High spatial resolution quantitative elemental imaging of foraminifer by laser ablation-inductively coupled plasma-mass spectrometry

Yuqiu Ke ab, Jianzong Zhou a, Lei Qiao a, Muhui Zhang a, Wei Guo a, Lanlan Jin a and Shenghong Hu *a
aState Key Laboratory of Biogeology and Environmental Geology, School of Earth Sciences, China University of Geosciences, Wuhan 430074, China. E-mail: shhu@cug.edu.cn
bSchool of Metallurgy and Chemical Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, China

Received 6th December 2018 , Accepted 4th March 2019

First published on 6th March 2019


Quantitative determination of the concentrations of elements and systematical characterization of their distribution in foraminifer is of vital significance. Quantitative elemental imaging by LA-ICP-MS is a useful alternative method; however, a methodology for high spatial resolution elemental imaging of foraminifer has not been obtained. In this work, a laser ablation spot size of 16 μm and a line scan speed of 8 μm s−1 were selected for elemental imaging after optimization; then, a two-point calibration strategy (TPCS) was established by combining NIST SRM 610 and NIST SRM 612 glasses with MACS-3 as external standards. The concentrations of Mg and Sr in four carbonate reference materials obtained by TPCS were close to the reference values with relative errors less than 10%. TPCS can avoid incorrect calibration caused by inhomogeneously distributed internal standards (e.g., 43Ca) between foraminifer shells and holes. A methodology for quantitative LA-ICP-MS elemental imaging of foraminifer was then developed, and high spatial resolution elemental images of Mg, Sr, and Ba were obtained. The spatial resolution of these images was calculated to be 16 × 0.40 μm per pixel. Elemental imaging of the Mg/Ca, Sr/Ca, and Ba/Ca ratios of a second foraminifer further confirmed the reproducibility of the elemental imaging methodology. The Mg/Ca ratio and the calcification temperature were found to gradually increase from the inner chambers (f-0) to the final chamber (f-1), while Sr was distributed more homogeneously and Ba showed little uptake in foraminifer shells. All these results demonstrate that this elemental imaging methodology is applicable to providing visual evidence to distinguish the elemental distributional differences in foraminifer.


Introduction

Foraminifer have many calcified chambers, which are sequentially added to existing chambers and grow gradually due to continuous calcification. During their adult life-cycle stages, they may migrate through the water column. For example, some species live in warmer surface water and then migrate into deeper, colder water to reproduce.1,2 During the calcification process, Mg is usually co-incorporated into foraminifer shells; this process has been proved be thermodynamically controlled and temperature-dependent (e.g., Mg/Ca = 0.167 × exp0.121×T).3–5 It can be concluded that the variability of seawater temperature occurring with foraminifer habitat migration during their adult life-cycle stages causes the Mg/Ca ratio to vary among distinct chambers.1,6,7 Therefore, the concentrations of Mg and its congeners Sr and Ba as well as their ratios to Ca in foraminifer are utilized as proxies for past seawater temperature and past climate changes and are widely used to reconstruct palaeoenvironmental conditions.3,8–12 In addition, it is also advisable to employ foraminifer proxies to monitor current changes in the polar oceans and atmosphere.13 For these reasons, quantitative determination of trace element concentrations and systematical characterization of their distribution in foraminifer are becoming increasingly essential for substantial understanding of trace elements as palaeoenvironmental proxies.

Bulk analysis was previously carried out by ICP-MS and other techniques to determine the element concentrations and element/Ca ratios;14–16 however, this technique cannot be used to distinguish the intratest variability of trace elements (i.e., distributional differences within an individual foraminifer). Meanwhile, a high spatial resolution imaging methodology based on quantitative microanalysis of trace elements in foraminifer would be particularly useful in this regard, because it enables us to quantitatively determine the element concentrations in each foraminifer chamber, to distinguish their distributional differences, and to provide visual elemental distributional images.

Many elemental imaging techniques have been reported previously, such as secondary ion mass spectrometry (SIMS),17,18 laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS),19–23 electron microprobe analysis (EMPA),24 and various X-ray fluorescence spectroscopy (XRF, e.g., micro synchrotron XRF).25–28 Among these imaging techniques, EMPA has been successfully used to characterize the contaminant phases and element/Ca ratios in foraminifer;24 however, a full view of the foraminifer elemental distributional images was not provided to compare the distributional differences among distinct chambers. The LA-ICP-MS elemental imaging methodology has a number of advantages, such as the lowest limits of detection at the ng g−1 range and high spatial resolution at a low μm scale;29,30 it has been widely used for elemental imaging of geological samples, such as ferromanganese nodules and fish otoliths.23,31 However, a quantitative LA-ICP-MS elemental imaging methodology has not been developed for foraminifer to distinguish distributional differences in trace elements and element/Ca ratios.

The aim of this paper was to develop a methodology for quantitative LA-ICP-MS elemental imaging of foraminifer. For this purpose, two critical LA-ICP-MS operating conditions, spot size and line scan speed, were systematically investigated, a quantification method combining NIST SRM 610 and NIST SRM 612 with MACS-3 as external standards was established, and a two-volume ablation cell was employed to improve the spatial resolution of the elemental images.30 The spatial resolution of the elemental images and the reproducibility of the elemental imaging methodology were also studied to ensure that this methodology has potential application in providing visual evidence for reconstructing palaeoenvironmental conditions.

