- Email: [email protected]
Localized quantification of anhydrous calcium carbonate polymorphs using micro-Raman spectroscopy
Vibrational Spectroscopy 95 (2018) 1–6
Contents lists available at ScienceDirect
Vibrational Spectroscopy journal homepage: www.elsevier.com/locate/vibspec
Localized quantification of anhydrous calcium carbonate polymorphs using micro-Raman spectroscopy Radek Šev9cík* , Petra Mácová Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Centre of Excellence Tel9 c, Batelovská 485-486, 588 56, Tel9c, Czech Republic
A R T I C L E I N F O
Article history: Received 20 October 2017 Received in revised form 13 December 2017 Accepted 13 December 2017 Available online 21 December 2017 Keywords: Micro-Raman Quantification Calcite Vaterite Aragonite Nanolime
A B S T R A C T
Micro-Raman spectroscopy is a powerful technique for qualitative and quantitative analysis of different mineral mixtures. In this paper, micro-Raman spectroscopy was used for quantification in local regions (180 180 mm area) of ternary mixtures of the synthetic calcium carbonate (CaCO3) polymorphs (vaterite, aragonite, calcite) as well as CaCO3 formed during the carbonation of nanolime suspension. The obtained results of localized quantification were in agreement with the detected concentrations obtained from bulk quantitative phase analysis of X-ray powder diffraction patterns. The detection limits were found to be below 0.5 wt.% for each CaCO3 polymorphs. Through the use of 2D mapping, localized quantification of CaCO3 polymorphs can be achieved. This information could be potentially useful for conservation of valuable Cultural Heritage objects, as it might influence the consolidation treatment chosen. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Calcium carbonate (CaCO3) is one of the most abundant minerals in the Earth’s crust. It can be present in three anhydrous polymorphs – vaterite, aragonite, calcite and three hydrated phases – amorphous CaCO3 (ACC), monohydrate (CaCO3H2O), ikaite (CaCO36H2O) [1,2]. CaCO3 is of high importance in natural systems as in the process of biomineralization [3,4], and in many industrial sectors where it is used as filler, extender and pigment, including the production of paper, rubber, plastic, pharmaceuticals, food, paint, textiles and numerous other different materials [5–8]. From ancient times, lime produced from CaCO3 (using limestone rocks) was the key component of mortars hardened through the carbonation reaction [7]. More recently, novel nanolime suspensions were introduced on the market for the consolidation of Cultural Heritage objects, like wall paintings and stones or for the conservation of paper, canvas and wood [9,10]. As in the case of lime mortars, the consolidation effect is based on the carbonation reaction of Ca(OH)2 resulting in the formation of CaCO3 phases [11]. It was found that different CaCO3 phases could be formed during nanolime carbonation and their formation is highly dependent on the climate conditions, such as relative humidity [12–14].
* Corresponding author. E-mail address: [email protected] (R. Šev9 cík). https://doi.org/10.1016/j.vibspec.2017.12.005 0924-2031/© 2017 Elsevier B.V. All rights reserved.
Raman spectroscopy is extensively used for qualitative analysis and its effectiveness has been demonstrated in the quantitative phase analysis of pharmaceuticals [15,16], lignin [17] and geological samples [18–21]. The main advantage of the using of micro-Raman spectroscopy (m-RS) is the ability to focus the laser beam onto the specific sites of the samples and thus perform localized quantifications. Depending on the microscope parameters used, a spatial resolution on the order of 1 mm can be potentially achieved. Another advantage is the relatively small amount of the sample needed, especially when compared with bulk techniques like X-ray powder diffraction (XRPD). In this respect, m-RS is very appealing for localized quantification and/or the analysis of samples that cannot be collected in large quantity, as in the case of the samples originated from Cultural Heritage objects. In the present paper, the methodology for the localized quantitative determination of the anhydrous calcium carbonate polymorphs – vaterite, aragonite and calcite – using m-RS is described. The validity of this method was tested on synthesized CaCO3 samples as well as on sample of carbonated nanolime. The results were compared with those obtained from quantitative phase analysis of the measured XRPD patterns, using Rietveld refinement, an established technique for quantification of components in different powder mixtures [22].
2
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
2. Experimental 2.1. Samples preparation Vaterite was synthesized following a procedures described elsewhere [23] from two supersaturated solutions of CaCl22H2O (p.a.) and K2CO3 (p.a.) without use of additives. The reagents were purchased from Lach-Ner (Czech Republic) and used as received. Aqueous solutions (c = 2 mol l1, V = 50 ml) of these reactants were prepared using deionized water (Barhstead Smart2pure, Thermo Scientific) of the following characteristics: l = 0.055 mS cm1; R = 18.2 MVxcm; pH = 8.2. The synthesis of aragonite was performed from the vaterite precursor refluxed for the 150 min [24]. The synthesis of calcite was previously described elsewhere [23]. Six mixtures containing various concentrations of CaCO3 polymorphs were prepared and the phase composition of each sample was verified using XRPD measurements. Each final powder mixture was homogenized in a zirconia ball mill (Pulverisette 23, Fritch) at 30 oscillations s1 for 30 s. The suspension of nanolime (CaLoSil E25, (IBZ-Salzchemie, Germany), c = 25 g l1, V = 5 ml) was cured in a climate chamber for the 4 weeks at constant relative humidity (RH = 65(5) %) and temperature (t = 20 (1) C). 2.2. Analytical techniques A micro-Raman spectrometer (DXR, Thermo Scientific) equipped with high resolution grating (1800 lines mm1) was used. The laser beam (l = 532 nm) was focused with a 20 objective (N.A = 0.40), and laser power of 10 mW was used. Maps were collected over an area of 180 180 mm using a 20 mm step size (100 points). Raman spectra were collected over a spectral range of 1300–50 cm1. Twenty exposures for each spectrum were recorded with exposure time 2 s. The Raman bands at 701, 711, 750 cm1 were used for quantitative analysis of aragonite, calcite and vaterite, respectively. The phase composition of the prepared CaCO3 polymorph mixtures and of the sample of the carbonated nanolime was investigated with X-ray powder diffraction (XRPD), employing a Bragg-Brentano D8 advance diffractometer (Bruker) equipped with a LynxEye 1-D silicon strip detector, using CuKa radiation and Ni filter. Patterns were collected at 40 kV and 40 mA at ambient temperature in the angular range 15–90 2u with a virtual step scan of 0.01 and counting time 0.4 s per step. The samples were allowed to spin at 15 rpm to increase the particle statistics. Quantitative phase analysis (QPA) was performed by the Rietveld method [25] using the Topas 4.2 software from Bruker AXS. Selected samples were examined with a high-resolution scanning electron microscope (SEM) Quanta 450 FEG (FEI) using a secondary electron detector. Analyses were performed at accelerating voltage between 5 and 20 kV. All samples were gold coated with a 5 nm thick layer.
