Micro- and Nano- Raman analyses of chalk Borromeo, L1, 2, *, Egeland N1, 2, Minde M1, 2, Zimmermann, U1, 2,

Andò, S3,Toccafondi, C4, Ossikovski R4

1Department of Petroleum Engineering, University of Stavanger, 4036 Stavanger, Norway, laura.borromeo@uis.no 2 The National IOR Centre of Norway, University of Stavanger, 4036 Stavanger, Norway; 3Department of Earth and Environmental Sciences, University of Milano-Bicocca,

20122, Milano, Italy; 4LPICM, CNRS, Ecole Polytechnique, Université Paris Saclay, 91128, Palaiseau, France

RAMAN SPECTROSCOPY

RAMAN SPECTROSCOPY IS AN INNOVATIVE, EFFICIENT, USER-FRIENDLY TECHNIQUE, REPRESENTING A POWERFUL TOOL TO CONFIDENTLY IDENTIFY MINERALS

Coralline algae, Maledives 16,54% MgCO3

CASE HISTORY Cirripedia, Italy 0,89 % MgCO3

Raman shift (cm-1) In

tens

ity

ν1 1086

ν1 1090

L 282

L 287

800 600 400 1000 200

500 micron

500 micron

RESULTS

Micro - Raman on Flooded chalk

T 156 ν4

712

T 160

ν4 717

Mg-calcites have not been thoroughly investigated by Raman spectroscopy [1-3].

HERE, we show that Raman spectra of carbonates are sufficiently sensitive to the structural and chemical changes occurring when Mg2+ substitutes Ca 2+ in the lattice (Fig. 1). More than 100 biological carbonate samples with variable Mg content (0 - 20 mol % MgCO3) have been studied with Raman spectroscopy, SEM-EDS, and EMPA-WDS. The peaks positions are directly linked to the amount of MgCO3:

THE SHIFT INCREASE WITH THE Mg CONTENT allowing us to estimate Mg from the positions of the Raman peaks (Figs. 4, 5).

The two peaks that resulted to be the most reliable for the estimation of the Mg content are: ν 4: 709-718 cm -1 (Fig. 2) ν1: 1084-1091 cm -1 (Fig. 3) Unfortunately there is a certain dispersion of peak positions and the same Raman position can be detected in samples having different Mg content (Figs. 2, 3). To better estimate the reliability of the equations and visualize the discrepancy between calculated and experimental results, we compared the measured MgCO3 content to the compositions obtained using the L, ν4, ν1 modes equations (Fig. 4). As expected, the difference between theoretical and experimental data is often significant. Therefore, it seems to be imprecise to quantify the exact content of Mg using a function based on a single Raman peak. However, the combination of three peaks gives mostly a powerful tool to classify Mg-calcites. When L, ν4, ν1 positions are represented in the same 3D plot, two groups characterized by different Mg concentration are present (Fig. 5): - from 0 to 5.5 % mol MgCO3: LOW - Mg CALCITES - from 10.5 to 20 % mol MgCO3: HIGH - Mg CALCITES

Advantages: ! Non destructive technique, quickness and microscopic resolution (down to 1-2 µm) ! No time-consuming and specific preparation of the sample ! Can be used directly on rock samples, thin sections, micromounts, can be applied to any group of minerals ! The very same crystal analized under the spectrometer could be later analysed by microprobe or other instrument ! Identification of polymorphs ! Semi-quantitative chemical analysis Disadvantages ! Fluorescence and low signal of some species ! Difficulty to prepare an appropriate database

It is based on the identification of diagnostic peaks and comparison with reference spectra. The position of Raman peaks for a substance is constant, depending only on structure and chemistry. Every Raman spectrum is like a fingerprint allowing a robust mineral identification. The Raman effect

