Authigenic Green Mica in Interflow Horizons within Late Cretaceous Deccan Volcanic Province, India and Its Genetic Implications

: Green authigenic mica, i.e., celadonite, is commonly associated with submarine alteration of basic igneous rock. However, very few studies have reported the formation of celadonite under nonmarine conditions. An integrated study involving field investigation, petrography, mineralogy, and mineral chemistry highlighted the origin of celadonite in two clay ‐ rich horizons (green boles) of the Late Cretaceous Deccan volcanic province. Within the Salher green bole, the celadonite oc ‐ curred as the dissolution and alteration of plagioclase, volcanic glass, and pore ‐ filling cement. In the case of the Pune green bole, the celadonite was formed by the alteration of plagioclase, pyroxene, and precipitation as film within intergranular pores, along with zeolite. The celadonite in the Salher green bole exhibited slightly lower K 2 O and Fe 2 O 3 and higher Al 2 O 3 than in the Pune. The mineral chemistry of the former showed a composition closer to ferro ‐ aluminoceladonite. Although the min ‐ eral chemistry of celadonite overlaps with glauconite, the distinct 10 Å and 15 Å reflections in XRD, euhedral lath and honeycomb morphology under SEM, and characteristic absorption bands in VNIR spectroscopy (0.4–2.5 μ m) and FTIR spectroscopy (400–4000 cm − 1 ) identified celadonite and Fe ‐ smectite within green boles. The green boles were formed either by the alteration of a volcani ‐ clastic deposit in local pools of water or by the in situ alteration of the fragmentary flow top. The present study is significant due to the occurrence of celadonite in a nonmarine environment, as it otherwise forms under submarine conditions. S.B..; formal analysis, P.S. E.L; investigation, P.S.; resources, S.B., K.P.; data curation, P.S., S.S.; writing—original draft preparation, P.S., S.B.; writing—review and editing, K.P., E.L..; visualization, S.B., E.L., K.P.; supervision, S.B., K.P. E.L; project administration, S.B.;

Several studies have highlighted the occurrences of authigenic clay-rich red interbasaltic layers (i.e., red bole) within a continental flood basalt province. Red bole mainly contains smectite along with hematite [29][30][31][32][33][34][35][36][37][38][39]. However, very few cases discussed the origin of an authigenic green silicate-rich layer within volcanic rocks. The green clay-rich interbasaltic horizons between lava flows in the Late Cretaceous Deccan volcanic province in India are known as green boles [23,32,40,41]. Green boles are relatively less common than authigenic clay-rich red interbasaltic layers (known as red boles). Green boles occur locally as a lenticular bed a few centimeters to a meter thick and as centimeter-scale cavityfilling deposits. VNIR spectra revealed the presence of nontronite and celadonite in green boles (e.g., [23]). Several studies have provided geochemical characteristics of green boles [23,32,40,41], but petrographic details are not yet available. However, factors controlling the formation of interbasaltic green bole beds and the source of potassium for the formation of authigenic celadonite are still a topic of discussion. This study aimed to understand the origin of authigenic green silicates within the green bole beds of the Deccan volcanic province based on textural, chemical, and mineralogical characteristics.

Geological Background and Field Evidence
The Late Cretaceous Deccan Volcanic Province is one of the largest mafic continental flood basalt provinces of the world, formed during a time interval of 69 to 63 Ma, extending over hundreds of kilometers across western and central India, with a present-day areal extent of ~0.5 × 10 6 km 2 (e.g., [42][43][44][45][46][47]). It includes interbasaltic clay horizons several centimeters to a few meters thick formed by the alteration of basalt and tuffaceous/volcaniclastic units. These interbasaltic bole horizons are extensively reported in western India [33,36,39,48,49]. The Western Deccan volcanic province has been classified into three major subgroups based on variation in the geochemical regime: the Kalsubai (~2000 m), Lonavala (~525 m), and Wai (~1100 m) subgroups, which are further divided into 12 stratigraphic units [50][51][52]. The Late Cretaceous Salher green bole is relatively older, and it belongs to the Kalsubai subgroup, whereas the Pune green bole belongs to the Wai subgroup.