Experimental

Instrumentation

All LA-ICP-MS experiments were performed on an Agilent 7700x series ICP-MS (Agilent Technologies, USA) coupled with a 193 nm ArF excimer LA system (GeoLas 2005, Lambda Physik, Göttingen, Germany) owned by the State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences (Wuhan). High purity helium (99.999%) was chosen as the carrier gas to increase the transport efficiency and enhance the signal sensitivity.32,33 The LA-ICP-MS was routinely tuned daily using NIST SRM 610 to obtain the maximum signal intensities of 25Mg+, 88Sr+, and 137Ba+, to maintain the 238U+/232Th+ ratio close to 1, and to ensure low oxide formation. Low oxide production was assured by an m/z 248/232 ratio (representing 232Th16O+/232Th+) that was consistently less than 0.5%. The isotopes 25Mg, 43Ca, 88Sr, and 137Ba were recorded during ablation with dwell times of 10 ms. A two-volume ablation cell was employed and a high carrier gas flow rate of 1000 mL min−1 was adopted to obtain a short washout time (∼0.5 s).30 Spot size and scan speed are two significant parameters that affect the sensitivity and spatial resolution of elemental images; therefore, they were optimized carefully. The typical operation parameters for LA-ICP-MS elemental imaging of foraminifer are given in Table 1.
Table 1 LA-ICP-MS operating conditions used in this work
Laser ablation parameters
Wavelength 193 nm, ArF excimer laser
Pulse duration 15 ns
Energy density 6 J cm−2
Repetition rate 10 Hz
Spot size 5, 10, 16, 24, 32, 44, 60, 90 μm
Scan speed 5, 8, 16, 24, 32 μm s−1
Carrier gas (He) flow rate 1.0 L min−1
ICP-MS conditions
Type Agilent 7700x
RF power 1500 W
Auxiliary gas (Ar) flow rate 0.8 L min−1
Plasma gas (Ar) flow rate 15 L min−1
Dwell time per isotope 10 ms
Sampling depth 6.0 mm
Detector mode Dual
Measured isotopes 25Mg, 43Ca, 88Sr, 137Ba


To determine the concentration of Ca, which was used as an internal standard in the LA-ICP-MS analysis, the concentrations of CaO were detected at the State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences (Wuhan), using a JEOL JXA-8100 Electron Probe Micro Analyzer equipped with four wavelength-dispersive spectrometers (WDS). The samples were first coated with a thin conductive carbon film prior to analysis.34 During the analysis, an accelerating voltage of 15 kV, a beam current of 5 nA and a 10 μm spot size were used to minimize sample damage and decrease the X-ray intensity fluctuations.35 Data were corrected on-line using a ZAF correction procedure. The peak counting time was 10 s for Na, Mg, Al, Si, K, Ca, Ba, Sr, and P. The background counting time was one-half of the peak counting time on the high- and low-energy background positions. The standards used for quantification included sanidine (K), pyrope garnet (Al), diopsode (Ca, Mg), jadeite (Na), olivine (Si), barite (Ba), celestite (Sr), and apatite (P).

Sample preparation

Two benthic foraminifer were handpicked from sediments that were collected from Ledong County, Hainan Province, China; they were disaggregated after leaching for 2 days in 10% hydrogen peroxide solution (Sinopharm Chemical Reagent Co., Ltd., China). The foraminifer were all cleaned by ultrasonication three times in methanol (Tianjin Tianli Chemical Reagent Co., Ltd., China) and repeatedly rinsed in 18.2 MΩ Milli-Q water (Labconco, Kansas, USA). After drying at room temperature, each foraminifer was mounted with epoxy resin (BUEHLER, EpoxiCure® Epoxy Resin, USA) and polished stepwise with sandpapers of 2000, 5000, and 7000 mesh, respectively, using a polishing machine (BUEHLER, USA) to obtain a flat surface for LA-ICP-MS imaging analysis. The first foraminifer was used to develop the methodology for elemental imaging by LA-ICP-MS, and the second was used to investigate the reproducibility of the established elemental imaging methodology.

Four Chinese carbonate powder reference materials (GBW07129, GBW07130, GBW07132, and GBW07135) were used to evaluate the accuracy of the two quantification methods. Each reference material was further milled in an agate mortar to obtain a grain size smaller than 400 mesh (∼38 μm) and pressed into one pellet at a pressure of 40 MPa using a powder press machine (BP-1, Dandong North Scientific Instrument Co., Ltd., China), respectively.36 Then, 4 pellets in total were mounted together into one sample target in epoxy resin (BUEHLER, EpoxiCure® Epoxy Resin, USA) and polished using the same procedure described above to obtain a flat surface for LA-ICP-MS analysis.

Quantification

A widely used quantification method is external calibration with internal standardization (ECIS).37–41 Because foraminifer have a matrix of calcium carbonate, a matrix-matched reference material, MACS-3, was used for external calibration. The unknown element concentration (Cisam) can be calculated using this equation:
 
Cisam = CISsam × (Cistd/CISstd) × (Iisam/IISsam)/(Iistd/IISstd)(1)
where Cisam and Cistd are the concentrations of element i in the sample and external standard, respectively; Iisam and Iistd are the signal intensities of element i in the sample and external standard, respectively; CISsam and CISstd are the concentrations of the internal standard (IS, i.e., 43Ca in this work) in the sample and external standard, respectively; and IISsam and IISstd are the signal intensities of the IS in the sample and external standard, respectively.