ACKNOWLEDGEMENTS: the authors acknowledge the Research Council of Norway and the industry partners;  ConocoPhillips Skandinavia AS, BP Norge AS, Det Norske Oljeselskap AS, Eni Norge AS, Maersk Oil Norway AS, DONG Energy A/S, Denmark, Statoil Petroleum AS, ENGIE E&P NORGE AS, Lundin Norway AS, Halliburton AS, Schlumberger Norge AS, Wintershall Norge AS of The National IOR Centre of Norway for support. REFERENCES: 1] Bischoff, W.D., Sharma, S.K., Mackenzie, F.T., 1985, Am. Mineral. 70, 581–589; 2] Unvros, J., Sharma S. K., Mackenzie T., 1991. Am. Mineral.,76, 641-646; 3]Vagenas N.V., Kontoyannis C.G. 2003. Vibr. Spec., 32, 261–264.

Fig. 7: ULTRA LONG TERM TEST (ULTT) was cut longitudinal and sample map with sample locations and associated results are indicated. Yellow arrow indicates flooding direction. Cal = Low-Mg calcite, Mg-cal = High-Mg calcite, Mg = Magnesite, Unknown = minerals that have not yet been identified.

Fig. 6: Sample map of LONG TERM TEST (LT1_p1) with sample locations and associated results indicated. Yellow arrow indicates flooding direction. Cal = Low-Mg calcite, Mg-cal = High-Mg calcite, Mg = Magnesite, Unknown = minerals that have not yet been identified. The rim of the core is indicated on the left side.

Tuning Fork (TF) which oscillates at a given frequency Gold tip glued to the TF

Laser 633 nm(illumination at 60°)

x

yz

far field near + far field

Fig. 1

Fig. 2

Fig. 5: 3D plot of L Vs ν4 Vs ν1. The colour legend represents the Mg content.

AFM – TERS Raman on chalk

Two samples of chalk (Liége, Belgium) were flooded with MgCl2 for c. 1.5 and 3 years at reservoir conditions (130°C, 1 PV/day, 12 MPa effective stress) comparable to important hydrocarbon reservoirs in the North Sea. A micro-Raman spectrometer was used to observe mineralogical changes due to fluid injection: when Mg2+ ions bond with (CO3)2- ions, magnesite or Mg-calcite will grow as new mineral phases. Micro-Raman spectroscopy could identify the presence of magnesite along the core of the Long Term Test (1.5-years-test) up to 4 cm from the injection surface. In the Ultra Long Term Test core (3-years-test) the recrystallization of MgCO3 affected nearly the entire core (7 cm). However, the small grain size of newly grown minerals far below 1 micron (50-600 nm) asks for a different imaging and smaller laser spots.

Higher resolution chemical analyses were possible using TERS (Tip Enhanced Raman Spectroscopy), which represents a new frontier technique, combining Raman Spectroscopy with Atomic Force Microscopy (AFM). At the contrary of micro-Raman, this methodology requires considerable expertise and sample preparation, but provides an outstanding spatial resolution (~ 20 nm) and and an impressive Raman signal enhancement ~ 10 4-5.

Fig. 3

Fig. 4: comparison between and experimental (WDS/EDS) MgCO3 data. The first ones were obtained using L, ν4, and ν1 modes equations.

TOPOGRAPHY RAMAN MAP, displaying the intensity of the ν1 peak; 1 pixel = TERS spectrum.1 micron 1 micron

1086

Raman shift (cm-1)

Inte

nsity

We graphed the Raman peaks positions Vs the Mg content (Fig. 2, 3), in order to obtain the equations (Fig. 4) that link the composition to the Raman bands positions: L, ν4, ν1 modes behave quite linearly (0.7 - 0.85 R2, with more than 100 data points).

MgCO3 (mol%)) MgCO3 (mol%))

ν 4 m

ode

(cm

-1)

ν 1 m

ode

(cm

-1)

MgCO3 (mol%))

sam

ples

MgCO3 (mol %))

L m

ode

(cm

-1)