A detailed investigation was undertaken on green boles exposed in two sections, Salher (20°44′19.14″ N, 73°56′32.34″ E) and Pune (18°30′28.36″ N, 73°47′2.64″ E), Maharashtra, India ( Figure 1). Green boles in Salher occur as 35 cm thick isolated pockets and lenses with ~2 m lateral extent at an elevation of 915 m above mean sea level (Figure 2A). They overlie the weathered flow top of the Deccan basalt. The flow top is highly altered, with secondary white and green minerals as vesicle fillings. The lava flow in Salher is dominated by compound lava flows. The upper contact of the green bole is sharp with the overlying basalt. Its lower contact is sharp, mostly convex downward, but irregular in places where it extends into the fractures on the underlying flow top (Figure 2A,B). The middle part of the green bole is homogeneous, showing distinct planar lamination. Three samples were collected from this location, including the fracture filling at the base of section (1), middle (2), and upper part (3) ( Figure 2B). All the samples included green and white vug filling minerals. The green bole bed in Pune occurs as an approximately 30 cm thick lens, at a 610 m elevation above the mean sea level. It occurs above the fragmentary flow top and exhibits gradational lower and sharp upper contacts with the basalt ( Figure  2C,D). Two samples were collected from this section.

Methods
Soft and friable samples were impregnated with a 1:1 mixture of East Coast Resin© (epoxy resin and hardener) before cutting and grinding for petrographic investigation under an optical microscope. The impregnated samples were kept in a hot-air oven at 50 °C overnight for hardening. The hardened samples were cut, ground, and polished using silicon carbide paper. The polished surface was mounted on a glass slide using a 2:1 mixture of cold setting Buehler© EpoThin TM 2 epoxy resin and Buehler© EpoThin TM 2 epoxy hardener. A Buehler© PetroThin sectioning system was used to cut the mounted samples. Thin sections were examined using a Leica DM 4500P polarizing microscope. Images were taken with a Leica DFC420 camera and processed using Leica Image Analysis software (LAS-v4.6). For clay mineralogy, XRD analysis was carried out using Cu Kα radiation (45 kV, 40 mA) and a nickel filter in an Empyrean X-ray diffractometer with a Pixel 3D detector at the Department of Earth Sciences, Indian Institute of Technology Bombay. Rock samples were powdered using an agate mortar and pestle. The <2 μm clay size fraction was separated by centrifugation after suspension and sedimentation of coarse nonclay fractions in Milli-Q water (18 MΩ) column [53,54]. Random preparations were made on a zero-background Si sample holder using <2 μm clay separate and scanned for 4 to 70° 2θ, with a scan speed of 43 s/step and a step size of 0.013° 2θ. Additionally, slow scans were performed in the region of 59 to 63° 2θ (0.013° 2θ step size and 148 s scan step time) to characterize 060 reflections of the clay minerals. An oriented clay mount of <2 μm clay size fraction was prepared on glass slides [54]. The oriented mounts were scanned from 4 to 30° 2θ with a 0.013° 2θ step size for qualitative analysis of clay minerals. They were analyzed each time after drying, glycolation (with ethylene glycol), and heating at 400 °C and 550 °C, using the same instrumental settings [55]. Qualitative identification of mineral phases was carried out using Panalytical© HighScore™ software. All three samples of Salher and two samples of Pune green bole were analyzed under XRD. XRD patterns of one sample each from the Salher (SL) and Pune (PN) green boles are presented in this paper due to their similarity in mineralogical composition. For morphological investigation, fresh rock chips of 5 mm × 2 mm dimension were mounted on the stub, and a platinum coating (15 nm) was applied on the sample with a sputter coater to avoid charging. The samples were analyzed under a JEOL JSM-7600F FEG-SEM at the Sophisticated Analytical Instrument Facility, Indian Institute of Technology Bombay.