In addition to ECIS, a two-point calibration strategy (TPCS) was studied comparatively in this work for quantitative determination of the element concentrations in foraminifer. The equation for determining an unknown element concentration (Cisam) set out by Longerich et al.42 is:

 
Cisam = Iisam/S(2)
where S is the normalized sensitivity of element i, which was modified to incorporate the two reference materials NIST SRM 610 and NIST SRM 612 and can be calculated according to the following equations:
 
S = (k1 × Ii610/Ci610 + k2 × Ii612/Ci612)/2(3)
 
k1 = (IISMACS/CISMACS)/(IIS610/CIS610)(4)
 
k2 = (IISMACS/CISMACS)/(IIS612/CIS612)(5)
where k1 and k2 are coefficients that were used to calibrate the relative sensitivities of NIST SRM 610 and NIST SRM 612, respectively. These were calculated using 43Ca as an “internal standard”. Ci610 and Ci612 are the concentrations of element i in NIST SRM 610 and 612, respectively; Ii610 and Ii612 are the signal intensities of element i in NIST SRM 610 and 612, respectively; CIS610, CIS612, and CISMACS are the respective concentrations of the IS in NIST SRM 610, NIST SRM 612, and MACS-3; and IIS610, IIS612, and IISMACS are the signal intensities of the IS in NIST SRM 610, NIST SRM 612, and MACS-3, respectively. Herein, MACS-3 and NIST SRM 610 and 612 were analyzed before the whole run to predetermine the sensitivity (S) of each element. The reference concentrations in NIST SRM 610 and NIST SRM 612 can be found in previous work reported by Pearce et al.,43 and those in MACS-3 can be found in a paper published by Jochum et al.44

LA-ICP-MS analyses (n = 10) in single spot mode were conducted on each carbonate reference material. A spot size of 60 μm and a laser pulse number of 200 were used here. Other LA-ICP-MS operation parameters are listed in Table 1. The analytical protocol was similar to that published previously.45 In brief, the total acquisition time for each spot analysis was 60 s, consisting of 20 s for gas background with the laser off, followed by 20 s for signal acquisition and 20 s for the gas background again. Off-line selection and integration of the background and analyte signals, as well as time drift correction and quantitative calibration, were performed using an in-house program called ICPDataCal.46 Then, the concentrations of Mg, Sr, and Ba were comparatively calculated by ECIS (43Ca was used as the internal standard) and TPCS, respectively, to investigate the features of these two quantification methods.

LA-ICP-MS analysis and elemental imaging protocol

The foraminifer was placed in the inner cell of the two-volume ablation cell and the calibration standards, NIST SRM 610, NIST SRM 612, and MACS-3, were placed in the outer cell. All the sample surfaces were cleaned with absolute ethanol (Tianjin Tianli Chemical Reagent Co., Ltd., China) before analysis.

To determine the concentrations and lateral distributions of Ca in the foraminifer, 10 random analyses were conducted on each foraminifer by electron microprobe analysis (EMPA). The results show that the first foraminifer contained CaO (54.76 ± 0.87 wt%) and the second one contained CaO (55.02 ± 0.85 wt%). This indicates that Ca is distributed homogeneously in the foraminifer with a low concentration variation (RSD < 2%). From this perspective, Ca can be employed as an internal standard in this work.

The relative sensitivity S in eqn (2) should be predetermined at the beginning of the imaging experiment. NIST SRM 610, NIST SRM 612, and MACS-3 were analyzed respectively with 5 random spots under the optimized operating conditions as listed in Table 1.

It is often too time-consuming to perform LA-ICP-MS imaging of element concentrations in single spot mode; meanwhile, using line scans has many advantages compared to a series of points for imaging, as summarized by Ulrich et al.47 Therefore, laser analysis along a number of parallel, adjoining, and equally long lines was adopted for the LA-ICP-MS imaging. After each line scan, a single spot analysis was performed on the “drift correction” standard NIST SRM 610 to correct time-related sensitivity drift using a standard-sample bracketing (SSB) method.

Each line scan produced a (.csv) file, which was reduced using Microsoft Excel 2007 and assembled into a 2D signal matrix. In the matrix, the X coordinate in the elemental distributional images was defined as the scan line number multiplied by the spot size (i.e., 16 μm after optimization), and the Y coordinate was calculated by multiplying the recorded time by the scan speed (i.e., 8 μm s−1 after optimization). The 2D signal matrix of the first foraminifer consisted of 105 scan lines, and that of the second foraminifer consisted of 90 scan lines.

After subtraction from the gas blank and correction of time-related sensitivity drift, the 2D signal matrix was converted into a 2D concentration matrix using our developed quantification method and imported into the imaging software Surfer (Ver. 12.0.626) for elemental imaging analysis. Other software, such as ICPDataCal46 and OriginPro (Version 8.0724), were used in this work to calculate the integrated signal intensities and plot the figures, respectively.

Results and discussion

Optimization of operating conditions

To perform external calibration in single spot mode and to obtain high quality elemental distributional images of foraminifer using LA-ICP-MS, careful optimization of the operation parameters is required. As described above, two crucial parameters, spot size and scan speed, that influence the sensitivity and spatial resolution of images were optimized. NIST SRM 610 was used here for optimization due to the homogeneous distribution of the trace elements.

Sensitivity and spatial resolution in LA-ICP-MS elemental imaging experiments primarily depends on the spot size of the laser. Eight different spot sizes, 5, 10, 16, 24, 32, 44, 60, and 90 μm, were used in single spot analysis mode (200 pulses), and the Mg (Fig. 1a) and Sr (Fig. 1b) signals were recorded for data evaluation. As shown in Fig. 1, unstable signals were recorded for spot sizes of 5 and 10 μm. Under these two conditions, both Mg and Sr showed obvious decreases in signal intensity during the 20 s ablation process.


image file: c8ay02664e-f1.tif
Fig. 1 Influence of spot size on quantification: (a) Mg; (b) Sr.