Fig. 1. Raman spectra of the pure anhydrous CaCO3 polymorphs – vaterite, aragonite and calcite – in the spectral range 1300–50 cm1. Main peaks of polymorphs are indicated (V = vaterite, C = calcite, A = aragonite).
the weight concentrations of aragonite, calcite and vaterite, respectively and IA, IC, IV are their intensities of Raman signal at selected wavelengths. Then, the following equations were used for calculating the concentration of CaCO3 anhydrous polymorphs: XA ¼
XC ¼
XV ¼
aCA I701 A I711 C
þ aCA I701 þ aCV I750 A V
I711 XA C aCV I701 A
aCV I750 I711 XA V C aCA I701 I711 A C
ð1Þ
ð2Þ
ð3Þ
where aCA, aCV are the slopes of the calibration curves.
2.3. Quantification methodology The baseline correction implemented in the software OMNIC for Dispersive Raman 9.1.24 was applied on the collected spectra. The values of the intensities at 701, 711, 750 cm1 (for aragonite, calcite and vaterite, respectively) were extracted using the software SpectraGryph vs. 1.2 and the outliers were identified using software Statistica vs. 10. The average values for the specific wavenumbers of each sample were used to plot calibration curves. The calibration method used is described in [19] using the two binary calibrations (calcite/aragonite and calcite/vaterite) for the ternary mixtures. The calibration curves were plotted as the ratios of IC711/IA701 vs. XC/XA and IC711/IV750 vs. XC/XV, where XA, XC, XV are
Fig. 2. Selected region of measured XRD spectra of synthesized mixtures and testing sample. Main peaks of polymorphs are indicated (V = vaterite, C = calcite, A = aragonite).
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
The limits of detection were calculated using the equation [19,26]: rffiffiffiffiffiffiffiffiffiffiffiffiffi Nþ1 LD ¼ t sB ð4Þ N where t is the Student’s t-value defined as: t = (x–m)/s, where s is the standard deviation of the measurements and (x–m) represents the absolute deviation from the mean value; sB is the standard deviation of the blank measurements; N is the number of blank measurements.
3
Table 1 The results of the QPA with Rietveld method of the investigated samples. Sample
mix 1 mix 2 mix 3 mix 4 mix 5 mix 6 test sample nanolime
cXRPD [wt.%] Aragonite
Calcite
Vaterite
64.6(8) 22.8(2) 18.1(2) 12.5(2) 6.5(2) 11.1(2) 18.7(2) 47.4(2)
22.6(3) 34.2(3) 59.1(5) 77.3(8) 77.9(6) 60.0(5) 42.8(5) 32.3(2)
13(5) 43(1) 23(1) 10(2) 16(1) 29(1) 39(2) 20(1)
3. Results and discussion The symmetric C O stretching mode (v1), doubly degenerated asymmetric CO stretching mode (v3) and doubly degenerated inplane OCO deformation bending mode (v4), are active vibrations for the anhydrous CaCO3 polymorphs and their assignment is well established [27–30]. Unfortunately, the most intense bands are overlapping at 1084 cm1 and thus other two regions have been used for phase identification, that is, vibrations modes in the
spectral range 760–680 cm1 and lattice vibrations modes in the spectral range 500–50 cm1 [18,19]. Also two other bands at 1437 cm1 and 1750 cm1, assigned to calcite, were reported [30]. Examples of measured spectra of the pure polymorphs in the selected regions are depicted in Fig. 1. The inset in Fig. 1 shows the region used for the quantification of the anhydrous CaCO3 polymorphs in the spectral range 800–
Fig. 3. SEM images of (a) the investigated sample prepared from the synthesized CaCO3 polymorphs and (b) the nanolime suspension carbonated for 4 weeks in climate chamber (RH = 65(5) %); t = 20(1) C). The corresponding CaCO3 polymorphs are highlighted (V = vaterite, C = calcite, A = aragonite).
4
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
Fig. 5. One of the Raman spectra of the investigated samples with the curve fitted to separate the contributions of the calcite band at 711 cm1 and aragonite bands at 705 cm1and 701 cm1.
Fig. 4. 2D micro-Raman map of calcite intensities in the test sample in 180 180 mm region; red colour corresponds to the regions with the highest calcite intensities, and blue colour corresponds to the regions with the lowest calcite intensities. Two Raman spectra at specific points within the Raman map are enclosed. Corresponding CaCO3 polymorphs are indicated (V = vaterite, C = calcite, A = aragonite). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
650 cm1, which was selected due to the limited overlapping of the bands. The Raman bands at 701, 711, 750 cm1 were used for aragonite, calcite and vaterite, respectively. Examples of the detected patterns are depicted in Fig. 2. The results of the Rietveld refinement are reported in Table 1. For the refinement of vaterite phase, two structural models were used as described in [23]. SEM images (see Fig. 3) show the morphology of the synthesized CaCO3 polymorphs (Fig. 3a), as well as the morphology of the newly formed CaCO3 polymorphs through the carbonation of Ca(OH)2 nanoparticles (Fig. 3b). In case of the synthesized CaCO3 (Fig. 3a), vaterite was present as big bodies with size ca. 4.3 6.7 mm and consisted of spherules composed of clearly identifiable nanosized particles. Calcite was largely composed by typical euhedral crystals with size in the range from 0.5 0.7 mm to 2.6 3.3 mm. Aragonite was present as rods with diameter around 0.5 mm and 4.3 mm in length. Smaller aragonite rods with sizes up to 0.2 0.2 mm, were present as a consequence of the milling procedure. The morphology of the carbonated nanolime sample is depicted in Fig. 3b. The main difference, in comparison with the synthetic CaCO3 described above, was in particle size. The vateritic spherules were around 50 nm. The subhedral calcite particles were ca. 140 nm in size, whereas and the needle- like aragonite had dimensions 100 300 nm. Such nanosized particles are known to be formed during the carbonation of nanolime [14].