Visible near-infrared (VNIR) spectroscopy of the three powdered samples from Salher (SL-1, 2, and 3) and one powdered sample from Pune (PN) was carried out using a FieldSpec ® 4 high-resolution spectroradiometer (Analytical Spectral Devices, Boulder, CO, USA) at the Space Applications Centre, Indian Space Research Organization, Ahmedabad. The reflectance spectra of the green bole were collected under laboratory conditions over a wavelength range of 0.35-2.5 μm (VNIR range), using a contact probe attached with the FieldSpec ® 4 high-resolution instrument. The contact probe was equipped with a highintensity light source; i.e., a 100 W halogen reflectorized lamp, with a fiber-optic input socket. For better results, a 30° angle was maintained between the halogen lamp and the fiber-optic input socket. Details of methodologies are provided in Bhattacharya et al. [56] and Naveen et al. [57]. The IR absorption spectra of <2 μm clay were recorded with a Bruker Vertex 80 FTIR system at the Sophisticated Analytical Instrument Facility, Indian Institute of Technology Bombay. For each sample, 256 scans were recorded (on the Bruker spectrometer) in the MIR region (400-4000 cm −1 ) with a resolution of 4 cm −1 . The IR absorbance conversion was made automatically with the OPUS 8.2.28 software package (Bruker Ltd., Billerica, MA, USA). All the samples were prepared using the KBr presseddisc technique. To make a KBr pellet, 0.5 mg of sample was dispersed in 200 mg of KBr, and the resulting mixture was placed in a 13 mm pellet die and pressed under room temperature. The pellet was then placed into a glass desiccant box with CaCl2 and heated in a furnace at 110 °C for 24 h. The peak deconvolution in the selected area of the OH-stretching region was done using OriginPro 2019b software. VNIR and FTIR spectroscopy were done for detailed characterization of clay minerals.
Mineral chemical investigations of different phases within the green boles were carried out using a Cameca SX 5 electron probe microanalyzer (EPMA) with an accelerating voltage of 15 kV, a specimen current of 20 nA, and a nominal beam diameter of 1 μm (peak: 10-20 s and background counting: 5-10 s) at the Department of Earth Sciences, Indian Institute of Technology Bombay. Minerals and synthetic phases were used as laboratory standards for the calibration of the EPMA data. Duplicate analysis of individual points shows an analytical error of less than 1%. All data points on the mineral of interest were selected using a combination of optical and backscattered electron image control. The EPMA data were ZAF corrected.

Petrographic and Morphological Description
Petrographic analysis of the green bole bed in Salher revealed volcanic lithic fragments, green authigenic clay, plagioclase, and vesicular volcanic glasses (up to 200 μm) as primary constituents ( Figure 3A). It also included mud clasts (up to 2 mm long), subrounded fine-grained green clasts (up to 500 μm), and highly altered basalt clasts with plagioclase laths (up to 100 μm). Pyroxene, apatite, and opaques occurred as minor constituents ( Figure 3A). Authigenic green minerals occurred as replacing plagioclase ( Figure 3B-D) and glassy phase within volcanic lithic fragments ( Figure 3E,F). It also occurred as fine precipitates within groundmass ( Figure 3A,E) and glass vesicles ( Figure 4B-D). Red clay minerals occurred as the replacement of basaltic fragments, as precipitate within intergranular pore spaces ( Figure  4A), and glass vesicles ( Figure 4B-D). In places, the red phases precipitated at the center of vesicles after the formation of green minerals ( Figure 4B  The green bole in Pune showed lithic and vitric grains of different sizes within finegrained groundmass. The long dimension of basalt clasts ranged from 500 μm to 2.5 mm, which included plagioclase laths, altering into yellow and green clay minerals ( Figure 5A). Olivine and pyroxene within basaltic clasts were mostly replaced by iron oxide and clay minerals, with a few relicts of original minerals ( Figure 5A). Plagioclase ( Figure 5B,C) and pyroxene ( Figure 5B) phenocrysts were variably altered to green clay minerals. Few pyroxene grains showed saw-edge-like dissolution features ( Figure 5E,F). Green clays also occurred as films around intergranular pores filled with zeolite ( Figure 5A) and as altered glasses ( Figure 5D). Some glass shards were altered to yellow clay minerals ( Figure 5D). Opaque phases were abundant within altered basaltic fragments and the matrix ( Figure  5A-C). Under the scanning electron microscope, the elongated lathlike morphology (less than 1 μm) ( Figure 6A) corresponded to green clay, and the honeycomb morphology corresponded to smectite ( Figure 6B).