As described in the literature, the ablation rate was high, up to ∼200 nm per pulse, during the first 100 pulses and gradually decreased later, which led to a decrease of the signal intensity in the LA-ICP-MS analysis.48 This phenomenon was particularly severe when a small spot size was employed and was another source of elemental fractionation.49 For example, Mank and Mason50 reported that elemental fractionation becomes significant when the crater depth/diameter ratio is larger than 6, and an effective approach to decreasing the depth/diameter ratio is to increase the spot size. The depth/diameter ratio was roughly estimated to be 8 (an ablation rate of ∼200 nm and 200 pulses were used for estimation here) when a spot size of 5 μm was employed. This demonstrated that severe elemental fractionation occurred under these conditions; this also can be observed in Fig. 1, where the signal intensities of both Mg and Sr dramatically decreased to the background level within 5 s. When lager spot sizes (≥10 μm) were employed, smaller depth/diameter ratios (<6) could be obtained. However, elemental fractionation was still found with a spot size of 10 μm on the basis of the elemental fractionation index (EFI, 0.6 to 0.8 as displayed in Table 2), which was calculated according to the equation published by Fryer et al.51 When a spot size of 16 μm was adopted, the EFI was >0.96, which indicates that the elemental fractionation was negligible. Although larger spot sizes (≥24 μm) can lead to milder elemental fractionation (EFI > 0.99), smaller spot sizes are known to be beneficial to improve the spatial resolution of elemental images; therefore, a compromised spot size, 16 μm, was ultimately employed in the subsequent experiments.

Table 2 EFI values obtained with different spot sizes
Spot size (μm) 5 10 16 24 32 44 60 90
Mg 0.485 0.737 0.971 0.994 0.992 0.993 1.000 1.007
Sr 0.325 0.607 0.960 0.990 0.992 0.990 0.999 1.000


The second crucial parameter for image quality is the scan speed employed at a certain spot size (16 μm). For each line scan, the total line scan distance and cycle time (dwell time × monitored isotope number) remain unchanged. A higher scan speed means less scan time and a lower data number (equal to scan time/cycle time), fewer pixels, and lower spatial resolution. A lower scan speed can ensure a higher data number and enough pixels; however, it always requires too long of a scan time for line scan and imaging analysis. Thus, the scan speed should also be carefully optimized. Because foraminifer have many holes in their bodies, in order to precisely distinguish the elemental distributional differences between the holes and shells in foraminifer, several ablation holes were produced in single spot analysis mode (500 pulses) along a straight line in NIST SRM 610. To simulate different sizes of foraminifer chamber holes, different spot sizes (10, 16, 24, 32, 44, 60, and 90 μm) were adopted to produce these ablation holes. Five repeated line scans along the straight line from left to right under 5 different scan speeds (32, 24, 16, 8, 5 μm s−1) were performed and the signal variations of Mg and Sr were recorded, as shown in Fig. 2. For larger ablation holes (≥24 μm), an obvious decrease in the signal intensity can be observed at any scan speed; meanwhile, for smaller ablation holes (10, 16 μm), only ablation at relatively low speeds (8 and 5 μm s−1) can distinguish the signal variations among the ablation holes. At a scan speed of 5 μm s−1, the signal intensity fell and rose inacutely because much time was required to ablate the margins between NIST SRM 610 and the ablation hole, which may cause image blurring effects.52 Furthermore, a lower scan speed requires a longer scan time for imaging analysis; compared to 5 μm s−1, about 40% time can be saved for elemental imaging of one sample at a scan speed of 8 μm s−1. For these reasons, an optimal scan speed of 8 μm s−1 was employed for the elemental imaging analysis.


image file: c8ay02664e-f2.tif
Fig. 2 Scheme of the line scans (a) and influence of the line scan speed (μm s−1): (b) 32; (c) 24; (d) 16; (e) 8; (f) 5. The numbers in red represent the spot size.

Quantification

Given the definition of a suitable internal standard, one important requirement is that the internal standard element is homogeneously distributed in both the samples and reference materials.53–55 For foraminifer, although Ca is distributed homogeneously (RSD < 2%) in foraminifer shells according to EMPA results, the Ca distribution between the shells and holes is obviously inhomogeneous.

When performing LA-ICP-MS external calibration with internal standardization (ECIS), another equation can be obtained according to eqn (1):

 
Cisam = Iisam × (IISstd/CISstd)/(Iistd/Cistd)/(IISsam/CISsam)(6)

If the internal standard is definitely distributed homogeneously in the external standard once a suitable reference material is selected, (IISstd/CISstd)/(Iistd/Cistd) must be constant except for slight variations due to random error. Here, we focus on the effects of Ca distribution (IISsam/CISsam) in foraminifer on the accuracy of this quantification method (ECIS). In order to simplify eqn (6), (IISstd/CISstd)/(Iistd/Cistd) was defined as a constant A:

 
A = (IISstd/CISstd)/(Iistd/Cistd)(7)

Therefore, eqn (6) can be simplified as follows:

 
Cisam = Iisam × A/(IISsam/CISsam)(8)

In theory, a matrix element is distributed homogeneously in the samples and therefore can be widely used as an internal standard; however, this is not always true. For example, Ca is distributed inhomogeneously between the shells and holes. Based on this, if ECIS is used for LA-ICP-MS calibration, incorrect results will be obtained, as deduced below:

 
Cisam-a = Iisam-a × A/(IISsam-a/CISsam-a)(9)
 
Cisam-b = Iisam-b × A/(IISsam-b/CISsam-b)(10)
where Cisam-a and Cisam-b are the concentrations of element i in different laser ablation locations a and b in a sample; Iisam-a and Iisam-b are the signal intensities of element i in locations a and b; CISsam-a and CISsam-b are the concentrations of the internal standard in locations a and b; and IISsam-a and IISsam-b are the signal intensities of IS in locations a and b.