2D chemical maps with dimension 180 180 mm showing the abundance of the anhydrous CaCO3 polymorphs mixtures were collected for the each sample. In Fig. 4, the 2D micro-Raman map of calcite intensities in test sample is shown together with the collected Raman spectra at the specific spots. All three CaCO3 polymorphs are clearly distinguishable in the selected spectral range. The curve fitted to separate the contribution of each band in the selected spectral region is shown in Fig. 5. One can see that the contributions of the two other bands at the local maxima at 711 and 701 cm1 were negligible. The calibrations lines created using the Raman bands at 701, 711, and 750 cm1 were plotted in Fig. 6. The significance of intercepts for each calibration line was tested using the t-test: t¼
b sb
ð5Þ
where b is the intercept of the y axes and sb is its standard deviations. In both cases, it was found that t < ta (a = 0.05) and thus
Fig. 6. Calibration lines constructed for the dependencies of IC/IA on XC/XA (black line) and IC/IV on XC/XV (red line), where I is the intensity and X is the concentration of the relevant CaCO3 polymorph. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
Table 2 The results of the m-RS quantification of the testing samples and carbonated nanolime sample as calculated using Eqs. (1)–(3). cRaman of [wt.%]
test s. 1. test s. 2 test s. 3 nanolime
Ara
Cal
Vat
19.5 19.8 19.0 48.8
41.9 41.3 41.7 27.0
39.2 38.9 39.2 24.2
the only the equation y = ax could be used. The slope of the ratios IC/ IA vs. XC/XA was found to be 1.46(3) with the correlation coefficient r2 = 0.9987 and the slope of the ratios IC/IV vs. XC/XV was 8.73(14) with the r2 = 0.9986. The detection limits for aragonite, vaterite and calcite, calculated using Eq. 4, were 0.2, 0.4 and 0.4 wt.%, respectively. The constructed calibration curves were verified using the testing sample consisting of 18.7(2), 42.8(5) and 39(2) wt.% for aragonite, calcite and vaterite, respectively. The obtained results of measurements of three replicates are summarized in Table 2. XRPD results were found to be in agreement with Raman results and relative errors were calculated to be 4.0, 2.7 and 1.1% for aragonite, calcite and vaterite, respectively. The obtained calibration lines were used also for the detection of the anhydrous CaCO3 polymorphs in the nanolime carbonated in the climate chamber, as described in the experimental part. The XRPD analysis showed that the sample was consisting of all three anhydrous polymorphs as it was also shown using SEM (Table 1, Fig. 3b). The calculated wt.% determined from the m-RS data was found to be 48.8, 27.0 and 24.2 wt.% for aragonite, calcite and vaterite, respectively. In comparison with the XRPD results (Table 1), the difference in detected concentrations is less than 5.3 wt.%. The calibration regression coefficients calculated in the work [19] were slightly closer to 1 (0.9999), which could be ascribed to the measurement of only one spectrum per sample using the FTRaman instrument. Dickinson and McGrath did not detect the linear correlation as published in [31]. However, from the Raman spectra of pure CaCO3 polymorphs as well as of the prepared mixture, it seems that, in their work, the CaCO3 polymorphs contained other species and thus these impurities likely affected their results causing the observed nonlinearity of the plotted curves. Dandeu et al. [18] used Partial Least Squares (PLS) method for the quantification of the ternary CaCO3 polymorphs mixture. The Raman bands observed in a low frequency region from 400 to 50 cm1, which corresponds to the lattice mode vibrations, were used for the multivariable regression. In such region, the vaterite bands overlap with aragonite and calcite at 205 cm1 and 281 cm1, respectively. Also overlapping of all three CaCO3 polymorphs exists at 150 cm1. In this case, the simple linear regression method described above cannot be used. With the increasing availability of the micro-Raman spectrometer with the imaging detectors, the total time needed for the measurements of all samples could be decreased from hours to minutes, and thus m-RS could be not only reliable but also the powerful and fast technique for the local quantification of the CaCO3 polymorphs mixtures. 4. Conclusion Micro-Raman spectroscopy was used for the localized quantification of ternary CaCO3 polymorphs mixtures. The calibration samples were prepared by mixing of synthesized
5
