Following up on previous attempts to date diatom frustules, further investigations were made on 1) extracting diatom frustules devoid of inorganic luminescent grains, 2) developing an equivalent dose estimating protocol based on the diatomite luminescence characterization, and 3) testing the applicability of this protocol on two lacustrine profiles.
Radiation Measurements 156 (2022) 106803 Contents lists available at ScienceDirect Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas Further investigations towards luminescence dating of diatoms P Morthekai a, *, P Tiwari a, M.K Murari b, P Singh a, c, B Thakur a, M.C Manoj a, S.N Ali a, V.K Singh a, K Kumar a, J Rai a, N Dubey d, P Srivastava e a Birbal Sahni Institute of Palaeosciences, 53 University Road, Lucknow, India National Geochronology Facilities, Inter-University Accelerator Centre (IUAC), New Delhi, India c Department of Geology, Banaras Hindu University, Varanasi, India d Addis Ababa Science and Technology University, Addis Ababa, Ethiopia e Department of Geology, University of Lucknow, Lucknow, India b A R T I C L E I N F O A B S T R A C T Keywords: Diatoms Frustules Luminescence dating Diatomite Vembanad wetland Mahanadhi Following up on previous attempts to date diatom frustules, further investigations were made on 1) extracting diatom frustules devoid of inorganic luminescent grains, 2) developing an equivalent dose estimating protocol based on the diatomite luminescence characterization, and 3) testing the applicability of this protocol on two lacustrine profiles Diatom frustules were extracted in such a way that they are almost devoid of non-biogenic polymineral grains, confirmed by field emission scanning electron microscope (FE-SEM) observation and en ergy dispersive X-ray (EDX) analysis The presence of opal was confirmed by X-ray diffraction (XRD) spectro metric analysis The optimized luminescence signal that could be used for equivalent dose estimation was blue stimulated UV emission with a preheat temperature of 200 ◦ C The thermoluminescence glow curve peaking at 245 ◦ C might be the source of this signal In this study the characteristic dose was found to be ~1500 Gy Two sediment profiles were explored for luminescence dating, fading rates and a-values were found different between profiles This discrepancy can be resolved 1) by measuring luminescence characteristics across different regions, or/and 2) by using species-specific luminescence measurements This attempt has yielded an encouraging set of luminescence ages, with diatom frustule ages comparable to fine grain polymineral ages Introduction of diatom frustules) experiences luminescence quenching (Shin et al., 1996) Moreover, the long-term stability is reduced due to relatively larger anomalous fading rate in amorphous materials compared to crystalline ones (Hayes et al., 2019; Rieser and Edsall, 2004) The second issue is the technical difficulty of extracting diatom frustules from inorganic luminescent minerals like quartz and feldspar Even though earlier studies suggested that the diatoms not have many useful dosimetric properties (Berger and Easterbrook, 1988; Hayes et al., 2019), other studies have shown that they produce luminescence signals (Cornett and Cornett, 2010; Hayes et al., 2019) Optically stim ulated luminescence (OSL; blue stimulated UV emission measured at 125 ◦ C after preheating to 260 ◦ C) was measured on the commercially available diatomaceous earth (deposits of exoskeleton material formed by the death of a large concentrated diatom population) yielded an unbleached dose of 180 Gy (Hayes et al., 2019) These researchers re ported a prominent TL peak at 300 ◦ C and a small peak at 120 ◦ C using a heating rate of ◦ C.s− In another attempt (Cornett and Cornett, 2010), arctic and alpine lake sediment diatoms were extracted using acid Dating siliceous biogenic materials such as diatom is advantageous in some geological settings where there are no other datable materials such as calcium carbonates or organic materials for the radiocarbon method (Anderson et al., 2002; Andrews et al., 1999; Sulpis et al., 2018), or little or unsuitable quartz/feldspar for luminescence method (Cabanes et al., 2011; Madella and Lancelotti, 2012; ODP, 