Optically stimulated luminescence (OSL) dating is used to measure the period from the present to a mineral's last exposure to sunlight (Aitken, 1998). In order to obtain an OSL age, the equivalent dose (
Notably, it was reported that the IRSL signal stimulated at a high-temperature following a preceding low-temperature IR stimulation was much more stable (Thomsen
A multiple-elevated-temperature post-IR IRSL (MET-pIRIR) dating protocol has been established with five or six steps of IR stimulation from 50°C to 250°C or 300°C, with an interval of 50°C (Li and Li, 2011b; 2012b). The anomalous fading test indicated that the fading rates of the MET-pIRIR signals at 250°C or 300°C were negligible (Li and Li, 2011b; 2012b). In order to extend the maximum dating limit, a protocol was proposed based on the dose-dependent sensitivity of MET-pIRIR signals using the multiple-aliquot regenerative-dose (MAR) protocol (Li
In the present study, K-feldspar samples from the Jingbian site on the northern margin of the Chinese Loess Plateau were dated using five pIRIR protocols: four SAR (pIR50IR290, pIR200IR290, MET-pIRIR250, MET-pIRIR300) protocols, and the ‘MAR with heat’ protocol. Subsequently, the
Samples were collected from two loess sections (A and B) at the Jingbian loess site, which is located on the northern margin of the Chinese Loess Plateau, adjacent to the Mu Us Desert (
Sample pretreatment was performed in a dark room under subdued red light. Sample material from the two ends of the tubes was removed, and then the inner part was wet sieved to separate the >63 μm fraction, which was then successively treated with 10% HCl and 30% H2O2 solutions to remove carbonates and organic matter, respectively. The samples were then rinsed several times with water, oven-dried at 50°C overnight, and then dry sieved. The 90–125 μm fraction was used to separate K-feldspar grains, using a sodium poly-tungstate solution (ρ <2.58 g/cm3). K-feldspar grains were immersed in 10% HF solution, with magnetic stirring, for ∼15min, in order to remove the outer layer exposed to alpha irradiation. According to Duval
The five post-IR IRSL dating protocols used in this study. The first four are the SAR protocols with pIR50IR290, pIR200IR290, MET-pIRIR250 and MET-pIRIR300 signals, respectively. The fifth is the MAR protocol with the MET-pIRIR300 signal and a ‘cutheat to 500°C’ treatment added before the test dose, simplified as ‘MAR with heat’ (modified from Li et al. (2013)).
1 | ||||||
2 | Preheat at 320°C, 60 s | Preheat at 300°C or 320°C, 60 s | Preheat at 320°C, 60 s | |||
3 | IRSL 200 s at 50/200°C | IRSL 100 s at 50°C | IRSL 100 s at 50°C | |||
4 | IRSL 200 s at 290°C | IRSL 100 s at 100°C | IRSL 100 s at 100°C | |||
5 | IRSL 100 s at 150°C | IRSL 100 s at 150°C | ||||
6 | IRSL 100 s at 200°C | IRSL 100 s at 200°C | ||||
7 | IRSL 100 s at 250°C | Lx (250) | IRSL 100 s at 250°C | |||
8 | IRSL 100 s at 300°C or not | Lx (300) | IRSL 100 s at 300°C | Lx (300) | ||
9 | ||||||
10 | ||||||
11 | Preheat at 320°C, 60 s | Preheat at 300°C or 320°C, 60 s | Preheat at 320°C, 60 s | |||
12 | IRSL 200 s at 50/200°C | IRSL 100 s at 50°C | IRSL 100 s at 50°C | |||
13 | IRSL 200 s at 290°C | IRSL 100 s at 100°C | IRSL 100 s at 100°C | |||
14 | IR at 325°C, 200 s | IRSL 100 s at 150°C | IRSL 100 s at 150°C | |||
15 | Return to step 1 | IRSL 100 s at 200°C | IRSL 100 s at 200°C | |||
16 | IRSL 100 s at 250°C | Tx (250) | IRSL 100 s at 250°C | |||
17 | IRSL 100 s at 300°C or not | Tx (300) | IRSL 100 s at 300°C | Tx (300) | ||
IR at 320 or 325°C, 100 s | ||||||
Return to step 1 |
In addition, the low-frequency magnetic susceptibility (MS) of dried samples was measured using a Bartington Instruments MS2 meter. Measurements were repeated three times, and the mean values were normalised by the sample weights to obtain the mass-specific MS (10−8 m3/kg).