CaCO3 polymorphs and 2D maps of area 180 180 mm were measured for each sample. The average intensities at 701, 711, 750 cm1 were used for quantitative analysis of aragonite, calcite and vaterite, respectively and two binary calibration curves were plotted. CaCO3 of two different origins were used for testing of constructed calibrations curves – mixture of synthesized CaCO3 and CaCO3 polymorphs formed during the carbonation of nanolime suspension. The obtained results were in agreement with the concentrations determined by quantitative phase analysis using Rietveld refinement of X-ray powder diffraction patterns with the differences of the calculated concentrations around 5 wt.%. The described methodology could be helpful for the local quantification of the newly formed CaCO3 polymorphs within Cultural Heritage objects during and/or after the consolidation treatment with lime-based agents. Acknowledgements The authors gratefully acknowledge the support of the project No. LO1219 under the Ministry of Education, Youth and Sports National sustainability program I of Czech Republic. We also thank anonymous referees for their insightful comments and suggestions. References [1] F. Lippmann, Sedimentary Carbonate Minerals, Springer-Verlag, Berlin, New York, 1973. [2] P. Bots, L.G. Benning, J.-D. Rodriguez-Blanco, T. Roncal-Herrero, S. Shaw, Cryst. Growth Des. 12 (2012) 3806–3814, doi:http://dx.doi.org/10.1021/cg300676b. [3] F.C. Meldrum, H. Cölfen, Chem. Rev. 108 (2008) 4332–4432, doi:http://dx.doi. org/10.1021/cr8002856. [4] N.K. Dhami, M.S. Reddy, A. Mukherjee, Front. Microbiol. 4 (2013), doi:http://dx. doi.org/10.3389/fmicb.2013.00314. [5] J. Dobrev, P. Markovic (Eds.), Calcite: Formation, Properties, and Applications, Nova Science Publishers, New York, 2012. [6] L. Bre9 cevi c, D. Kralj, Croat. Chem. Acta (2007) (2007) 467–484. [7] A.D. Cowper, Building Research Station Building Research Establishment, Lime and lime mortars, Donhead, Shaftesbury, 1998. [8] D.B. Trushina, T.V. Bukreeva, M.V. Kovalchuk, M.N. Antipina, Mater. Sci. Eng. C 45 (2014) 644–658, doi:http://dx.doi.org/10.1016/j.msec.2014.04.050. [9] P. Baglioni, D. Chelazzi, R. Giorgi, Nanotechnologies in the Conservation of Cultural Heritage, Springer, Netherlands, Dordrecht, 2015. [10] R. van Hees, R. Veiga, Z. Slížková, Mater. Struct. 50 (2016) 65, doi:http://dx.doi. org/10.1617/s11527-016-0894-5. [11] K. Van Balen, Cem. Concr. Res. 35 (2005) 647–657, doi:http://dx.doi.org/ 10.1016/j.cemconres.2004.06.020. [12] L.S. Gomez-Villalba, P. López-Arce, M. Alvarez de Buergo, R. Fort, Appl. Phys. A 104 (2011) 1249–1254, doi:http://dx.doi.org/10.1007/s00339-011-6457-2. [13] L.S. Gomez-Villalba, P. López-Arce, R. Fort, Appl. Phys. A 106 (2012) 213–217, doi:http://dx.doi.org/10.1007/s00339-011-6550-6. [14] C. Rodriguez-Navarro, K. Elert, R. Šev9cík, CrystEngComm 18 (2016) 6594–6607, doi:http://dx.doi.org/10.1039/C6CE01202G. [15] B. Nagy, A. Farkas, A. Balogh, H. Pataki, B. Vajna, Z.K. Nagy, G. Marosi, J. Pharm. Biomed. Anal. 128 (2016) 236–246, doi:http://dx.doi.org/10.1016/j. jpba.2016.05.036. [16] H. Wang, L. Williams, S. Hoe, D. Lechuga-Ballesteros, R. Vehring, Appl. Spectrosc. 69 (2015) 823–833, doi:http://dx.doi.org/10.1366/14-07812. [17] H. Miyafuji, K. Komai, T. Kanbayashi, Vib. Spectrosc. 88 (2017) 9–13, doi:http:// dx.doi.org/10.1016/j.vibspec.2016.10.011. [18] A. Dandeu, B. Humbert, C. Carteret, H. Muhr, E. Plasari, J.M. Bossoutrot, Chem. Eng. Technol. 29 (2006) 221–225, doi:http://dx.doi.org/10.1002/ ceat.200500354. [19] C.G. Kontoyannis, N.V. Vagenas, Analyst 125 (2000) 251–255, doi:http://dx. doi.org/10.1039/a908609i. [20] T. Dörfer, W. Schumacher, N. Tarcea, M. Schmitt, J. Popp, J. Raman Spectrosc. 41 (2010) 684–689, doi:http://dx.doi.org/10.1002/jrs.2503. [21] N.V. Vagenas, C.G. Kontoyannis, Vib. Spectrosc. 32 (2003) 261–264, doi:http:// dx.doi.org/10.1016/S0924-2031(03)00027-4. [22] A.F. Gualtieri, A. Viani, C. Montanari, Cem. Concr. Res. 36 (2006) 401–406, doi: http://dx.doi.org/10.1016/j.cemconres.2005.02.001. [23] R. Šev9 cík, M. Pérez-Estébanez, A. Viani, P. Šašek, P. Mácová, Powder Technol. 284 (2015) 265–271, doi:http://dx.doi.org/10.1016/j.powtec.2015.06.064. [24] R. Šev9 cík, P. Mácová, M. Pérez-Estébanez, Adv. Mater. Res. 1119 (2015) 466–470, doi:http://dx.doi.org/10.4028/www.scientific.net/AMR.1119.466. [25] H.M. Rietveld, J. Appl. Crystallogr. 2 (1969) 65–71, doi:http://dx.doi.org/ 10.1107/S0021889869006558.
6
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
[26] D.A. Skoog, D.M. West, F.J. Holler, S.R. Crouch, Fundamentals of Analytical Chemistry, 9th ed., Brooks Cole, Belmont, CA, 2013. [27] B. Schrader, D. Bougeard (Eds.), Infrared and Raman Spectroscopy: Methods and Applications, VCH Weinheim, New York, 1995. [28] U. Wehrmeister, A.L. Soldati, D.E. Jacob, T. Häger, W. Hofmeister, J. Raman Spectrosc. 41 (2010) 193–201, doi:http://dx.doi.org/10.1002/jrs.2438.
[29] U. Wehrmeister, D.E. Jacob, A.L. Soldati, N. Loges, T. Häger, W. Hofmeister, J. Raman Spectrosc. 42 (2011) 926–935, doi:http://dx.doi.org/10.1002/jrs.2835. [30] S. Gunasekaran, G. Anbalagan, S. Pandi, J. Raman Spectrosc. 37 (2006) 892–899, doi:http://dx.doi.org/10.1002/jrs.1518. [31] S.R. Dickinson, K.M. McGrath, Analyst 126 (2001) 1118–1121, doi:http://dx.doi. org/10.1039/B103004N.