It is supposed that the true concentration of Cisam-a is equal to Cisam-b, and CISsam-a is higher than CISsam-b. Thus, it is not difficult to see that Iisam-a equals Iisam-b and IISsam-a is higher than IISsam-b according to their relative sensitivity coefficients. However, CISsam-a is constrained to be equal to CISsam-b when performing internal standardization due to the definition of the internal standard. (IISsam-a/CISsam-a) thus becomes higher than (IISsam-b/CISsam-b), and an incorrect result where Cisam-a is lower than Cisam-b is obtained. Furthermore, if Cisam-a is truly higher than Cisam-b, an inverse result where Cisam-a is lower than Cisam-b will probably be obtained according to the above deductions. For example, the concentration of Mg in foraminifer shells is definitely higher than that in the holes; however, we obtained an inverse result where the concentration of Mg in the shells was calculated to be lower than that in the holes, which was confirmed by quantitative imaging of the foraminifer, as described below. In summary, the quantification method with ECIS is not suitable for quantitative imaging of foraminifer because Ca is distributed inhomogeneously between the shells and holes.

Therefore, an alternative quantification approach, TPCS, was developed in this work. To evaluate the accuracy of this quantification method, four carbonate reference materials, GBW07129, GBW07130, GBW07132, and GBW07135, were treated as unknown samples, and 10 random analyses on each sample were conducted. The concentrations of Mg, Sr, and Ba in these samples were comparatively calculated by ECIS and TPCS, as shown in Fig. 3. It can be seen that the results of Mg and Sr obtained by TPCS are similar to those obtained by ECIS and are close to the reference values with relative errors less than 10%. With regard to Ba, larger errors (>15%) were found in GBW07129, GBW07130 and GBW07132 by both ECIS and TPCS. This can be ascribed to the low concentrations of Ba in GBW07129 (8 μg g−1) and GBW07130 (4.9 μg g−1) and the extremely high concentration of Ba in GBW07132 (1330 μg g−1), which may be beyond the linear range because NIST SR612, NIST SRM610, and MACS-3 have Ba contents of 39.7, 435, and 58.7 μg g−1, respectively. Because Ba is not obviously uptaken in foraminifer shells (as discussed below) and the more important elements, Mg and Sr, could be accurately determined, TPCS was considered to be reasonable.


image file: c8ay02664e-f3.tif
Fig. 3 Comparison of reference values and measured values (n = 10) by both ECIS and TPCS of Mg (a), Sr (b), and Ba (c). Error bars: 1 SD.

More importantly, for ECIS, although Ca is distributed homogeneously in the shells (Fig. 4b), it is undoubtedly distributed inhomogeneously between the shells and holes; this led to inverse results in the quantification and elemental imaging analysis. As shown in Fig. 4c and d, in theory, low concentrations of Mg (Fig. 4c) and Sr (Fig. 4d) will be detected for foraminifer holes and epoxy resin, and high concentrations will be detected for the shells. Actually, high concentrations of Mg and Sr were found in the holes and epoxy resin and Mg and Sr were hardly detected in the shells. This incorrect result can be avoided by TPCS because it does not involve calculations in terms of element/Ca ratios. Using TPCS, the expected element images were obtained, as shown in Fig. 4e and g. A similar TPCS process (employing powder otolith reference material FEBS-1) was successfully used by McGowan et al.31 to perform elemental imaging analysis of fish otoliths, which further reveals the potential of the TPCS developed in this work for quantitative determination of element concentrations in foraminifer by LA-ICP-MS. In summary, TPCS was adopted for quantification in the quantitative elemental imaging of foraminifer.


image file: c8ay02664e-f4.tif
Fig. 4 Micrograph of the first foraminifer (a), Ca signal intensity distribution (b), elemental distributional images of Mg/Ca (c) and Sr/Ca (d) quantified by ECIS, and elemental distributional images of Mg (e), Sr (f), and Ba (g) quantified by TPCS.

Quantitative elemental imaging of a foraminifer

Quantitative elemental distributional images based on 105 line scans on the first foraminifer by LA-ICP-MS were produced for Mg, Sr, and Ba, as shown in Fig. 4e–g. A micrograph of the foraminifer obtained in reflected light mode (Fig. 4a) is also shown to compare the elemental distributional images with the real sample morphology.

4400 data were recorded for each line scan, and a total of 105 line scans were carried out. The spatial resolution of these elemental distributional images was 105 × 4400 pixels in an area of 1680 × 1765 μm2; thus, the pixel size was 16 × 0.40 μm. This high spatial resolution of the elemental images provided significant visual evidence for investigating the elemental distributional behaviors in the foraminifer.

