Figure 1

(Color online) The complex dielectric susceptibility $\chi$ (measured with an ac field of $3\mathrm{V}∕\mathrm{cm}$ rms at $50\mathrm{Hz}$ and a $T$-sweep rate of $3\mathrm{K}∕\min$) is plotted against temperature for PMN/PT (90/10). Cooling and heating curves are shown for the in-phase susceptibility.Reuse & Permissions

Figure 2

(Color online) Aging behavior of ${\chi }^{″}$ is shown for several $T_A$. Inset of panel (a) shows the temperature profile for a typical aging experiment. Curve 1 is taken on cooling from initial annealing at temperature $T_{AN}$ to aging temperature $T_A$ held there for aging time $t_A$. Curve 2 is taken on subsequent cooling to $T_{EX}$ with immediate reheating back to the $T_{AN}$ along curve 3. Curves 4 and 5 are reference cooling and heating curves (respectively). Measurements were taken on PMN/PT (90/10) [PMN/PT (88/12)] with an ac field of $2.9\mathrm{V}∕\mathrm{cm}$ rms at $50\mathrm{Hz}$ ($1\mathrm{V}∕\mathrm{cm}$ at $1\mathrm{kHz}$) with $T$-sweep rates of about $3\mathrm{K}∕\min$ [$2\mathrm{K}∕\min$, except (h), which was taken at about $4\mathrm{K}∕\min$]. For PMN/PT (90/10), $T_{AN}=(\mathrm{a},\mathrm{b})400\mathrm{K}$, (c) $340\mathrm{K}$, (d) $330\mathrm{K}$; $t_A=(\mathrm{a},\mathrm{c},\mathrm{d})16\mathrm{h}$, (b) $4\mathrm{h}$; $T_{EX}=\left(\mathrm{a}\right)240\mathrm{K}$, (b) $250\mathrm{K}$ (curve 3a), $240\mathrm{K}$ (curve 3b), (c,d) $130\mathrm{K}$. For PMN/PT (88/12) $T_{AN}=450\mathrm{K}$, $t_A=10\mathrm{h}$, and $T_{EX}=\left(\mathrm{e}\right)250\mathrm{K}$, (f) $220\mathrm{K}$, (g) $180\mathrm{K}$, (h) $130\mathrm{K}$. For clarity, damped oscillations from 302 to $308\mathrm{K}$ lasting about $0.5\mathrm{h}$ arising from the initial temperature overshoot have not been shown in (a). (b) shows aging curves for two sets of experiments with different $T_{EX}$’s. Curves 1 and 3 overlay over each other for the two experiments and only the heating curve 3a of the first experiment is shown for clarity. Curve 3a in (d) is a polynomial fit of curve 3 with the “hole” from 165 to $200\mathrm{K}$ excised and is used to separate the aging memory into cumulative and hole-like components (see text). The sample spent about $2\mathrm{h}$ around $140\mathrm{K}$ on cooling to $T_{EX}$ in (h) because of a slowing cooling rate, but with little evidence of a resulting aging hole. High-frequency noise in the aging curve has been removed by adjacent averaging for (h).Reuse & Permissions

Figure 3

(Color online) The ability to create successive aging holes in ${\chi }^{″}$ with independent memory is shown. PMN/PT (90/10) is aged at $175\mathrm{K}$ (1a) and subsequently $130\mathrm{K}$ (1b) for $8\mathrm{h}$ each, then cooled to $100\mathrm{K}$ (2) and immediately reheated (3). $T$-sweep rates were about $3\mathrm{K}∕\min$ and ${\chi }^{″}$ was measured with an ac field of $2.9\mathrm{V}∕\mathrm{cm}$ rms at $50\mathrm{Hz}$.Reuse & Permissions

Figure 4

(Color online) Dimensionless aging rate at ${10}^4\mathrm{s}$ for PMN/PT (90/10) (ac frequency $50\mathrm{Hz}$), PMN, PMN/PT(72/28) $\left(10\mathrm{Hz}\right)$, and PLZT $\left(40\mathrm{Hz}\right)$.Reuse & Permissions

Figure 5

(Color online) Representative ${\chi }^{″}\left(t_W\right)$ is shown for each of the aging regimes of PMN/PT (90/10) along with best fits to various functional forms (chosen depending on which give physically reasonable parameters or a smaller chi-squared). (a) ${\chi }^{″}$ at $300\mathrm{K}$ was fitted to a stretched exponential form [$\beta \left(\omega \right)$ ranged from 0.63 at $10\mathrm{Hz}$ to 0.50 at $1\mathrm{kHz}$]. (b) $280\mathrm{K}$ at $50\mathrm{Hz}$ fitted to power law $(\gamma =0.19)$, (c) $230\mathrm{K}$ fitted to logarithmic form, and (d) $150\mathrm{K}$ again to power-law form $(\gamma =0.14)$. (c,d) were measured with the same set of frequencies as (a), with $\chi \left(\omega \right)$ decreasing monotonically as a function of frequency [only for early $t_W$ in (c)]. Data for some frequencies have been omitted for clarity.Reuse & Permissions

Figure 6

(Color online) ${\chi }^{″}(\omega ,t_W)-{\chi }_o^{″}\left(\omega \right)$ is plotted against $\omega t_W$ for several frequencies at $300\mathrm{K}$ (right axis), $260\mathrm{K}$, and $150\mathrm{K}$ (left axis). Scaling is found in PMN/PT (90/10) for lower $T$.Reuse & Permissions

Figure 7

(Color online) Aging memory is shown as a function of initial aging time after a cooling cycle to $T_{EX}$ at sweep rates of $\pm 3\mathrm{K}∕\min$. For $T_A=175\mathrm{K}$ $\left(150\mathrm{K}\right)$, $T_{EX}=120\mathrm{K}$ $\left(105\mathrm{K}\right)$.Reuse & Permissions

Figure 8

(Color online) Aging memory after small thermal excursions of $\mathrm{\Delta }T$ about $T_A$ for PMN/PT (90/10) at 175 and $150\mathrm{K}$. Sweep rates were about $\pm 3\mathrm{K}∕\min$.Reuse & Permissions

Figure 9

(Color online) Aging memory after cooling cycles to $T_{EX}$ well below $T_A$ is shown. Memory has been separated into cumulative and holelike components [see Fig. 2d]. For $T_A$ of $175\mathrm{K}$ and above, $T_{EX}=130\mathrm{K}$. For $T_A=165\mathrm{K}$, $T_{EX}=115\mathrm{K}$. $T_{EX}$ is below $95\mathrm{K}$ for the lowest $T_A$.Reuse & Permissions

Figure 10

(Color online) Effect of a $1\mathrm{h}$ field application after aging in zero field for $6\mathrm{h}$ on ${\chi }^{″}$ is shown for PMN/PT (90/10).Reuse & Permissions

Figure 11

(Color online) A measure of the effect of a sudden application of an electric field (as in Fig. 10) on aging of ${\chi }^{″}$ is shown as a function of applied field. We define the amount of initial zero-field aging reduction of ${\chi }^{″}$ remaining $30\mathrm{s}$ after the dc field application as a measure of aging recovery.Reuse & Permissions