Abstract
The effect of time-gating method in Raman spectroscopy for fluorescence suppression of Iranian region calcite is investigated. Experiments are done using an Nd:YAG laser with a pulse durations of 10 ns at wavelength 532 nm. Seven samples from different places are examined. In order to obtain the optimum gate width for fluorescence suppression, a series of experiments is carried out at different gate widths. Raman-to-fluorescence (R/F) and fluorescence-to-laser peak (F/L) ratios are compared at gated and nongated experiments. Applying the optimum gate width leads to an effective reduction of fluorescence background and improvement in both ratios of R/F and F/L. Raman signals of some samples in nongated experiments are completely hidden by fluorescence while emerged in gated experiments.
1. Introduction
Raman spectroscopy is a robust method which informs about chemical structure of different types of samples. The accuracy and nondestructive features of this method make it applicable into wide fields, including mineralogy [1] and planetary surfaces [2–4]. One of the applications of Raman spectroscopy is molecular and structural analysis of mineral samples. Calcium carbonates are famous components in mineralogy with three common polymorphs including calcite, vaterite, and aragonite. Calcite and vaterite have a hexagonal form and aragonite has an orthorhombic form [5]. Calcite is the most common form of carbonates, which can be found in dry and semidry areas and restrained sewages [6]. Although calcites are almost pure, some impurities such as Mg, Fe, and Mn may be found in those.
Limestone and marble are two rocks that are made of calcite and make up a large amount of earth’s surface. These components are important in construction materials [7, 8], construction aggregates, and pigments [9, 10].
Three different polymorphs of calcite can be readily distinguished by using vibrational spectroscopy including infrared and Raman spectroscopy, because of their different crystal forms. These techniques provide valuable information on the frequencies of atomic vibrations in the crystal, which in turn depend intimately on the structure and the symmetry of the crystal and make the fingerprint of a molecular structure.
One of the difficulties that decreases the distinguishability of Raman peaks is intense fluorescence background. This background provides significant problems in many subjects such as analysis of biological specimens in long exposure times and also study of impure samples. Different research groups tried to solve this problem by theoretical and experimental methods [11–17]. When a pulsed laser is used, Raman peaks appear simultaneously with laser pulse while fluorescence emission is spread in times after the laser pulse. Therefore, the time-gating method can eliminate long lifetime fluorescence [11]. In this case, using intensified charge coupled device (ICCD) and optical Kerr gating is common [14, 18]. On the other hand, Fourier-domain filtering [13] and polynomial fitting of the background [16] are two theoretical methods for suppression of fluorescence.
Different types of calcite can be found in Iran, due to its paleogeographical condition. Calcite products play an important role in economy of Iran and so, evaluation of the quality of different resources is prominent. In our knowledge, no previous studies have been performed on the time gating using pulsed lasers in carbonates of different Iranian regions. The main objective of this study is fluorescence rejection by using time-gating approach in Raman spectroscopy of calcites from different regions of Iran. Various gate times were evaluated and according to experiments, 23 ns gate width was selected as optimum. Results show that time-gating method is suitable for suppression of fluorescence.
2. Experiment
2.1. Samples
Seven samples collected from different places in Iran were used. These samples were selected by consultation with Department of Earth Science from places such as Vanarj Qum (Figure 1(a)), Semnan (Figure 1(b)), Alborz (Figure 1(c)), Mashhad (Figure 1(d)), Mahalat Arak (Figure 1(e)), Zagross Fars (Figure 1(f)), and Kerman (Figure 1(g)). These samples were selected without attention to color, quality, impurity, and crystal structure. Distribution and location of study places are shown in Figure 2.
2.2. Experimental Setup
Schematic of experimental setup is illustrated in Figure 3. The excitation source was the second harmonic of an Nd:YAG pulsed laser at 532 nm with full width at half maxima (FWHM) of 10 ns, and a repetition rate of 8 Hz. The laser pulses were reflected by a mirror (M) to illuminate the sample at distance of 2 meters from the laser. 70 mJ Laser pulses were illuminated an area of 50 mm2 on the sample without focusing. Because of variation in mass and volume of mineral samples, a flexible holder was used. The light emitted from the sample was collected by a lens with 3.5 cm focal length (LN) at a right angle to the incident beam. All samples were fixed in the focal length and hence, a parallel beam passed through a 532 nm notch filter (NF) to eliminate Reyleigh scattering. Then the scattered light was focused by a 10× Olympus objective lens (OB) on the core of a multimode fiber optics. Light was transmitted by the fiber optics to an Echelle spectrograph, and spectra were recorded by an intensified charge coupled device (ICCD). A delay generator controls timing by triggering the laser Q-switch (L1), flash lamp (L2), and ICCD camera (L3). In order to find a maximum Raman-to-fluorescence ratio (R/F) a series of experiments at different time gating was carried out. In addition, two series of experiments with and without time gating was done to show the effect of time gating on fluorescence suppression. To compare time-gated with non-gated spectra, experiments were done in the same conditions without any movement of sample or making any changes in the setup. Fourteen spectra were recorded from seven calcite samples, from two separate areas of each sample. Each spectrum was an accumulation of 160 laser pulses. Fluorescence and Raman emissions overlap during the laser pulse duration as shown in Figure 4. By considering different life times for these two emissions, applying gate width helps to reduce remaining fluorescence emission.