The determination of the unit cell parameters, i.e., a, b, c, and β of the authigenic green mica structure, was carried out using experimental d(hkl) values. The b and a parameters of the structure were assumed to be 6 × d(060) and b/√3, respectively [63,64]. The unit cell parameters c and β were calculated from the d values of the 112 and 112 reflections (3.0879 Å and 3.6361 Å for SL; 3.08523 Å and 3.6305 Å for PN) using the formulae given in the paper by Sakharov et al. [64] ( Table 1). The unit-cell parameters of the studied samples occurred within the ranges provided for celadonite in the literature [59,63], which corroborated the identification of the green authigenic minerals as celadonite.

Visible Near-Infrared Spectroscopy
Spectra of samples from the Salher green bole showed weak bands near 0.41 μm, 0.45, μm and ~0.50 μm; broad and shallow absorption bands near ~0.74 μm, ~0.94 μm, and 1.085 μm; a shallow asymmetrical absorption band at 1.41 μm; a sharp and deep asymmetrical band near 1.91 μm; and shallow peaks between 2.20 and 2.50 μm (Figure 9). The green bole of Pune showed similar reflectance spectra as well as more prominent ~0.50 μm, 1.41 μm, and 2.21 μm absorption bands compared to the Salher green bole.
The reflectance spectra of the green bole samples revealed weak absorption features in the extended visible region, i.e., 0.40-1.10 μm due to crystal-field and charge-transfer effects in iron-bearing minerals, viz. celadonite and smectite [22,65]. The absorption features between 1.

Mineral Chemistry
The EPMA data revealed considerable variation in major element compositions among the green celadonite minerals from the Salher and Pune green boles (  Table 2). The subrounded brown-orange vesicular glasses ( Figure 4B-D) F and  4A,B). The Na2O content of plagioclase grains varied from 3.99 to 11.51 wt %. The zeolite was Ca-rich, with SiO2 ranging from 58.66 to 72.59 wt %, Al2O3 ranging from 11.11 to 16.82 wt %, and CaO varying from 4.85 to 7.80 wt %. The plagioclase composition was consistent with labradorite-bytownite and albite (av. albite). The authigenic zeolites belonged to the Ca-rich variety, similar to heulandite. X-ray mapping revealed higher contents of Si and K ( Figure 11B,C) and lower contents of Fe and Mg of the celadonite than in the associated red clay (Figure 11D,E).
Green authigenic clay; i.e., celadonite, within the Pune green bole showed a negligible variation in mineral chemistry ( Table 2). The celadonite showed a range of K2O from 8.83 to 9.76 wt %, Al2O3 from 5.13 to 6.98 wt %, MgO from 4.37 to 5.97 wt %, and Fe2O3 (total) from 16.11 to 19.53 wt %. The Fe-smectite clay showed SiO2 ranging from 41.04 to 45.52 wt %, Al2O3 from 10.86 to 12.00 wt %, Fe2O3 (total) from 21.61 to 24.98 wt %, MgO from 9.07 to 14.13 wt %, and CaO less than 5 wt %. X-ray mapping showed the alteration of plagioclase into celadonite along cleavage and weak planes in green bole samples from Pune ( Figure 12). All EPMA values were normalized to 100 wt % on an anhydrous basis to present the data in cross-plots. Celadonite data of Salher and Pune were compared with the published data reported from continental celadonite [5,14] using different symbols in Figure 13. The significant results of different cross-plots reveal the following significant aspects.
(a) A cross-plot of Al2O3 vs. Fe2O3 (total) showed two clusters for Salher and Pune celadonites. Salher celadonite showed higher Al2O3 and slightly lower Fe2O3 (total) contents than Pune celadonite ( Figure 13A). The Al2O3 content, varying from 9.25 to 12.28 wt %, caused a scattering of data points for Salher.