Contents lists available at ScienceDirect
Vibrational Spectroscopy journal homepage: www.elsevier.com/locate/vibspec
Localized quantification of anhydrous calcium carbonate polymorphs using micro-Raman spectroscopy Radek Šev9cík* , Petra Mácová Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Centre of Excellence Tel9 c, Batelovská 485-486, 588 56, Tel9c, Czech Republic
A R T I C L E I N F O
Article history: Received 20 October 2017 Received in revised form 13 December 2017 Accepted 13 December 2017 Available online 21 December 2017 Keywords: Micro-Raman Quantification Calcite Vaterite Aragonite Nanolime
A B S T R A C T
Micro-Raman spectroscopy is a powerful technique for qualitative and quantitative analysis of different mineral mixtures. In this paper, micro-Raman spectroscopy was used for quantification in local regions (180 180 mm area) of ternary mixtures of the synthetic calcium carbonate (CaCO3) polymorphs (vaterite, aragonite, calcite) as well as CaCO3 formed during the carbonation of nanolime suspension. The obtained results of localized quantification were in agreement with the detected concentrations obtained from bulk quantitative phase analysis of X-ray powder diffraction patterns. The detection limits were found to be below 0.5 wt.% for each CaCO3 polymorphs. Through the use of 2D mapping, localized quantification of CaCO3 polymorphs can be achieved. This information could be potentially useful for conservation of valuable Cultural Heritage objects, as it might influence the consolidation treatment chosen. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Calcium carbonate (CaCO3) is one of the most abundant minerals in the Earth’s crust. It can be present in three anhydrous polymorphs – vaterite, aragonite, calcite and three hydrated phases – amorphous CaCO3 (ACC), monohydrate (CaCO3H2O), ikaite (CaCO36H2O) [1,2]. CaCO3 is of high importance in natural systems as in the process of biomineralization [3,4], and in many industrial sectors where it is used as filler, extender and pigment, including the production of paper, rubber, plastic, pharmaceuticals, food, paint, textiles and numerous other different materials [5–8]. From ancient times, lime produced from CaCO3 (using limestone rocks) was the key component of mortars hardened through the carbonation reaction [7]. More recently, novel nanolime suspensions were introduced on the market for the consolidation of Cultural Heritage objects, like wall paintings and stones or for the conservation of paper, canvas and wood [9,10]. As in the case of lime mortars, the consolidation effect is based on the carbonation reaction of Ca(OH)2 resulting in the formation of CaCO3 phases [11]. It was found that different CaCO3 phases could be formed during nanolime carbonation and their formation is highly dependent on the climate conditions, such as relative humidity [12–14].
* Corresponding author. E-mail address: [email protected] (R. Šev9 cík). https://doi.org/10.1016/j.vibspec.2017.12.005 0924-2031/© 2017 Elsevier B.V. All rights reserved.
Raman spectroscopy is extensively used for qualitative analysis and its effectiveness has been demonstrated in the quantitative phase analysis of pharmaceuticals [15,16], lignin [17] and geological samples [18–21]. The main advantage of the using of micro-Raman spectroscopy (m-RS) is the ability to focus the laser beam onto the specific sites of the samples and thus perform localized quantifications. Depending on the microscope parameters used, a spatial resolution on the order of 1 mm can be potentially achieved. Another advantage is the relatively small amount of the sample needed, especially when compared with bulk techniques like X-ray powder diffraction (XRPD). In this respect, m-RS is very appealing for localized quantification and/or the analysis of samples that cannot be collected in large quantity, as in the case of the samples originated from Cultural Heritage objects. In the present paper, the methodology for the localized quantitative determination of the anhydrous calcium carbonate polymorphs – vaterite, aragonite and calcite – using m-RS is described. The validity of this method was tested on synthesized CaCO3 samples as well as on sample of carbonated nanolime. The results were compared with those obtained from quantitative phase analysis of the measured XRPD patterns, using Rietveld refinement, an established technique for quantification of components in different powder mixtures [22].
2
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
2. Experimental 2.1. Samples preparation Vaterite was synthesized following a procedures described elsewhere [23] from two supersaturated solutions of CaCl22H2O (p.a.) and K2CO3 (p.a.) without use of additives. The reagents were purchased from Lach-Ner (Czech Republic) and used as received. Aqueous solutions (c = 2 mol l1, V = 50 ml) of these reactants were prepared using deionized water (Barhstead Smart2pure, Thermo Scientific) of the following characteristics: l = 0.055 mS cm1; R = 18.2 MVxcm; pH = 8.2. The synthesis of aragonite was performed from the vaterite precursor refluxed for the 150 min [24]. The synthesis of calcite was previously described elsewhere [23]. Six mixtures containing various concentrations of CaCO3 polymorphs were prepared and the phase composition of each sample was verified using XRPD measurements. Each final powder mixture was homogenized in a zirconia ball mill (Pulverisette 23, Fritch) at 30 oscillations s1 for 30 s. The suspension of nanolime (CaLoSil E25, (IBZ-Salzchemie, Germany), c = 25 g l1, V = 5 ml) was cured in a climate chamber for the 4 weeks at constant relative humidity (RH = 65(5) %) and temperature (t = 20 (1) C). 2.2. Analytical techniques A micro-Raman spectrometer (DXR, Thermo Scientific) equipped with high resolution grating (1800 lines mm1) was used. The laser beam (l = 532 nm) was focused with a 20 objective (N.A = 0.40), and laser power of 10 mW was used. Maps were collected over an area of 180 180 mm using a 20 mm step size (100 points). Raman spectra were collected over a spectral range of 1300–50 cm1. Twenty exposures for each spectrum were recorded with exposure time 2 s. The Raman bands at 701, 711, 750 cm1 were used for quantitative analysis of aragonite, calcite and vaterite, respectively. The phase composition of the prepared CaCO3 polymorph mixtures and of the sample of the carbonated nanolime was investigated with X-ray powder diffraction (XRPD), employing a Bragg-Brentano D8 advance diffractometer (Bruker) equipped with a LynxEye 1-D silicon strip detector, using CuKa radiation and Ni filter. Patterns were collected at 40 kV and 40 mA at ambient temperature in the angular range 15–90 2u with a virtual step scan of 0.01 and counting time 0.4 s per step. The samples were allowed to spin at 15 rpm to increase the particle statistics. Quantitative phase analysis (QPA) was performed by the Rietveld method [25] using the Topas 4.2 software from Bruker AXS. Selected samples were examined with a high-resolution scanning electron microscope (SEM) Quanta 450 FEG (FEI) using a secondary electron detector. Analyses were performed at accelerating voltage between 5 and 20 kV. All samples were gold coated with a 5 nm thick layer.