2007; Tr´ eguer et al., 1995) There are attempts to date 1) sillafin which is a protein intrinsic to diatom frustules (Hatt´ e et al., 2008), 2) other organic compounds such as long-chain polyamines entrapped in diatom frustules during silicifi cation (Ingalls et al., 2004), 3) phytolith occluded carbon (Zuo and Lu, 2019) using the radiocarbon method The luminescence method was also attempted on diatom frustules (Berger and Easterbrook, 1988; Cornett and Cornett, 2010; Hayes et al., 2019; Rieser and Edsall, 2004) Previous attempts to date diatom frustules have been hampered by two main difficulties The first one is a conceptual issue based on an un founded suspicion that amorphous material like opaline silica (make up * Corresponding author E-mail address: morthekai@gmail.com (P Morthekai) https://doi.org/10.1016/j.radmeas.2022.106803 Received 17 December 2021; Received in revised form 21 May 2022; Accepted 24 May 2022 Available online June 2022 1350-4487/© 2022 Elsevier Ltd All rights reserved P Morthekai et al Radiation Measurements 156 (2022) 106803 digestions with aqua regia and hydrogen peroxide OSL was measured without preheating and at a lower stimulation temperature (30 ◦ C) in ultraviolet (UV) after blue stimulation In these earlier attempts, 1) no infra-red stimulated luminescence (IRSL) was observed, 2) OSL signals were brought down to 10% after 30 of natural sunlight exposure (in Ontario, Canada), and 3) two OSL components were fitted to the measured OSL decay curve (Cornett and Cornett, 2010) In another refined attempt, diatom species-specific luminescence dating was con ducted on, 1) ~35 ka old (26–55 ka) freshwater diatom ooze (Cyclo tellastelligera), and 2) ~120 ka (115–130 ka) old marine diatom ooze (Thalassiothrixlongissima), but the complete data is yet to be published as a full research paper (Rieser and Edsall, 2004) Learning from earlier attempts to date diatom frustules using the luminescence method, we further investigate the dating potential of diatom frustules by 1) refining the method to extract diatom frustules free of conventional abiotic luminescing minerals such as quartz and feldspar, 2) studying its luminescence characteristics to arrive at a protocol to estimate equivalent dose and, 3) applying the findings to date two sedimentary deposits This study will provide the methodo logical aspect of diatom frustule dating shown in Fig The first two stages entailed removing carbonates and organic materials from the sediments using N HCl and 30% H2O2 respectively Stage involved extracting the clay particles by treating them with 5% sodium hexametaphosphate (SHP) overnight The floating clay particles were pipetted out Stage involved cleaning the settled fraction with distilled water before storing it in sodium poly tungstate with a specific density of 2.3 g cm− Settling of grains was expedited by applying a gentle centrifugal force at 1500 rpm for 15 using a centrifuge tube of 50 ml The floating portion was pipetted out and cleaned with distilled water, and oven-dried This procedure is slightly a modified one of Morley et al (2004) After the extraction of diatom frustules, fine grain polymineral was separated from MN and MT samples by treating with 0.01 N sodium oxalate which deflocculates the finer grains and the associated coarser grains Using Stokes’ settling time difference, 4–11 μm grains were separated by allowing them to settle between 1.5 and 15 in the ethanol column These grains were mixed in acetone and allowed to deposit in cleaned scratched Al discs that were placed inside individual flat bottom glass vials of cm height (Morthekai and Ali, 2014) Diatom frustules were extracted and deposited in Al discs as mentioned above Materials and methods 2.2.2 Testing of diatom frustules free of non-biogenic luminescing materials The geochemical and mineralogical compositions of the extracted diatom frustules were tested using energy dispersive X-ray analysis (EDX; also known as energy dispersive analysis, EDS) and X-ray diffraction spectrometric analysis (XRD) respectively The EDX mea surements were carried out on the identified spots (EDS Spots) using a field emission scanning electron microscope (FE-SEM) A Schottky-type field emission (T-FE) gun was used as