(b) The K2O vs. Fe2O3 (total) cross-plot formed two clusters for Salher and Pune celadonites ( Figure 13B). Celadonites from Salher showed a lower content of K2O, ranging from 7.90 to ~9.00 wt %. while those from Pune exhibited a consistently high content of K₂O (average 9.00 wt %).
(d) A cross-plot between Si 4+ and interlayer cations M⁺ (i.e., K⁺ Ca 2 ⁺ Na⁺) showed more than 3.85 Si⁴⁺ atoms per formula unit (pfu) in most celadonite data of Salher and Pune ( Figure 13D). A few celadonites from Salher showed relatively lower content (3.77-3.89 atom pfu). The interlayer cation ranged from 0.75 to 0.82 atom pfu in celadonite from Salher, and from 0.85 to 0.96 atom pfu in celadonite from Pune.
(e) A cross-plot between octahedral (VI)Al 3+ and interlayer cation (M⁺) showed two clusters for the celadonite data of Salher and Pune ( Figure 13E).

Mineralogical and Chemical Characteristics of the Celadonite in Green Boles
The green authigenic minerals in the Salher and Pune green bole were Fe-and Kcontaining dioctahedral mica. Mineral chemistry of celadonite overlapped with glauconite in several cases [6,59,89,90]. Therefore, the presence of celadonite in the green boles needed to be confirmed by integrating data using an optical microscope, scanning electron microscope, mineral chemistry, X-ray diffractogram, visible near-infrared spectroscopy, and FTIR spectroscopy.
The celadonite within the studied green boles formed primarily by the dissolution and precipitation mechanism of plagioclase ( Figure 3B-D) and glassy phase within volcanic lithic fragments ( Figure 3E,F), and as vesicle and intergranular void-filling. However, glauconite is mostly pellet-shaped, formed either by the alteration of fecal pellets, K-feldspar, and quartz, and as precipitates within the bioclasts [88]. Glauconite, containing more than 8 wt % K2O and more than 25 wt % Fe2O3 (total) characterized the highly evolved variety, exhibiting lamellar morphology [3,[90][91][92], whereas, the studied green clay; i.e., celadonite, showed fine (less than 1 μm) euhedral lathlike morphology, which was similar to the celadonite morphology reported in previous studies [3,14,27]. The interlayer cation is one of the criteria to distinguish between mica; i.e., celadonite (interlayer cation more than 0.85 a.p.f.u.), and interlayer deficient mica; i.e., glauconite (interlayer cation less than 0.85 a.p.f.u.) [85]. Although the interlayer cation content of the Pune celadonite exceeded 0.85 a.p.f.u., it varied between 0.75-0.82 a.p.f.u in the case of the Salher celadonite. Therefore, the interlayer cation of the green phase in the Pune green bole satisfied the composition of celadonite, but it was slightly lower in the case of the Salher celadonite. The green clay from both Salher and Pune showed low Al tetrahedral substitution (less than 0.2 a.p.f.u.), which was characteristic of celadonite [1,59,60,83,92]. Further, several studies indicated that the interlayer cation content may be less than 0.85 in the case of aluminoceladonite [59,62,63,83,86]. Therefore, the composition of the primary green phase in the case of the Salher green bole was closer to aluminoceladonite, whereas the same in the case of the Pune green bole was closer to celadonite. X-ray diffraction parameters of the studied samples exhibited sharp hkl reflection, split-up 003 and 022 peaks, and d060 spacing less than 1.51 Å, which was coherent with the XRD data of celadonite reported in the literature. The 1.51 Å d060 value was given as a discriminating value by the AIPEA nomenclature [80]. Unlike celadonite, glauconite exhibits reduced hkl reflection and broader basal peaks than celadonite, as well as merged 003 and 022 peaks and d060 spacing less than 1.51 Å [2,3,11,14,[58][59][60][61]84]. Therefore, the X-ray diffraction parameters ruled out glauconite and confirmed the presence of celadonite in both green boles. The equally strong metal-OH combination band at ~2.30 μm and 2.35 μm and the strong slope between 1.00 μm and ~2.10 μm under VNIR spectra corroborated the presence of celadonite. In FTIR, celadonite showed sharp stretching and bending vibrations, with characteristic absorbance in OH-stretching region [81][82][83].
