Experimental crystallization of analcime zeolite from clay and feldspar precursors

Abstract

Analcime is a common authigenic mineral in siliciclastic rocks and has widespread industrial applications in catalysis and ceramic production. Although its formation from volcanic glass has been extensively studied, the role of clay minerals and feldspar precursors in its genesis remains poorly constrained. This study investigates the hydrothermal formation of analcime and associated zeolites from arkosic sandstone of the Oligocene–Miocene Al Wajh Formation, northwest Saudi Arabia using Na₂CO₃ solutions (0.1 M and 0.5 M) at 80, 150, 200, and 250 °C for 336 h. The results indicate that analcime crystallized through the dissolution of feldspars, kaolinite, smectite, and illite, forming cubic to trapezohedral crystals. Mordenite and chabazite formed as minor zeolite phases at ≥ 200 °C, associated with higher silica activity and the breakdown of silica gel, smectite, and illite. Analcime precipitated as grain-coating, replacive, and pore-filling cement, with intercrystalline pores reaching up to 9.7 µm in size. The synthesised analcime may enhance the mechanical stability of the sandstone framework by reducing susceptibility to compaction. Furthermore, if exposed to acidic meteoric waters, dissolution of analcime could generate secondary intracrystalline porosity that could significantly improve the overall reservoir quality of the sandstone.

Introduction

Over the last few decades, zeolites have attracted increasing global attention due to their broad economic and scientific significance^1–7. They are widely used in numerous industrial applications, including pollution control, soil conditioners and fertilizers, gas purification, petroleum production etc.^8,9. Therefore, zeolites are often regarded as the “mineral future” in many countries, where significant efforts are underway to explore, develop, and utilize these resources effectively^10,11. Zeolites are naturally-occurring, hydrated aluminosilicate minerals that occur in a variety of igneous, low-grade metamorphic and sedimentary rocks, with over 20 known mineral species^4,12. However, the most widely known species of zeolites are analcime, clinoptilolite, chabazite, erionite, mordenite, and phillipsite⁸. Analcime occurs in clastic sedimentary rocks spanning a wide range of geological ages, lithologies, and depositional environments in basins worldwide^3,6. For instance, analcime has been documented in the Mesozoic Stuttgart Formation of northern Germany⁷; the Eocene Shahejie Formation in the Bohai Basin of northeastern China¹³; the Eocene Green River Formation of Utah, USA¹⁴; and the Pleistocene strata of the Lake Lewis Basin in central Australia¹⁵ etc. In addition, analcime can form during steam injection for heavy oil recovery and as a result of hot fluid circulation in geothermal systems¹⁶. Although basins reporting analcime occurrences are not particularly extensive in number or scale, its presence within reservoir rocks can significantly influence reservoir quality by modifying porosity and permeability^17,18.

Diagenesis is widely recognized as the primary mechanism responsible for the formation of authigenic minerals in siliciclastic rocks^19–49. Analcime is commonly reported to form through the diagenetic transformation of clinoptilolite, which in turn forms from hydration of abundant volcanic glass during shallow burial diagenesis. Iijima² noted that this transformation occurs at temperatures ranging between 84 and 91 °C. Numerous studies have therefore focused on investigating the origin of analcime through the transformation of clinoptilolite^1,2,50–52.

In addition, clay minerals and feldspars are frequently cited as potential precursors for authigenic analcime^6,53. However, the mechanisms governing analcime formation from clay and feldspar precursors remain poorly constrained. This knowledge gap is chiefly because the reaction commonly occurs at depths > 1000 m⁶, where precursor phases are commonly consumed during diagenesis and thus become difficult to identify in the rock record.

To address this limitation, the present study employs controlled hydrothermal-reactor experiments to investigate the crystallization of analcime from clay and feldspar precursors using an arkosic braided-river sandstone from the Oligocene–Miocene Al Wajh Formation (NW Saudi Arabia; Fig. 1A–C). The sedimentology and early diagenetic features of the Al Wajh Formation have been described in detail by Bello⁵⁴. The specific objectives of the present study are to:

1. Characterize the mode of occurrence and textural evolution of diagenetic analcime derived from clay and feldspar precursors.

2. Investigate the formation of associated zeolite phases during the hydrothermal experiments.

3. Evaluate the influence of Na concentration on the formation and thermal stability of the synthesized analcime.

Fig. 1.

(A) Regional view of the Midyan Basin, northwest Saudi Arabia. (B) Location map of the study area showing the Al Wajh outcrop investigated in this study (blue arrow). (C) Stratigraphic column illustrating the age and depositional environment of the Al Wajh Formation. This map was prepared using Adobe Illustrator software, version 22.1 (https://adobe-illustrator-cc.software.informer.com/22.1/).

Geological background of the Al Wajh sample

The present study utilized an outcrop, sandstone sample from the Al Wajh Formation, situated in the Midyan Basin, northwest Saudi Arabia (Fig. 1A–C). The sampled unit has been interpreted as a braided fluvial sandstone⁵⁴. The Al Wajh Formation represents the oldest syn-rift succession in the Midyan Basin, deposited during the Early Oligocene to Early Miocene (Fig. 1C)The sedimentology, depositional facies, diagenesis, and reservoir characteristics of the formation have been extensively documented in previous studies^54–57. Overall, the Al Wajh Formation is interpreted to have been deposited in predominantly alluvial, fluvial, and lacustrine settings, locally influenced by marginal marine conditions^56,58.