the electron gun probe It has a 1.5 nm beam width at an accelerating voltage of kV in GB mode It is a JOEL JSM 7610f model of Joel India Pvt Ltd Energy dispersive X-ray measurements were carried out at liquid nitrogen temperature and the EDX spectra were collected at a resolution of 127 eV The PANalytical Xpert’3 Powder with Cu as the anode material was used to measure the diffraction pattern of the samples (diatom frustules and inorganic polymineral fine grains of size 4–11 μm) deposited on the Al discs The diffraction pattern was measured from 2θ = 5◦ –80◦ with a time step of 18.870 s, at room temperature The measured diffraction spectra were analyzed using powdR, an R package (Butler and Hillier, 2020), and the rockjock_mixtures data set (Eberl, 2003) was used to quantify the mineral concentrations present in the samples Opal_282, Opal_264, Opal_253, Intermediate_Microcline, Orthoclase, and Albite_ Cleavelandite were considered as the possible minerals that are present in the sample, and corundum was considered as the standard mineral The presence of quartz grains was checked by comparing the shapes of the measured photon arrival time distribution of diatom frustules and quartz The photon arrival time distribution (PATD) of diatom frustules was measured using time-resolved luminescence (TRL) detecting system attached to Risoe TL/OSL Reader DA-20 (Lapp et al., 2009) The stim ulation was achieved by pulsed blue LEDs (470 ± 20 nm) using ON and OFF times of 10 μs and 30 μs respectively The total stimulation time was 200 s and hence million pulses with a pulse period of 40 μs were used The emitted photons were collected using PMT EMI 9835QA through a 7.5 mm Hoya U-340 filter 2.1 Sample details Two categories of samples were used in this study A natural diato mite deposit was collected from Lake Ashenge, Ethiopia and this is the first category The diatom frustules that were extracted from this sample was used to characterize the luminescence properties such as stability (both thermal and a-thermal), bleachability, and dose-response No age estimation was done for this sample The second category of samples was used to apply the luminescence dating method to the diatom extracts, and compare the ages with other independent ages There are two sets of samples in the second category One set of samples (MN series) was a revisit of radiocarbon-dated 200 cm deep sedimentary sequence from the margin of Mahanadi river near Chhuipali village, Bargarh district, Odisha (Tripathi et al., 2013) The fine-grain polymineral (abiotic sediment) and diatom frustules of this series are called MNS and MND respectively The sediment samples used in this study are recent depositions with a moderate to a high degree of mottling and are essentially unconsolidated to semi-consolidated, allowing the sediment to be found in chunks rather than loosely asso ciated (Fig S1) The sediment type is defined by sandy to silty loams and has a high humus content (Tripathi et al., 2013) Samples were collected not in opaque metallic pipes but as chunks at depths of 70 cm, 120 cm, and 180 cm from the top The age of diatom frustules are validated against the available conventional radiocarbon age The second set of samples (MT series) belongs to a 100 cm deep core retrieved from Mundro Thuruth, Ashtamudi Lake, Kollam, Kerala Similarly, the fine-grain polymineral (abiotic sediment) and diatom frustules of this series are called MTS and MTD respectively Four samples were collected at 20 cm, 36 cm, 50 cm, and 80 cm from the top of the core As there is no other independent age control for this series, luminescence-based diatom frustule ages are compared to fine-grain polymineral ages 2.2.3 Equivalent dose estimation The equivalent dose (De) of diatom frustules was measured using a SAR procedure that was adapted from that of quartz dating (Murray and Wintle, 2003) Preheat that was used in the SAR procedure is discussed in Section 3.2.7 The De of the fine grain polymineral sample was measured using the SAR procedure with preheat temperature of 200 ◦ C for a holding time of 60 s, before IRSL measurement (Banerjee et al., 2001; Blair et al., 2005) The measured IRSL was majorly contributed by feldspar grains Fading rate (g-value, %/decade) was