Celadonite composition differed in the green boles of Salher and Pune. The composition of celadonite of the green boles in these places and those reported from other places [5,13,14] were compared ( Figure 13). Both Salher and Pune celadonites exhibited higher Al2O3 and slightly lower Fe2O3 (total) than those reported by Baker et al. [5] and Weiszburg et al. [14]. However, some of the published data of Weiszburg et al. [14] overlapped with the Fe2O3 wt % of the studied celadonites. The K2O and MgO wt % of the studied celadonites overlapped with most of the published data of Baker et al. [5] and Weiszburg et al. [14]. A comparison of the studied celadonites with ferroceladonite and ferroaluminoceladonite data of Li et al. [13] revealed that the Salher celadonite showed lower interlayer cation and higher octahedral Al 3 ⁺ than ferroceladonite, and overlapping octahedral Al 3 ⁺ with ferroaluminoceladonite. Celadonite from Pune partially overlapped with the ferroceladonite data of Li et al. [13]. In addition, the Salher celadonite showed higher Al (total) and lower Fe (total) compared to typical celadonite, and lower Al (total) and higher Fe (total) compared to aluminoceladonite; whereas the Pune celadonite showed a slightly lower value of Fe(total) and a slightly higher value of Al (total) than typical celadonite [59,62,63,83]. The above discussion indicates that the Salher celadonite had a composition closer to ferroaluminoceladonite, and the Pune celadonite was closer to typical celadonite composition.

Origin of Green Authigenic Celadonite under the Subaerial Condition
Celadonite formation is primarily associated with the submarine alteration of volcaniclastic and intermediate to mafic igneous rocks. However, recent studies suggested that celadonite can also form in subaerial conditions [4,5]. Unlike celadonite, glauconite is found exclusively in marine deposits, and is extremely rare in the continental setting [90]. The essential elements required for the formation of celadonite are K, Mg, Fe, Al, and Si. The potassium content in the seawater facilitates the alteration of basalt on the seafloor at a low temperature [3][4][5][6][7][8][9][10][11]. Recent studies showed the formation of this mineral in subaerial conditions as well [4,5,14,15]. It may form along with Fe-smectite during the intermediate stage of alteration of basalt into the soil [4]. Baker et al. [5] suggested the dissolution of glasses in basalt under a closed system as the primary source of K for celadonite formation in the continental setting. The alteration of volcanic rocks by a potassic hydrothermal fluid at the late stage of volcanic activity also forms celadonite [93].
Petrographic investigation of the studied samples demonstrated the formation of celadonite by the partial dissolution of plagioclase and almost complete alteration of pyroxene and olivine along with volcanic glasses in the Salher green bole. Mg, Fe, and Al were released to the pore water because of the dissolution of plagioclase and mafic minerals, whereas potassium was supplied by high-K2O-containing volcanic glasses. As illustrated by elemental mapping, the alteration of plagioclase into celadonite proceeded with the addition of K, Fe, and Mg ( Figure 12D-F) and removal of Al ( Figure 12C), with iron oxyhydroxide as a byproduct. Celadonite usually forms in slightly reducing conditions as the altering fluids replace the original substrate [3]. The reducing pore water promoted the substitution of Fe into the silicate structure, rather than forming a separate oxyhydroxide/oxide phase [4,5]. The d003 peak position of celadonite provided an estimation of the oxygen fugacity under which it formed [7,9]. The d003 spacing for celadonite in green boles was 3.31 Å, indicating an oxygen fugacity slightly lower than the hematite-magnetite buffer.
