancy-solute complexes in the as-quenched sample. The mean positron lifetime shows a two-step decrease with aging time at both 50 and 100°C. The trend obtained in this experiment is basically consistent with the results obtained by Banhart et al.9) The second-step decreases in the mean positron lifetime occur after 20 h of aging at 50°C and after 5 h of aging at 100°C, and seem to correspond to increases in the Vickers hardness for both 50 and 100°C. In samples that have been aged for such a long time, the vacancy-solute complexes have migrated and formed nano-clusters.8) According to Fig. 1(b), the mean positron lifetime at both 50 and 100°C is much longer than the positron lifetime in bulk Al (160 ps). Thus, the quenched-in vacancies are incorporated into the nano-clusters formed by the aging at both 50 and 100°C, i.e., the nano-clusters contain the vacancies.

Fig. 1

Change in (a) Vickers hardness and (b) mean positron lifetime for isothermal aging at 50 (black circles) and 100°C (red circles) in the quenched sample. As-Q denotes as-quenched.

Figure 2(a) shows (LMCF, HMCF) points of the CDB spectra for samples having undergone isothermal aging at 50 and 100°C. The (LMCF, HMCF) points for well-annealed pure Al, Mg, and Si are also plotted as references. Figure 2(b) shows an enlarged view of the (LMCF, HMCF) points of the CDB spectra. In the as-quenched state, the (LMCF, HMCF) points sit near the pure Si point rather than the pure Mg point. Thus, the positrons seem to detect mainly vacancy-Si complexes rather than vacancy-Mg complexes.

Fig. 2

(a) Aging behavior of low and high momentum component fractions (LMCF, HMCF) points of the CDB spectra for aging at 50 (black circles) and 100°C (red circles). The (LMCF, HMCF) points for well-annealed pure Al, Mg, and Si are also plotted as reference points. (b) Enlarged view of the (LMCF, HMCF) plot.

For aging at 50°C, the (LMCF, HMCF) point moves toward the Mg point for aging up to 20 h and then moves toward the Al point after 20 h. In the initial stage of aging, the positrons detect a sequence of vacancy-Si complexes migrating and forming Si-rich Al–Mg–Si-nano-clusters. Subsequently, more Mg and Si are added to the nano-clusters with aging time, corresponding to an increase in the nano-cluster size observed by 3D-AP.8) The chemical composition around the vacancies in the nano-clusters during the size increase changes so as to increase the Mg/Si chemical composition ratio for aging up to 20 h. During the change in the Mg/Si chemical composition ratio, the mean positron lifetime decreases, as shown in Fig. 1(b), which corresponds to the first step of the two-step decrease. This first step of the two-step decrease in the mean positron lifetime is not due to a decrease in the vacancy concentration but rather the change in the chemical composition around vacancies, because the (LMCF, HMCF) point does not move toward the Al point in this aging range. The chemical compositions around vacancies reach a certain value of the Mg/Si ratio at around 20 h of aging. After aging for 20 h, the (LMCF, HMCF) point moves toward the pure Al point, indicating that the vacancy concentration decreases with aging time after 20 h. This is consistent with the decrease in the mean positron lifetime in the second step of the two-step decrease. This decrease in the vacancy concentration also corresponds to an increase in the Vickers hardening. Therefore, the vacancies are dissociated from the nano-clusters after aging for 20 h and the dissociated vacancies increase the size of the nano-clusters by diffusion of solute atoms.

For aging at 100°C, the (LMCF, HMCF) point moves toward a point between that of pure Mg and pure Al for aging up to 10 min. This means that the chemical composition of the nano-clusters is changed so as to increase the Mg/Si ratio with decreasing vacancy concentration, which contrasts with the case for aging at 50°C, for which the vacancy concentration decreases after a composition change. Therefore, the decrease in the mean positron lifetime in the first-step of the two-step decrease for aging at 100°C is larger than that for aging at 50°C because the vacancy concentration decreases in the initial stage of the aging at 100°C. After aging for 5 h, the (LMCF, HMCF) point moves toward the pure Al point, indicating that the concentration of vacancies with almost fixed chemical composition decreases with aging time. This is consistent with the large decrease in the mean positron lifetime and the increase in the Vickers hardening after aging for 5 h at 100°C.