3. Results and Discussion
Calcite has four major Raman peaks [19] at 156, 282, 713, and 1087 cm−1. The aragonite peaks can be found at 205, 700, 705, and 1085 cm−1 and vaterite has peaks [20, 21] at 301, 738, 750, 1074, and 1089 cm−1.
Fluctuation of the fluorescence background is not convenience for extracting precise information from Raman spectra. To evaluate and solve this problem, by introducing two ratios, a robust method is present. These ratios are Raman to fluorescence (R/F) [22–25] and fluorescence to laser peak ratio (F/L). Here, the Raman signals are obtained by subtracting the fluorescence background from the spectrum. Laser peak means intensity of transmitted laser through the notch filter. The fluorescence signal is the intensity of background emission under the Raman signal. Additionally, we assumed that the background is dominated by fluorescence. Higher amount of fluorescence causes reduction in the R/F and increasing the F/L ratio. For finding suitable gate width, R/F and F/L ratios at different gate widths, 2 to 30 ns, were evaluated. All spectra were recorded from the same place of the sample and experimental conditions remained Results show that there is a fluctuation in the amount of R/F and F/L ratios at gate widths less than 10 ns and this is due to the amount of R/F and F/L ratios at gate widths is less than 10 ns and this is due to missing of a part of the laser pulse or fluctuation in gate response of ICCD. Figure 5 shows the amount of R/F and F/L ratios from 7 to 25 ns. As it can be seen, optimum R/F and F/L ratios are related to 23 ns.
The R/F ratio according to each sample is shown in Figure 6 for 1085 cm−1 and in Figure 7 for 280 cm−1 peaks. Rhombs and squares are indicator of R/F ratio of gated and nongated experiments, respectively. As illustrated in Figures 6 and 7, in most samples, R/F ratio increases at gated experiments. For 1085 cm−1 peaks, the two first maxima of R/F ratios of gated experiments are 39.62 and 22.04, which are related to two different points of Vanarj Qum sample depicted in Figure 1(a). The average of R/F ratio of gated and nongated experiments is 9.07 and 2.81, respectively. This means that time gating enhances the R/F ratio of 1085 cm−1 more than three times. For R/F ratios of 280 cm−1 peaks depicted in Figure 7, two first maximum are 73.58 and 37.47 for gated experiments. The first maximum is related to Vanarj Qum illustrated in Figure 1(a) and the second one is related to Kerman depicted in Figure 1(g). The average of R/F ratios of 280 cm−1 at gated experiment is intensified more than 4 times in comparison to the nongated experiment.
The F/L ratio is depicted in Figures 8 and 9 for fluorescence of 1085 and 280 cm−1 peaks, respectively. Gated and nongated results are indicated by rhombs and squares, respectively. As illustrated, the fluorescence obviously decreases in the gated experiments. Less value of the F/L ratio signifies fewer fluorescence and better quality of spectra. In Figure 7, the two first minimums of the F/L ratio of time gated experiments are 0.00290 and 0.00296 both recorded from two different places of sample from Vanarj Qum depicted in Figure 1(a). The average of F/L ratio for gated experiments is 0.0603 and for nongated is 0.111, decreasing around 0.54 times. Figure 8 shows F/L ratio of 280 cm−1 peaks with two first minimum of 0.00068 and 0.0014 in time-gated experiment. Both of them are related to Vanarj Qum sample. Time-gated experiments have average of 0.043 and nongated experiments have average of 0.074, which means less fluorescence background and better spectrum quality for time-gated experiments.
To demonstrate the effectiveness of time-gating method, Raman spectra of Semnan sample (Figure 1(b)) are shown with and without time gating (see Figure 10). As seen, a giant fluorescence background is observed in nongated experiment. This huge background conceals the Raman peaks of the sample and causes difficulty in distinction of the peaks. By using 23 ns gate width, fluorescence signal decreases effectively and Raman peaks of 301, 738, 750, 1074, and 1089 cm−1 [20, 21] related to vaterite are obviously observed (see Figure 11).
4. Conclusion
Raman spectra excited by 532 nm pulses from an Nd:YAG laser on various calcite samples collected from different places of Iran are studied. Raman to fluorescence (R/F) and fluorescence to laser (F/L) ratios are investigated in gated and nongated experiments. A series of experiments is carried out and an optimum gate width of 23 ns is obtained. Both ratios, R/F and F/L, are improved in the time gating method at optimum gate width. Fluorescence background from biological origin has very fast life time in comparison to fluorescence from rare-earth impurities. These samples will not show much difference between the gating and nongating experiments. Samples 2, 9, 12, 13, and 14 may have biological contents. Samples containing rare-earth impurities such as having fluorescence with ms life time will show significant improvement in gating condition (samples 5, 6, 7, and 8). The Raman signals relevant to vaterite are completely hidden by fluorescence in nongated experiments while emerged in gated experiments. Results show the potential of time gating Raman spectroscopy for application in mineralogy.
Acknowledgments
The authors thank Dr. Mehrdad Behzadi and Dr. Mahboubeh Hosseini, Faculty of Department of Earth Science, Shahid Beheshti University, for consultation. In addition, thanks are due to graduate students of Department of Earth Science for assisting in gathering samples.