The occurrence of flaky smectite in a few intergranular spaces and within vesicles of glass suggested subsequent enrichment of Fe and Mg under low silica activity (e.g., [4]). The formation of brown smectite within the voids of green celadonite indicated precipitation from a more reducing pore fluid [17,94]. High K2O and SiO2 contents in brown-orange and yellow glasses suggested a different source. Contemporaneous volcanic activity to the north possibly supplied felsic ash periodically during the breaks of Deccan Basalts flows (cf. [95,96]). A similar condition was reported in the Grande Ronde Formation, and in sedimentary interbeds between the flows of overlying Columbia River Basalts in central Washington and Picture Gorge subgroup of the Columbia River basalt, Oregon, where anomalously high K contents in paleosols within these basalts corresponded to felsic ash input [29,97]. The alteration of calcic plagioclase; i.e., labradorite into zeolite (heulandite), suggests increased temperature caused by the overlying flow [98][99][100]. Zeolites in veins and vesicles could form by the late-stage hydrothermal fluid circulation [99,101]. The intergranular pores and veins, filled partially by celadonite, were subsequently occupied by zeolite in the Pune green bole, reflecting rising temperatures during the overlying flow and hydrothermal activity ( Figure 5A). The hydrothermal alteration process was responsible for the devitrification of volcanic glass, followed by the hydration and crystallization of the smectite (cf. [102]). The saw-edge-like dissolution feature in pyroxene from the Pune green bole (indicating congruent dissolution) is a common product of weathering in buried truncated volcanic paleosols [103]. Physical, textural, mineralogical, and chemical characteristics of the Salher and Pune green boles showed different characteristics. Field features in Salher suggested the alteration deposition of weathered volcaniclastic in deposit on a local pool of standing water on the undulatory flow top of basalt, whereas the Pune green bole was formed by the in situ alterations of the fragmentary flow top of basalt. The slightly reducing conditions of the formation of the green bole facilitated the retention of Fe 2 ⁺ in the celadonite structure in the subaerial condition in both cases.
The formation of nonmarine celadonite within green boles is important for understanding the paleoenvironmental implications of ancient green mica. During alteration of basic rock in a marine environment, potassium is easily available in the seawater. However, the formation of celadonite in a continental setting requires an alternate source of potassium. One possible source of potassium is felsic volcaniclastic input [14][15][16]. Understanding the origin of green mica is crucial for exploring the paleoenvironmental condition of these minerals on the Martian surface [20][21][22][23][24][25][26][27][28].

Conclusions
Detailed petrographical, mineralogical, and geochemical studies made it possible to highlight the following conclusions: 1. Celadonite-bearing green boles were investigated in two study areas of the Late Cretaceous Deccan volcanic province; i.e., Salher and Pune. Green boles were formed by the subaerial alteration of basaltic rocks during the lulls of the eruption of basaltic flows. 2. Combined mineralogical and chemical investigation, as well as VNIR and FTIR spectroscopy, identified the green authigenic mineral as celadonite. 3. Celadonite from the Salher green bole contained slightly lower K2O (7.90 to ~9 wt %), Fe2O3 (total), and MgO, and higher Al2O3 contents than those from Pune. The mineral chemistry indicated Salher celadonite was closer to ferro-aluminoceladonite, whereas the Pune celadonite, with a narrow range of K2O, was closer to typical celadonite composition. 4. The green bole of Salher formed as volcaniclastic deposits in local pools of water on the flow top of basalt under slightly reducing conditions, whereas the green bole in Pune formed by the in situ alteration of the flow top of basalt. The sources of Mg, Fe, and Al of celadonite were linked to the dissolution of pyroxene, plagioclase, and metastable interstitial glasses, respectively; whereas K and Si were contributed by the felsic input in Salher celadonite. K and Si contents of celadonite from Pune were possibly sourced by potassic hydrothermal fluid or groundwater flow. 5. Therefore, this study is important to provide further insight on the formation of authigenic celadonite in a nonmarine environment. Acknowledgments: The authors are thankful to their host institutes for the infrastructure support. S.B. and K.P. acknowledge financial support from the Space Applications Centre (ISRO) for Project 467 STC0254, entitled "Physico-chemical conditions of formation of bole beds within Deccan basalt for 468 Martian analogue". PSis thankful to CSIR for the fellowship.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations were used in this manuscript: VNIR Visible near-infrared spectroscopy FTIR Fourier-transform infrared spectroscopy MIR Middle-infrared region spectroscopy FEG-SEM Field emission gun-scanning electron microscope XRD X-ray diffraction analysis