Reproducibility of the elemental imaging methodology

As significant proxies for studying past seawater temperatures and past climate changes, Mg/Ca, Sr/Ca, and Ba/Ca were calculated and plotted into 2D distributional images based on 90 line scans on the second foraminifer, as shown in Fig. 5c–e. Mg is usually co-incorporated into foraminifer shells in a thermodynamically controlled and temperature-dependent process (e.g., Mg/Ca = 0.167 × exp0.121×T); in these equations, the Mg to Ca ratio is expressed as a molar ratio (mmol mol−1) of the molar element to Ca. These were calculated according to their mass ratios and the corresponding relative atomic masses, as shown in Fig. 5c–e.
image file: c8ay02664e-f5.tif
Fig. 5 Micrographs of the second foraminifer using reflected light (a) and transmitted light (b) and distributional images of Mg/Ca (c), Sr/Ca (d), Ba/Ca (e), and calcification temperature (f).

As shown, high spatial resolution elemental images were also obtained for the second foraminifer (90 × 3760 pixels in an area of 1424 × 1504 μm2). This successfully confirmed the reproducibility of the elemental imaging methodology. Additionally, micrographs of the foraminifer obtained in reflected light mode (Fig. 5a) and in transmitted light mode (Fig. 5b) are shown to compare the elemental distributional images with the real sample morphology. Labels f-1 to f-15 in Fig. 5a indicate the chamber calcification order, counting back from the final chamber (f-1). Rather than representing individual chambers as f-1 to f-15, f-0 represents the inner chambers that could not be distinguished from each other. Almost homogeneous distribution of Sr (Fig. 5d) among these chamber shells was found, which can be attributed to isomorphism substitution of Ca by Sr in the shells; meanwhile, Mg (Fig. 5c) had interesting elemental distributional behaviour, where the lowest Mg/Ca ratio was found in the inner chambers f-0 and then gradually increased from f-15 to f-1, except for f-13 and f-2. This was also confirmed through statistical analysis of the average ratios of Mg/Ca and Sr/Ca in each chamber (shell), which were calculated based on 200 acquired data points, as shown in Fig. 6. The exceptionally high Mg/Ca ratio in chamber f-13 can be attributed to contamination from the final chamber f-1; however, the reason for the exceptional Mg/Ca ratio in chamber f-2 is not clear. However, it did not significantly influence the variation trend where the Mg/Ca ratio gradually increased from the inner chambers (f-0) to the final chamber (f-1). Completely different from Mg and Sr, Ba was concentrated either in the gaps between the shells or in the final chamber due to contamination from the surrounding environment (e.g. interstitial water).56 In other words, little Ba was uptaken, and low Ba/Ca ratios were detected in the foraminifer shells.


image file: c8ay02664e-f6.tif
Fig. 6 Mg/Ca and Sr/Ca ratios in foraminifer chambers, showing that the Mg/Ca ratio gradually increases from the inner chambers (f-0) to the final chamber (f-1).

Because the incorporation of Mg into foraminifer calcite is temperature-dependent, the Mg/Ca distributional image (Fig. 5c) was transformed into a calcification temperature image according to the equation developed by de Nooijer et al. (i.e., Mg/Ca = 0.167 × exp0.121×T).3 As shown in Fig. 5f, the calcification temperature generally increased from the inner chambers f-0 to the final chamber f-1, which provides visual evidence for evaluating the calcification temperature and has potential applications in the construction of palaeoenvironmental conditions. Overall, these results demonstrate that our developed high spatial resolution quantitative LA-ICP-MS elemental imaging methodology can be used to distinguish differences in the element/Ca ratio distributions and calcification temperatures in foraminifer.

Conclusions

Heterogeneously distributed trace elements in foraminifer are important proxies for studying palaeoenvironmental conditions. LA-ICP-MS can be used to quantitatively determine element concentrations and systematically characterize their distributions in foraminifer; however, a methodology for high spatial resolution elemental imaging of foraminifer has not been obtained. In this work, a laser ablation spot size of 16 μm and line scan speed of 8 μm s−1 were selected for high spatial elemental imaging after optimization because sufficient sensitivity and negligible elemental fractionation (EFI: 0.971 for Mg and 0.960 for Sr) were found and because it can distinguish signal variations in the range of 10 μm under these operating conditions. Then, a two-point calibration strategy (TPCS) was established by combining NIST SRM 610 and NIST SRM 612 glasses with MACS-3 as external standards. The concentrations of Mg and Sr in four carbonate reference materials obtained by TPCS were close to the reference values with relative errors less than 10%, which is comparable to the widely used external calibration with internal standardization (ECIS). More importantly, TPCS can avoid incorrect calibration caused by inhomogeneously distributed internal standards (e.g., 43Ca). On this basis, a methodology for quantitative LA-ICP-MS elemental imaging of foraminifer was developed and high spatial resolution elemental images of Mg, Sr, and Ba were obtained. The spatial resolution of these images was calculated to be 16 × 0.40 μm per pixel, which verifies the high spatial resolution of the elemental images. Furthermore, elemental imaging of Mg/Ca, Sr/Ca, and Ba/Ca for a second foraminifer confirmed the reproducibility of this elemental imaging methodology. For the second foraminifer, the Mg/Ca ratio and the calcification temperature were found to gradually increase from the inner chambers (f-0) to the final chamber (f-1), while Sr was distributed more homogeneously and Ba showed little uptake in foraminifer shells. All these results demonstrate that this elemental imaging methodology is applicable to provide visual evidence to distinguish the elemental distributional differences in foraminifer.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 41603025), the National Key Research and Development Program of China (No. 2017YFD0800304), and the China Postdoctoral Science Foundation (No. 2016M602386). We sincerely thank the editor and anonymous reviewers for their constructive comments and suggestions on this article.