measured using the SAR procedure (Auclair et al., 2003) and the fading correction was done using Huntley and Lamothe method (2001) 2.2 Methods and instrument details 2.2.1 Extraction of diatom frustules The MN samples were collected in chunks and the outer light exposed layer of a minimum of cm was removed The inner light un exposed portion was used further The MT samples were collected in PVC pipe vertically as a core The core was halved inside the dark lab oratory and sampled at four depths A mm thick sample that was in touch with PVC pipe was removed The diatom frustules from these light unexposed samples and the diatomite were extracted in four stages as P Morthekai et al Radiation Measurements 156 (2022) 106803 Fig Diatom frustules extraction procedure in four stages All the luminescence measurements were made using an automated Risø TL-OSL-DA-20 reader that was equipped with an EMI 9835QA photomultiplier tube, blue (470 ± 20 nm), IR (870 ± 40 nm), and violet (405 nm; violet laser + interference filter [AHF F39-404, mm] + glass filter [AHF GG 395–12.5]) stimulation sources (Bøtter-Jensen et al., 2010; Lapp et al., 2015) The emitted UV photons were detected either using a 7.5 mm Hoya U-340 (blue stimulation) or AHF F39-340 Bright-Line HC 340/26 mm filter (violet stimulation) Beta particle irradiations were carried out using an on-plate 90Sr/90Y beta source and it deliver a dose rate of 0.071 Gy s− significantly Alpha efficiency was calculated by comparing the lumi nescence induced by known beta and alpha dose Three required pa rameters to calculate a-value are equivalent beta dose, the flux of alpha particles, and alpha exposure time The formula that was used to calculate a-value is ‘a-value = [De,β/(13 x N x α)]’, where De,β is an equivalent beta dose (Gy), N is alpha particle flux from 241Am (#.min− mm− 2) in a vacuum, and α is alpha irradiation time (min) Results and discussion 3.1 Performance assessment of the diatom frustules extraction procedure 2.2.4 Dose rate estimation The concentrations of U, Th and K in the bulk sediment were measured using gamma spectrometry The samples were crushed, sealed in airtight plastic boxes, and stored to achieve secular equilibrium After a month of storage, concentrations of U, Th, and K were calculated by comparing the concentration of the aforesaid long-lived radioactive el ements with that of a standard NUSSY (Preusser and Kasper, 2001) Canberra-made reverse electrode coaxial Ge detector (REGe; GR-2018 model) with a 16,000 channels multichannel analyzer (DSA-LX) was used to detect the gamma-ray photons Background counts were reduced by keeping the detector at liquid nitrogen temperature and within the cm thick Perspex shield that is kept within a cm thick lead bricks chamber Uranium concentration was calculated from the arithmetic mean of concentration of its radioactive daughters 226Ra (186 keV), 214 Pb (295.2 keV), and 214Bi (609.3 keV, 1120.3 keV, and 1700 keV) Similarly, thorium concentration was calculated from its radioactive daughters 212Pb (238.6 keV), 228Ac (911.1 keV), and 208Tl (2614.5 keV) The photopeak at 1460 keV was used to calculate the concentration of potassium The concentrations of U, Th, and K were converted into dose rate (Adamiec and Aitken, 1998) after considering the water content, alpha efficiency (a-value), beta attenuation factors (Mejdahl, 1979), and cosmic ray dose (Prescott and Hutton, 1994) using online Dose Rate and Age Calculator (DRAC v1.2) (Durcan et al., 2015) As the diatom frustules are of different shapes with a hollow inside, essentially the ionizing radiation deposits their energy only in the thinner walls (5–50 μm) Since the dose deposition occurs only in the walls of diatom frustules whose thickness is comparable to the pene tration depth of alpha particles (~25 μm), alpha efficiency (a-value) becomes an important parameter that influences the dose rate The presence of non-biogenic luminescing minerals such as quartz or feldspar was examined on the diatom frustules mounted discs The samples were mounted and examined using FE-SEM The samples were dominated by different diatom assemblages (Fig a-d) There were a few occurrences of phytoliths and sponge spicules (Fig e, f and Fig S2 d, f) This