Fig. 1. Raman spectra of the pure anhydrous CaCO3 polymorphs – vaterite, aragonite and calcite – in the spectral range 1300–50 cm1. Main peaks of polymorphs are indicated (V = vaterite, C = calcite, A = aragonite).
the weight concentrations of aragonite, calcite and vaterite, respectively and IA, IC, IV are their intensities of Raman signal at selected wavelengths. Then, the following equations were used for calculating the concentration of CaCO3 anhydrous polymorphs: XA ¼
XC ¼
XV ¼
aCA I701 A I711 C
þ aCA I701 þ aCV I750 A V
I711 XA C aCV I701 A
aCV I750 I711 XA V C aCA I701 I711 A C
ð1Þ
ð2Þ
ð3Þ
where aCA, aCV are the slopes of the calibration curves.
2.3. Quantification methodology The baseline correction implemented in the software OMNIC for Dispersive Raman 9.1.24 was applied on the collected spectra. The values of the intensities at 701, 711, 750 cm1 (for aragonite, calcite and vaterite, respectively) were extracted using the software SpectraGryph vs. 1.2 and the outliers were identified using software Statistica vs. 10. The average values for the specific wavenumbers of each sample were used to plot calibration curves. The calibration method used is described in [19] using the two binary calibrations (calcite/aragonite and calcite/vaterite) for the ternary mixtures. The calibration curves were plotted as the ratios of IC711/IA701 vs. XC/XA and IC711/IV750 vs. XC/XV, where XA, XC, XV are
Fig. 2. Selected region of measured XRD spectra of synthesized mixtures and testing sample. Main peaks of polymorphs are indicated (V = vaterite, C = calcite, A = aragonite).
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
The limits of detection were calculated using the equation [19,26]: rffiffiffiffiffiffiffiffiffiffiffiffiffi Nþ1 LD ¼ t sB ð4Þ N where t is the Student’s t-value defined as: t = (x–m)/s, where s is the standard deviation of the measurements and (x–m) represents the absolute deviation from the mean value; sB is the standard deviation of the blank measurements; N is the number of blank measurements.
3
Table 1 The results of the QPA with Rietveld method of the investigated samples. Sample
mix 1 mix 2 mix 3 mix 4 mix 5 mix 6 test sample nanolime
cXRPD [wt.%] Aragonite
Calcite
Vaterite
64.6(8) 22.8(2) 18.1(2) 12.5(2) 6.5(2) 11.1(2) 18.7(2) 47.4(2)
22.6(3) 34.2(3) 59.1(5) 77.3(8) 77.9(6) 60.0(5) 42.8(5) 32.3(2)
13(5) 43(1) 23(1) 10(2) 16(1) 29(1) 39(2) 20(1)
3. Results and discussion The symmetric C O stretching mode (v1), doubly degenerated asymmetric CO stretching mode (v3) and doubly degenerated inplane OCO deformation bending mode (v4), are active vibrations for the anhydrous CaCO3 polymorphs and their assignment is well established [27–30]. Unfortunately, the most intense bands are overlapping at 1084 cm1 and thus other two regions have been used for phase identification, that is, vibrations modes in the
spectral range 760–680 cm1 and lattice vibrations modes in the spectral range 500–50 cm1 [18,19]. Also two other bands at 1437 cm1 and 1750 cm1, assigned to calcite, were reported [30]. Examples of measured spectra of the pure polymorphs in the selected regions are depicted in Fig. 1. The inset in Fig. 1 shows the region used for the quantification of the anhydrous CaCO3 polymorphs in the spectral range 800–
Fig. 3. SEM images of (a) the investigated sample prepared from the synthesized CaCO3 polymorphs and (b) the nanolime suspension carbonated for 4 weeks in climate chamber (RH = 65(5) %); t = 20(1) C). The corresponding CaCO3 polymorphs are highlighted (V = vaterite, C = calcite, A = aragonite).
4
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
Fig. 5. One of the Raman spectra of the investigated samples with the curve fitted to separate the contributions of the calcite band at 711 cm1 and aragonite bands at 705 cm1and 701 cm1.
Fig. 4. 2D micro-Raman map of calcite intensities in the test sample in 180 180 mm region; red colour corresponds to the regions with the highest calcite intensities, and blue colour corresponds to the regions with the lowest calcite intensities. Two Raman spectra at specific points within the Raman map are enclosed. Corresponding CaCO3 polymorphs are indicated (V = vaterite, C = calcite, A = aragonite). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
650 cm1, which was selected due to the limited overlapping of the bands. The Raman bands at 701, 711, 750 cm1 were used for aragonite, calcite and vaterite, respectively. Examples of the detected patterns are depicted in Fig. 2. The results of the Rietveld refinement are reported in Table 1. For the refinement of vaterite phase, two structural models were used as described in [23]. SEM images (see Fig. 3) show the morphology of the synthesized CaCO3 polymorphs (Fig. 3a), as well as the morphology of the newly formed CaCO3 polymorphs through the carbonation of Ca(OH)2 nanoparticles (Fig. 3b). In case of the synthesized CaCO3 (Fig. 3a), vaterite was present as big bodies with size ca. 4.3 6.7 mm and consisted of spherules composed of clearly identifiable nanosized particles. Calcite was largely composed by typical euhedral crystals with size in the range from 0.5 0.7 mm to 2.6 3.3 mm. Aragonite was present as rods with diameter around 0.5 mm and 4.3 mm in length. Smaller aragonite rods with sizes up to 0.2 0.2 mm, were present as a consequence of the milling procedure. The morphology of the carbonated nanolime sample is depicted in Fig. 3b. The main difference, in comparison with the synthetic CaCO3 described above, was in particle size. The vateritic spherules were around 50 nm. The subhedral calcite particles were ca. 140 nm in size, whereas and the needle- like aragonite had dimensions 100 300 nm. Such nanosized particles are known to be formed during the carbonation of nanolime [14].