Standard growth curves (SGCs) were constructed to save measurement time. SGC constructions for the SAR protocols applied the least-squares normalisation (LS-normalization) method implemented with the R package ‘numOSL’ (Li
For the “MAR with heat” protocol, samples from both the Jingbian section and the Luochuan section were used to construct the MAR SGC. Test doses were fixed at 100 Gy. Aliquots were first bleached using the ORIEL solar simulator (Newport Corporation, model 94042A) with a 1000W xenon arc lamp for more than 8 hr. Experiments showed that the residual doses of the MET-pIRIR300 signals are reduced to ∼10 Gy after bleaching for 8 hr (
Environmental dose rates were also estimated. Thick source alpha counting was performed to quantify the contribution of U and Th to the dose rate, and the K content of the whole rock was measured by X-ray fluorescence (XRF) analysis. Conversion factors were adopted from Adamiec and Aitken (1998). The internal K and Rb concentrations of the K-feldspar were assumed to be 13 ± 1% and 400 ± 100 ppm, respectively (Huntley and Baril, 1997; Zhao and Li, 2005). The dose rate of cosmic rays was calculated according to the sampling altitude, depth, and geomagnetic latitude (Prescott and Hutton, 1994). For the Jingbian section, the water content was generally set as 15 ± 5% for samples in paleosol layers, and 10 ± 5% for samples in loess layers (Stevens
The
De values obtained using the five protocols. In the ‘MAR with heat’ protocol, the first ‘n’ indicates the aliquots used to measure the natural signal, and the second refers to the aliquots used to measure the signal of a regenerative dose.
A-570 | 570 | 373 ± 35 | 7 | 424 ± 11 | 19 | 415 ± 9 | 15 | \ | \ | 381 ± 11 | 12;12 |
A-1050 | 1050 | 615 ± 54 | 7 | 706 ± 34 | 8 | 706 ± 17 | 8 | \ | \ | 747 ± 53 | 16;8 |
B-0 | 220 | 543 ± 17 | 6 | 673 ± 57 | 4 | 640 ± 13 | 8 | 712 ± 57 | 6 | 856 ± 73 | 24;14 |
B-610 | 830 | 652 ± 20 | 10 | 734 ± 21 | 14 | 753 ± 17 | 9 | 773 ± 45 | 6 | 925 ± 62 | 24;20 |
B-640 | 860 | 613 ± 51 | 6 | 741 ± 18 | 9 | 773 ± 22 | 7 | 749 ± 57 | 6 | 901 ± 51 | 14;10 |
B-1000 | 1190 | 649 ± 61 | 6 | 822 ± 35 | 5 | 762 ± 25 | 7 | 800 ± 47 | 6 | 897 ± 71 | 24;20 |
Dose rates and ages of the samples using the five protocols.
A-570 | 15 ± 5 | 2.03 | 9.55 ± 0.15 | 3.21 ± 0.12 | 116 ± 12 | 132 ± 6 | 129 ± 6 | \ | 119 ± 6 |
A-1050 | 15 ± 5 | 2.19 | 10.83 ± 0.16 | 3.44 ± 0.13 | 179 ± 17 | 206 ± 13 | 206 ± 9 | \ | 217 ± 17 |
B-0 | 15 ± 5 | 1.95 | 10.14 ± 0.18 | 3.30 ± 0.12 | 165 ± 8 | 204 ± 19 | 194 ± 8 | 216 ± 19 | 260 ± 24 |
B-610 | 19.3 ± 5 | 2.50 | 11.88 ± 0.19 | 3.55 ± 0.13 | 183 ± 9 | 206 ± 10 | 212 ± 9 | 217 ± 15 | 260 ± 20 |
B-640 | 19.3 ± 5 | 2.40 | 11.85 ± 0.19 | 3.59 ± 0.13 | 171 ± 16 | 206 ± 9 | 215 ± 10 | 209 ± 18 | 251 ± 17 |
B-1000 | 15 ± 5 | 2.22 | 10.74 ± 0.18 | 3.43 ± 0.13 | 189 ± 19 | 240 ± 14 | 222 ± 11 | 233 ± 16 | 261 ± 23 |
In section B, the
It was reported that the pIR50IR290 signal was not applicable for dating samples with
Previous studies have shown that the pattern of sensitivity changes of K-feldspar in the first cycle of the SAR protocol may be dissimilar to the pattern of sensitivity change in the subsequent cycles and that the failure of sensitivity correction of the first cycle would result in the underestimation of
With the SAR protocol, the growth curve of K-feldspar has
We can provide no credible hypothesis to explain the SAR
As shown in
The SAR ages of the four samples from section B range from 160–190 ka with the pIR50IR290 signal, and 190–240 ka with the pIR200IR290, MET-pIRIR250, MET-pIRIR300 signals, which correspond to loess layer L2 and paleosol layer S2, respectively. The ages obtained using the ‘MAR with heat’ protocol are systematically 30–40 ka older than the SAR ages with the same MET-pIRIR300 signal. Considering the errors of ∼20 ka for the MAR ages, it is uncertain whether the samples are located in S2 layer or L3 layer.
By comparing
Compared to SAR protocols, the ‘MAR with heat’ protocol can provide the most reliable