could be because the samples have not been sieved at any stage (less than 10 μm) Generally, broken diatoms would be of the size less than 10 μm and the larger diatoms are bigger than 75 μm size EDX measurements suggest that oxygen was the most abundant element in diatomite (60–70%), phytolith (65–70%), and sponge spicules (70%), with Si accounting for the remaining 25–30% (Fig S2 b,d-h) These observations support the fact that the samples are predominantly of SiO2 XRD spectra were measured on the Al discs that had 1) diatom frustules from the diatomite sample, 2) diatom frustules from sediment (MND1), and 3) polymineral grains (4–11 μm) extracted from the sediment sample (MNS1) XRD was measured on the Al disc itself and it has prominent peaks at 2θ values of 38.52 ◦ C, 44.76 ◦ C, 65.14 ◦ C, and 78.26 ◦ C Background (Al disc spectra) subtracted XRD spectra of diat omite, diatom (MND1), and sediment (MNS1) are given with the posi tion of opal in the spectra (Fig 3a and S3) These XRD spectra were analyzed, and it was found that the concentration of opal was 76% in both diatomite and diatom, but only 55% in sediment Feldspar’s con centration was found to be 20% in both diatomite and diatom, and 42% in sediment The shape of photon arrival time distribution (PATD) of quartz and feldspar is different (Ankjærgaard et al., 2010) Blue light stimulated TRL (TR-OSL) was measured on the irradiated (35 Gy) and preheated P Morthekai et al Radiation Measurements 156 (2022) 106803 Fig SEM images of extracted diatom frustules from diatomite sample Along with diatom assemblages (a Epithermia, b Cyclotellaocellata, c Cyclotellameneghiana, d Navicula sp.), phytoliths (e) and sponge spicules (f) were also present in the samples Mix of benthic (a & d) and planktonic (b & c) diatom species were there (200 ◦ C for 10 s) diatomite sample with 10 μs as ON time, and 30 μs as OFF time for 200 s (Fig 3b) The shape of TRL of quartz exhibit a slow rise during the ON time and a slow decay since the LED is OFF with a decay constant of ~35 μs (Chithambo and Galloway, 2000) Although the observed TRL’s shape resembles to that of feldspar (Ankjærgaard et al., 2009), the significantly lower intensity of IRSL (Fig 4b) suggests TRL has arisen from opaline diatom frustules themselves, not from feldspar grains There was a mix of benthic and planktonic species from diatomite, MN series and MT series of samples Only the proportion was different For MN series, there was an 80% benthic diatoms, and MT series constituted 89% benthic diatoms characterized 3.2.1 Thermoluminescence glow curves The extracted diatom frustules from the diatomite were given ~3.5 Gy The Tmax -Tstop method was used to know whether a measured TL peak is single or continuous (McKeever, 1980) The temperature at which the TL (up to 400 ◦ C @ ◦ C/s) shows its first maximum (Tmax) was plotted against Tstop which is similar to preheat temperature that varied from 50 ◦ C to 350 ◦ C at an interval of 20 ◦ C (Fig 4a) Before every Tstop, the aliquot was given 3.5 Gy This plot shows that there is a quasi-continuous TL peak around 200 ◦ C that arises not from a single electron trap but a distribution of it 3.2.2 Prominent luminescence signal One aliquot of the diatomite sample was given a 350 Gy dose and preheated to 200 ◦ C IR stimulated signals were measured at a temper ature of 30 ◦ C (IRSL) for 40 s After IRSL, the sample was again pre heated to 200 ◦ C to avoid any IR photo-transferred signal, and blue 3.2 Dosimetric characteristics In this section, basic luminescence properties such as the shape of the TL glow curve, prominent optically stimulated luminescence signal, the source of luminescence, bleaching, and fading behavior are Fig a) Background (XRD spectra measured on Al disc) subtracted XRD spectra of diatomite, diatom frustules extracted from MN series samples, and its sediment counterpart Only the cropped (2θ from 15◦ to 35◦ ) spectra were shown for better visibility, although measurement was done from 5◦ to 80◦ Fitted spectra for quantification were shown in Fig S2 b) Photon arrival time distribution (PATD) plot derived from the time-resolved luminescence data P Morthekai et al Radiation Measurements 156 (2022) 106803 Fig Luminescence characteristics