2D chemical maps with dimension 180 180 mm showing the abundance of the anhydrous CaCO3 polymorphs mixtures were collected for the each sample. In Fig. 4, the 2D micro-Raman map of calcite intensities in test sample is shown together with the collected Raman spectra at the specific spots. All three CaCO3 polymorphs are clearly distinguishable in the selected spectral range. The curve fitted to separate the contribution of each band in the selected spectral region is shown in Fig. 5. One can see that the contributions of the two other bands at the local maxima at 711 and 701 cm1 were negligible. The calibrations lines created using the Raman bands at 701, 711, and 750 cm1 were plotted in Fig. 6. The significance of intercepts for each calibration line was tested using the t-test: t¼
b sb
ð5Þ
where b is the intercept of the y axes and sb is its standard deviations. In both cases, it was found that t < ta (a = 0.05) and thus
Fig. 6. Calibration lines constructed for the dependencies of IC/IA on XC/XA (black line) and IC/IV on XC/XV (red line), where I is the intensity and X is the concentration of the relevant CaCO3 polymorph. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
Table 2 The results of the m-RS quantification of the testing samples and carbonated nanolime sample as calculated using Eqs. (1)–(3). cRaman of [wt.%]
test s. 1. test s. 2 test s. 3 nanolime
Ara
Cal
Vat
19.5 19.8 19.0 48.8
41.9 41.3 41.7 27.0
39.2 38.9 39.2 24.2
the only the equation y = ax could be used. The slope of the ratios IC/ IA vs. XC/XA was found to be 1.46(3) with the correlation coefficient r2 = 0.9987 and the slope of the ratios IC/IV vs. XC/XV was 8.73(14) with the r2 = 0.9986. The detection limits for aragonite, vaterite and calcite, calculated using Eq. 4, were 0.2, 0.4 and 0.4 wt.%, respectively. The constructed calibration curves were verified using the testing sample consisting of 18.7(2), 42.8(5) and 39(2) wt.% for aragonite, calcite and vaterite, respectively. The obtained results of measurements of three replicates are summarized in Table 2. XRPD results were found to be in agreement with Raman results and relative errors were calculated to be 4.0, 2.7 and 1.1% for aragonite, calcite and vaterite, respectively. The obtained calibration lines were used also for the detection of the anhydrous CaCO3 polymorphs in the nanolime carbonated in the climate chamber, as described in the experimental part. The XRPD analysis showed that the sample was consisting of all three anhydrous polymorphs as it was also shown using SEM (Table 1, Fig. 3b). The calculated wt.% determined from the m-RS data was found to be 48.8, 27.0 and 24.2 wt.% for aragonite, calcite and vaterite, respectively. In comparison with the XRPD results (Table 1), the difference in detected concentrations is less than 5.3 wt.%. The calibration regression coefficients calculated in the work [19] were slightly closer to 1 (0.9999), which could be ascribed to the measurement of only one spectrum per sample using the FTRaman instrument. Dickinson and McGrath did not detect the linear correlation as published in [31]. However, from the Raman spectra of pure CaCO3 polymorphs as well as of the prepared mixture, it seems that, in their work, the CaCO3 polymorphs contained other species and thus these impurities likely affected their results causing the observed nonlinearity of the plotted curves. Dandeu et al. [18] used Partial Least Squares (PLS) method for the quantification of the ternary CaCO3 polymorphs mixture. The Raman bands observed in a low frequency region from 400 to 50 cm1, which corresponds to the lattice mode vibrations, were used for the multivariable regression. In such region, the vaterite bands overlap with aragonite and calcite at 205 cm1 and 281 cm1, respectively. Also overlapping of all three CaCO3 polymorphs exists at 150 cm1. In this case, the simple linear regression method described above cannot be used. With the increasing availability of the micro-Raman spectrometer with the imaging detectors, the total time needed for the measurements of all samples could be decreased from hours to minutes, and thus m-RS could be not only reliable but also the powerful and fast technique for the local quantification of the CaCO3 polymorphs mixtures. 4. Conclusion Micro-Raman spectroscopy was used for the localized quantification of ternary CaCO3 polymorphs mixtures. The calibration samples were prepared by mixing of synthesized
5