References

  1. S. Eggins, P. De Deckker and J. Marshall, Earth Planet. Sci. Lett., 2003, 212, 291–306 CrossRef CAS.
  2. C. Hemleben, Modern Planktonic Foraminifera, Springer, 1989, p. 363 Search PubMed.
  3. L. J. de Nooijer, E. C. Hathorne, G. J. Reichart, G. Langer and J. Bijma, Mar. Micropaleontol., 2014, 107, 32–43 CrossRef.
  4. K. Billups and D. P. Schrag, Earth Planet. Sci. Lett., 2003, 209, 181–195 CrossRef CAS.
  5. D. W. Lea, T. A. Mashiotta and H. J. Spero, Geochim. Cosmochim. Acta, 1999, 63, 2369–2379 CrossRef CAS.
  6. P. Anand and H. Elderfield, Geochem. Geophys. Geosyst., 2005, 6, 1–15 CrossRef.
  7. L. Jonkers, B. Buse, G. J. A. Brummer and I. R. Hall, Earth Planet. Sci. Lett., 2016, 451, 177–184 CrossRef CAS.
  8. C. H. Lear, H. Elderfield and P. A. Wilson, Science, 2000, 287, 269–272 CrossRef CAS PubMed.
  9. J. B. Creech, J. A. Baker, C. J. Hollis, H. E. G. Morgans and E. G. C. Smith, Earth Planet. Sci. Lett., 2010, 299, 483–495 CrossRef CAS.
  10. K. A. Allen, B. Honisch, S. M. Eggins and Y. Rosenthal, Earth Planet. Sci. Lett., 2012, 351, 270–280 CrossRef.
  11. J. Yu, H. Elderfield, Z. Jin, P. Tomascak and E. J. Rohling, Quat. Sci. Rev., 2014, 98, 1–6 CrossRef.
  12. N. Glock, V. Liebetrau, A. Eisenhauer and A. Rocholl, Chem. Geol., 2016, 447, 40–53 CrossRef CAS.
  13. R. Schiebel, R. F. Spielhagen, J. Gamier, J. Hagemann, H. Howa, A. Jentzen, A. Martinez-Garcia, J. Meilland, E. Michel, J. Repschlager, I. Salter, M. Yamasaki and G. Haug, Mar. Micropaleontol., 2017, 136, 1–13 CrossRef.
  14. J. M. Yu, J. Day, M. Greaves and H. Elderfield, Geochem. Geophys. Geosyst., 2005, 6, 1–9 CrossRef.
  15. L. Lo, C.-C. Shen, C.-J. Lu, Y.-C. Chen, C.-C. Chang, K.-Y. Wei, D. Qu and M. K. Gagan, J. Asian Earth Sci., 2014, 81, 115–122 CrossRef.
  16. P. Chen, J. Yu and Z. Jin, Geochem. Geophys. Geosyst., 2017, 18, 617–630 CrossRef CAS.
  17. N. Winograd, Anal. Chem., 2015, 87, 328–333 CrossRef CAS PubMed.
  18. J. S. Becker, U. Breuer, H. F. Hsieh, T. Osterholt, U. Kumtabtim, B. Wu, A. Matusch, J. A. Caruso and Z. Y. Qin, Anal. Chem., 2010, 82, 9528–9533 CrossRef CAS PubMed.
  19. A. Balbekova, M. Bonta, S. Torok, J. Ofner, B. Dome, A. Limbeck and B. Lendl, Anal. Methods, 2017, 9, 5464–5471 RSC.
  20. A.-C. Niehoff, J. Gruenebaum, A. Moosmann, D. Mulac, J. Soebbing, R. Niehaus, R. Buchholz, S. Kroeger, A. Wiehe, S. Wagner, M. Sperling, H. von Briesen, K. Langer and U. Karst, Anal. Chim. Acta, 2016, 938, 106–113 CrossRef CAS PubMed.
  21. A. Gundlach-Graham, M. Burger, S. Allner, G. Schwarz, H. A. O. Wang, L. Gyr, D. Grolimund, B. Hattendorf and D. Günther, Anal. Chem., 2015, 87, 8250–8258 CrossRef CAS PubMed.
  22. M. Burger, A. Gundlach-Graham, S. Allner, G. Schwarz, H. A. O. Wang, L. Gyr, S. Burgener, B. Hattendorf, D. Grolimund and D. Günther, Anal. Chem., 2015, 87, 8259–8267 CrossRef CAS PubMed.
  23. J. Hirata, K. Takahashi, Y. V. Sahoo and M. Tanaka, Chem. Geol., 2016, 427, 65–72 CrossRef CAS.
  24. L. D. Pena, I. Cacho, E. Calvo, C. Pelejero, S. Eggins and A. Sadekov, Geochem. Geophys. Geosyst., 2008, 9, 1–12 CrossRef.
  25. J. Garrevoet, B. Vekemans, P. Tack, B. De Samber, S. Schmitz, F. E. Brenker, G. Falkenberg and L. Vincze, Anal. Chem., 2014, 86, 11826–11832 CrossRef CAS PubMed.
  26. D. S. Gholap, A. Izmer, B. De Samber, J. T. van Elteren, V. S. Selih, R. Evens, K. De Schamphelaere, C. Janssen, L. Balcaen, I. Lindemann, L. Vincze and F. Vanhaecke, Anal. Chim. Acta, 2010, 664, 19–26 CrossRef CAS PubMed.
  27. Z. Qin, J. A. Caruso, B. Lai, A. Matusch and J. S. Becker, Metallomics, 2011, 3, 28–37 RSC.