of extracted diatom frustules from diatomite sample The Tmax -Tstop plot was measured on an aliquot of irradiated (3.5 Gy) diatom frustule (a) An aliquot of diatom frustules was given 350 Gy and preheated to 200 ◦ C before the measurement of b) various luminescence signals (IR, Blue, and Violet stimulated luminescence), c) remaining TL signals, and d) OSL signals at different stimulation temperatures (the error is sqrt(N) where is N is photon counts) Bleachability (e) and fading (f) characteristics were also tested The error in the data point reflects the relative error from L/T that is converted to dose Note: The TL glow curve with No B-UV measurement in (c) was taken from the data of (a) before IR, Blue and Violet stimulated luminescence signals were measured That is why the remaining TL signals after B-UV @ 50 ◦ C are apparently higher stimulated signal (BSL) was measured at a temperature of 30 ◦ C for 40 s and preheated to 200 ◦ C Then violet light stimulated signal (VSL) was measured at a temperature of 30 ◦ C for 40 s Among these three signals, BSL was prominent (Fig 4b) 142 Gy and preheated to 200 ◦ C for 10 s A prompt OSL measurement was carried out using blue stimulation at 125 ◦ C for 40 s Another OSL measurement was made with a delay of 38 h between the end of beta irradiation and the measurement Visibly no fading occurred in 38 h since the cessation of irradiation (Fig 4f) Four aliquots of diatomite were used to measure g-value (%/decade) using the SAR procedure (Auclair et al., 2003) An average value of 1.5 ± 2.6% per decade was calculated, and that too suggests a non-fading luminescence signal in diatomite 3.2.3 Source of OSL signal TL was measured on the irradiated (350 Gy) and preheated (200 ◦ C) diatomite sample, and after OSL measurements at 50 ◦ C, 100 ◦ C, and 150 ◦ C (Fig 4c) The TL peak was observed at a temperature of ~245 ◦ C The OSL measurements were lowering that TL peak intensity Presence of TL signal even after measurement of OSL at high temperature (150 ◦ C) confirms hard to bleach signal So, the source of OSL signals from diatomite is the TL glow curve peaking at 245 ◦ C 3.2.7 Refined protocol for equivalent dose estimation The remaining OSL signal after progressive preheating of the irra diated diatomite aliquot to temperature from 70 ◦ C to 350 ◦ C at an in terval of 40 ◦ C was measured (Fig 5a) The OSL depletion rate was higher at the preheat temperature of 230 ◦ C Earlier studies measured OSL (blue light stimulation and UV photons detection) of diatom frus tules at 1) 30 ◦ C without preheating (Cornett and Cornett, 2010), and 2) 125 ◦ C after preheating to 220 ◦ C (Hayes et al., 2019) The second study used the SAR procedure which is similar to quartz OSL dating (Hayes et al., 2019; Murray and Roberts, 1998) We refined their protocol based on the luminescence characteristics that are studied above The preheat temperature was chosen to be 200 ◦ C, with a holding time of 30 s (Fig 5b), and this low holding time (30 s) was chosen arbitrarily A laboratory given dose of 361 Gy was recovered by the above protocol with 10 ± 4% over-estimation (399 ± 14 Gy) The character istic dose (D0) of 750 Gy was estimated using a single saturating expo nential to the constructed dose response curve (Fig 5c) This suggests that it is feasible to date diatom frustules that are typically 400 ka old (2 x D0 = 1500 Gy) using a dose rate ranging from to Gy.ka− As dose recovery suggests, the dose values of diatom frustules might have been overestimated by 10% 3.2.4 Optimizing the stimulation temperature The set of OSL signals measured at a temperature of 50 ◦ C, 100 ◦ C, and 150 ◦ C as discussed in the section above was used to optimize the stimulation temperature Compared to 50 ◦ C, the OSL signal measured at a temperature of 150 ◦ C was larger by 23% (Fig 4d) Higher stimulation temperature didn’t change the shape of the OSL decay curve but only the intensity (inset) This observation also implies that the substance being measured is not quartz Because in quartz OSL thermal quenching (Pagonis et al., 2010; Wintle, 1975) rather than thermal assistance, has been observed 3.2.5 Bleachability A single aliquot of diatomite was given a dose of 142 Gy and exposed to sunlight for