CaCO3 polymorphs and 2D maps of area 180 180 mm were measured for each sample. The average intensities at 701, 711, 750 cm1 were used for quantitative analysis of aragonite, calcite and vaterite, respectively and two binary calibration curves were plotted. CaCO3 of two different origins were used for testing of constructed calibrations curves – mixture of synthesized CaCO3 and CaCO3 polymorphs formed during the carbonation of nanolime suspension. The obtained results were in agreement with the concentrations determined by quantitative phase analysis using Rietveld refinement of X-ray powder diffraction patterns with the differences of the calculated concentrations around 5 wt.%. The described methodology could be helpful for the local quantification of the newly formed CaCO3 polymorphs within Cultural Heritage objects during and/or after the consolidation treatment with lime-based agents. Acknowledgements The authors gratefully acknowledge the support of the project No. LO1219 under the Ministry of Education, Youth and Sports National sustainability program I of Czech Republic. We also thank anonymous referees for their insightful comments and suggestions. References [1] F. Lippmann, Sedimentary Carbonate Minerals, Springer-Verlag, Berlin, New York, 1973. [2] P. Bots, L.G. Benning, J.-D. Rodriguez-Blanco, T. Roncal-Herrero, S. Shaw, Cryst. Growth Des. 12 (2012) 3806–3814, doi:http://dx.doi.org/10.1021/cg300676b. [3] F.C. Meldrum, H. Cölfen, Chem. Rev. 108 (2008) 4332–4432, doi:http://dx.doi. org/10.1021/cr8002856. [4] N.K. Dhami, M.S. Reddy, A. Mukherjee, Front. Microbiol. 4 (2013), doi:http://dx. doi.org/10.3389/fmicb.2013.00314. [5] J. Dobrev, P. Markovic (Eds.), Calcite: Formation, Properties, and Applications, Nova Science Publishers, New York, 2012. [6] L. Bre9 cevi c, D. Kralj, Croat. Chem. Acta (2007) (2007) 467–484. [7] A.D. Cowper, Building Research Station Building Research Establishment, Lime and lime mortars, Donhead, Shaftesbury, 1998. [8] D.B. Trushina, T.V. Bukreeva, M.V. Kovalchuk, M.N. Antipina, Mater. Sci. Eng. C 45 (2014) 644–658, doi:http://dx.doi.org/10.1016/j.msec.2014.04.050. [9] P. Baglioni, D. Chelazzi, R. Giorgi, Nanotechnologies in the Conservation of Cultural Heritage, Springer, Netherlands, Dordrecht, 2015. [10] R. van Hees, R. Veiga, Z. Slížková, Mater. Struct. 50 (2016) 65, doi:http://dx.doi. org/10.1617/s11527-016-0894-5. [11] K. Van Balen, Cem. Concr. Res. 35 (2005) 647–657, doi:http://dx.doi.org/ 10.1016/j.cemconres.2004.06.020. [12] L.S. Gomez-Villalba, P. López-Arce, M. Alvarez de Buergo, R. Fort, Appl. Phys. A 104 (2011) 1249–1254, doi:http://dx.doi.org/10.1007/s00339-011-6457-2. [13] L.S. Gomez-Villalba, P. López-Arce, R. Fort, Appl. Phys. A 106 (2012) 213–217, doi:http://dx.doi.org/10.1007/s00339-011-6550-6. [14] C. Rodriguez-Navarro, K. Elert, R. Šev9cík, CrystEngComm 18 (2016) 6594–6607, doi:http://dx.doi.org/10.1039/C6CE01202G. [15] B. Nagy, A. Farkas, A. Balogh, H. Pataki, B. Vajna, Z.K. Nagy, G. Marosi, J. Pharm. Biomed. Anal. 128 (2016) 236–246, doi:http://dx.doi.org/10.1016/j. jpba.2016.05.036. [16] H. Wang, L. Williams, S. Hoe, D. Lechuga-Ballesteros, R. Vehring, Appl. Spectrosc. 69 (2015) 823–833, doi:http://dx.doi.org/10.1366/14-07812. [17] H. Miyafuji, K. Komai, T. Kanbayashi, Vib. Spectrosc. 88 (2017) 9–13, doi:http:// dx.doi.org/10.1016/j.vibspec.2016.10.011. [18] A. Dandeu, B. Humbert, C. Carteret, H. Muhr, E. Plasari, J.M. Bossoutrot, Chem. Eng. Technol. 29 (2006) 221–225, doi:http://dx.doi.org/10.1002/ ceat.200500354. [19] C.G. Kontoyannis, N.V. Vagenas, Analyst 125 (2000) 251–255, doi:http://dx. doi.org/10.1039/a908609i. [20] T. Dörfer, W. Schumacher, N. Tarcea, M. Schmitt, J. Popp, J. Raman Spectrosc. 41 (2010) 684–689, doi:http://dx.doi.org/10.1002/jrs.2503. [21] N.V. Vagenas, C.G. Kontoyannis, Vib. Spectrosc. 32 (2003) 261–264, doi:http:// dx.doi.org/10.1016/S0924-2031(03)00027-4. [22] A.F. Gualtieri, A. Viani, C. Montanari, Cem. Concr. Res. 36 (2006) 401–406, doi: http://dx.doi.org/10.1016/j.cemconres.2005.02.001. [23] R. Šev9 cík, M. Pérez-Estébanez, A. Viani, P. Šašek, P. Mácová, Powder Technol. 284 (2015) 265–271, doi:http://dx.doi.org/10.1016/j.powtec.2015.06.064. [24] R. Šev9 cík, P. Mácová, M. Pérez-Estébanez, Adv. Mater. Res. 1119 (2015) 466–470, doi:http://dx.doi.org/10.4028/www.scientific.net/AMR.1119.466. [25] H.M. Rietveld, J. Appl. Crystallogr. 2 (1969) 65–71, doi:http://dx.doi.org/ 10.1107/S0021889869006558.
6
R. Šev9 cík, P. Mácová / Vibrational Spectroscopy 95 (2018) 1–6
[26] D.A. Skoog, D.M. West, F.J. Holler, S.R. Crouch, Fundamentals of Analytical Chemistry, 9th ed., Brooks Cole, Belmont, CA, 2013. [27] B. Schrader, D. Bougeard (Eds.), Infrared and Raman Spectroscopy: Methods and Applications, VCH Weinheim, New York, 1995. [28] U. Wehrmeister, A.L. Soldati, D.E. Jacob, T. Häger, W. Hofmeister, J. Raman Spectrosc. 41 (2010) 193–201, doi:http://dx.doi.org/10.1002/jrs.2438.
[29] U. Wehrmeister, D.E. Jacob, A.L. Soldati, N. Loges, T. Häger, W. Hofmeister, J. Raman Spectrosc. 42 (2011) 926–935, doi:http://dx.doi.org/10.1002/jrs.2835. [30] S. Gunasekaran, G. Anbalagan, S. Pandi, J. Raman Spectrosc. 37 (2006) 892–899, doi:http://dx.doi.org/10.1002/jrs.1518. [31] S.R. Dickinson, K.M. McGrath, Analyst 126 (2001) 1118–1121, doi:http://dx.doi. org/10.1039/B103004N.