The (LMCF, HMCF) point for aging at 50 and 100°C lies on almost the same line after aging for several hours, showing that the chemical composition around the vacancies in the nano-clusters formed for aging at 50 and 100°C is almost the same.

The aging time dependence of the chemical composition around the quenched-in vacancies at both 50 and 100°C is Si-rich initially, exhibits an increase in the Mg/Si chemical composition ratio with increasing size of the nano-clusters, and converges to a certain Mg/Si value. This trend is consistent with previously reported 3D-AP results.7,19) The 3D-AP technique observes the nano-clusters and the positron annihilation technique observes vacancies and their chemical environment. This means that nano-clusters contain vacancies. The reason that the chemical composition around the quenched-in vacancies is Si-rich initially is probably due to that the formation energy of vacancy-Si-rich solute complexes is lower than that of vacancy-Mg-rich solute complexes in the case that number of solute atoms around the vacancies is small.20,21) The reason that the chemical composition around the quenched-in vacancies converges to a certain Mg/Si value is probably due to that the vacancy-solute complexes have stable configuration of solute atoms around vacancy. Most stable configuration have certain Mg/Si value in the case that the number of solute atoms around vacancy is large. Therefore, when the number of solute atoms around vacancy is increased with the aging time, the chemical composition around the vacancies is changed from the initial Si-rich to the most stable one. Quantitative estimation of the Mg/Si value from the CDB results in not straight forward and is difficult to estimate. The momentum distribution in high-momentum region in Al (Z = 13), Mg (Z = 12), and Si (Z = 14) corresponding to the momentum distribution of core electrons, where Z is atomic number, is not so different among these elements with slightly different atomic numbers even if the momentum distribution of core electrons is characteristic in each chemical element. Therefore, it is difficult to estimate Mg/Si values quantitatively from the shape of the momentum distribution in high-momentum region. Moreover, positron annihilation fraction with Si and Mg around vacancy depends on the lattice relaxation of Si and Mg around vacancy, which is also obstacle to quantitative estimation of the Mg/Si value. In this CDB result, the chemical composition around vacancies with long aging time was observed to be almost the same for both 50 and 100°C aging. However, in the initial stage, difference of the chemical composition around vacancies was observed, i.e., vacancy-Si-rich solute complexes was formed at 50°C, while the formation of vacancy-Si-rich solute complexes was avoided at 100°C. As pointed out in Ref. 19), it is expected that the typical Si-rich clusters are hard to transform continuously into the β′′ phases during the artificial aging, i.e., the typical Si-rich clusters can lead to the retardation of the hardness increase during the artificial aging. The vacancy-Si-rich solute complexes formed in the initial aging at 50°C observed by CDB may correspond to the Si-rich clusters observed by 3D-AP. Therefore, the avoidance of the formation of vacancy-Si-rich solute complexes by pre-aging around 100°C before storage at room temperature may be a key to avoid the negative effect of the artificial aging.

4. Conclusion

The vacancy behavior in excess Si type Al–Mg–Si-alloys aging at 50 and 100°C was investigated by positron annihilation spectroscopy. Many quenched-in vacancies exist in the as-quenched state and form vacancy-solute complexes. In the sequence from the formation of vacancy-solute complexes to the generation of nano-clusters, the vacancies become incorporated in the nano-clusters. The chemical composition around the quenched-in vacancies is Si-rich in the initial stage. The fraction of Mg around vacancies increases with aging time and the Mg/Si ratio around vacancies reaches a certain value for long time aging. The final Mg/Si composition ratio around the vacancies is almost the same for aging at both 50 and 100°C. However, the formation of the nano-clusters containing vacancies with Si-rich which are observed in the initial stage of the aging at 50°C is suppressed for the aging at 100°C. The suppression of the formation of the Si-rich nano-clusters may be important for hardness increase at artificial aging because the Si-rich nano-clusters are hard to transform into the β′′ phases during the artificial aging.

Acknowledgments

We would like to thank Mrs. Y. Xaio for her support during the experiments.

REFERENCES
 
© 2019 The Japan Institute of Light Metals
feedback
Top