  28. L. Gomez-Nubla, J. Aramendia, S. F. O. de Vallejuelo, A. Alonso-Olazabal, K. Castro, M. C. Zuluaga, L. A. Ortega, X. Murelaga and J. M. Madariaga, Anal. Bioanal. Chem., 2017, 409, 3597–3610 CrossRef CAS PubMed.
  29. S. M. Eggins, L. P. J. Kinsley and J. M. G. Shelley, Appl. Surf. Sci., 1998, 127, 278–286 CrossRef.
  30. Y. Liu, Z. Hu, H. Yuan, S. Hu and H. Cheng, J. Anal. At. Spectrom., 2007, 22, 582–585 RSC.
  31. N. McGowan, A. M. Fowler, K. Parkinson, D. P. Bishop, K. Ganio, P. A. Doble, D. J. Booth and D. J. Hare, Sci. Total Environ., 2014, 494, 177–186 CrossRef PubMed.
  32. D. Günther and C. A. Heinrich, J. Anal. At. Spectrom., 1999, 14, 1369–1374 RSC.
  33. I. Horn and D. Günther, Appl. Surf. Sci., 2003, 207, 144–157 CrossRef CAS.
  34. R. X. Zhang and S. Y. Yang, Microsc. Microanal., 2016, 22, 1374–1380 CrossRef CAS PubMed.
  35. X. Zhang, S. Y. Yang, H. Zhao, S. Y. Jiang, R. X. Zhang and J. Xie, J. Earth Sci., 2019 DOI:10.1017/S143192761601182X.
  36. Y. Ke, W. Guo, L. Jin, L. Qiao, H. Yang and S. Hu, Chem. Res. Chin. Univ., 2017, 33, 360–364 CrossRef CAS.
  37. J. Cizdziel, K. X. Bu and P. Nowinski, Anal. Methods, 2012, 4, 564–569 RSC.
  38. J. S. Crain and D. L. Gallimore, J. Anal. At. Spectrom., 1992, 7, 605–610 RSC.
  39. T. Vaculovic, K. Breiter, Z. Korbelova, N. Venclova, K. Tomkova, S. Jonasova and V. Kanicky, Microchem. J., 2017, 133, 200–207 CrossRef CAS.
  40. Y. Q. Ke, Y. J. Sun, P. J. Lin, J. Z. Zhou, Z. F. Xu, C. F. Cao, Y. Yang and S. H. Hu, Microchem. J., 2019, 145, 642–647 CrossRef CAS.
  41. D. J. Hare, J. Lear, D. Bishop, A. Beavis and P. A. Doble, Anal. Methods, 2013, 5, 1915–1921 RSC.
  42. H. P. Longerich, S. E. Jackson and D. Günther, J. Anal. At. Spectrom., 1996, 11, 899–904 RSC.
  43. N. J. G. Pearce, W. T. Perkins, J. A. Westgate, M. P. Gorton, S. E. Jackson, C. R. Neal and S. P. Chenery, Geostand. Newsl., 1997, 21, 115–144 CrossRef CAS.
  44. K. P. Jochum, B. Stoll, U. Weis, D. E. Jacob, R. Mertz-Kraus and M. O. Andreae, Geostand. Geoanal. Res., 2014, 38, 265–292 CrossRef CAS.
  45. Y. T. Li, W. Guo, Z. C. Hu, L. L. Jin, S. H. Hu and Q. H. Guo, J. Agric. Food Chem., 2019, 67, 935–942 CrossRef CAS.
  46. Y. Liu, Z. Hu, S. Gao, D. Günther, J. Xu, C. Gao and H. Chen, Chem. Geol., 2008, 257, 34–43 CrossRef CAS.
  47. T. Ulrich, B. S. Kamber, P. J. Jugo and D. K. Tinkham, Can. Mineral., 2009, 47, 1001–1012 CrossRef CAS.
  48. M. Hola, J. Ondracek, H. Novakova, M. Vojtisek-Lom, R. Hadravova and V. Kanicky, Spectrochim. Acta, Part B, 2018, 148, 193–204 CrossRef CAS.
  49. M. Guillong and D. Günther, J. Anal. At. Spectrom., 2002, 17, 831–837 RSC.
  50. A. J. G. Mank and P. R. D. Mason, J. Anal. At. Spectrom., 1999, 14, 1143–1153 RSC.
  51. B. J. Fryer, S. E. Jackson and H. P. Longerich, Can. Mineral., 1995, 33, 303–312 CAS.
  52. M. Bonta, H. Lohninger, V. Laszlo, B. Hegedus and A. Limbeck, J. Anal. At. Spectrom., 2014, 29, 2159–2167 RSC.
  53. Y. Q. Ke, X. F. Yao, S. H. Hu, W. Guo, Q. H. Hu, Z. L. Zhu and Z. C. Hu, Anal. Lett., 2015, 48, 830–842 CrossRef CAS.
  54. J. O'Reilly, D. Douglas, J. Braybrook, P. W. So, E. Vergucht, J. Garrevoet, B. Vekemans, L. Vincze and H. Goenaga-Infante, J. Anal. At. Spectrom., 2014, 29, 1378–1384 RSC.
  55. D. A. Frick and D. Günther, J. Anal. At. Spectrom., 2012, 27, 1294–1303 RSC.
  56. L. Vetter, H. J. Spero, A. D. Russell and J. S. Fehrenbacher, Geochem. Geophys. Geosyst., 2013, 14, 2916–2931 CrossRef.

This journal is © The Royal Society of Chemistry 2019