The Al Wajh Formation was selected for this study because of its arkosic composition, the presence of clay minerals, and its abundance of plutonic and volcanic fragments⁵⁴. Additionally, the formation provides an excellent natural laboratory for investigating early diagenetic processes and their impact on reservoir quality, as it has undergone only shallow burial diagenesis⁵⁴.

Materials and methods

Analytical procedures

Bulk and clay-fraction X-ray diffraction (XRD) analyses were conducted on both pre- and post-reacted samples of the studied Al Wajh sandstone primarily to achieve qualitative identification of mineral phases and to document and mineralogical changes induced during the experiments, following the established criteria for sandstone and clay mineral analysis outlined by Bello^59,60 and Salisu^61,62. Due to the inherent limitations of XRD for precise mineral quantification, particularly in heterogeneous sandstones and clay-rich fractions, the analyses were intentionally focused on mineral peak identification rather than absolute quantification.

Bulk samples were gently crushed to a fine powder (< 63 μm) and analyzed as random mounts, whereas the clay fraction (< 2 μm) was separated by sedimentation after dispersion in deionized water, following the procedure described by^60,63. XRD measurements were performed using a Malvern Empyrean PANanalytical diffractometer operated at 45 kV and 40 mA. Random powder mounts were scanned over a 2θ range of 4–70° using CuKα radiation, with a step size corresponding to a counting time of 8.7 s per step. Oriented clay mounts were prepared and analyzed under air-dried conditions to enhance basal reflections and facilitate clay mineral identification.

Mineral phase identification was performed using HighScore Plus software (v. 4.9) equipped with the ICDD PDF-4 + (2024) mineralogical database. Quantitative phase analysis of bulk samples was carried out using the Rietveld refinement method implemented in HighScore Plus. The refinement procedure included background modeling, refinement of scale factors, zero shift, unit cell parameters, and peak profile parameters. Phase weight percentages were calculated from the refined scale factors using the crystallographic models incorporated in the database.

Refinement quality was evaluated using standard agreement indices (e.g., Rwp, Rp, and goodness-of-fit), and refinements were accepted once convergence was achieved without systematic misfit in the difference plots. Although Rietveld refinement provides quantitative phase estimates with high numerical precision, the practical analytical uncertainty of quantitative XRD analysis in heterogeneous sandstone samples is typically on the order of ± 1–3 wt.% for major phases and higher for minor phases, owing to factors such as sample preparation, grain statistics, preferred orientation, and microstructural effects. Thus, the mineral abundances are interpreted within this expected uncertainty, and small differences between samples are discussed in terms of overall trends rather than precise numerical variations. Clay mineral abundances were assessed qualitatively to semi-quantitatively based on relative basal peak intensities from oriented mounts, following established XRD clay mineral analysis protocols⁶³. Owing to the effects of preferred orientation, variable crystallinity, mixed-layering, and interlayer hydration, clay mineral proportions are expressed as relative abundances, and no attempt was made to derive fully quantitative clay mineral contents. XRD interpretations were integrated with petrographic observations and SEM–EDX analyses to strengthen mineral identification and mitigate uncertainties inherent in clay mineral XRD analysis.

Standard petrographic thin section for the pre-reacted Al Wajh sandstone was prepared and studied using an Olympus BX53F petrographic microscope to investigate the primary, framework composition, pore-filling matrix, and secondary diagenetic cements. In addition, the petrographic analysis was chiefly conducted to complement the XRD data and to establish the average sandstone composition. Compositional point counting (using 300 points) was conducted on the pre-reacted sample using a Petrog software package. Additionally, the software was employed to establish the textural characteristics of the pre-reacted sample (e.g., grain size and sorting), by measuring the longest axes of 100 detrital quartz and undegraded feldspar gains.

Stub-mounted samples of both the reacted and unreacted sandstones of the Al Wajh Formation were investigated using a Zeiss Gemini 550 scanning electron microscope (SEM), which was fitted with an Aztec energy dispersive X-ray spectrometer (EDX). The samples were coated with 30 nm palladium-gold using Quorum QR150R sputter coater prior to running the SEM analysis and were analyzed at the voltage and current ranging between 2 and 20 kV and 100 and 2 nA, respectively.

Hydrothermal-reactor experiments

Hydrothermal-reactor experiments were conducted using a mechanically stirred autoclave at the Experimental Diagenesis Laboratory, Center for Integrative Petroleum Research, College of Petroleum Engineering and Petroleum Geoscience, King Fahd University of Petroleum and Minerals, Saudi Arabia. The starting material, Al Wajh sandstone, was divided into eight equal portions (approximately 20 g each). The samples were reacted with 0.1 M and 0.5 M Na₂CO₃ solutions at constant temperatures of 80, 150, 200, 250 °C for 336 h each (Table 1), following the experimental setup and procedures described by Bello^60,64.

Table 1.

Experimental conditions for the autoclave reactions.

Sample ID	pH (25 °C)	Solution (M)	Temperature (°C)	Final pressure (bar)	Experimental duration (hours)
SF1	11.45	0.1 Na2CO3	25
SF2	11.4	0.5 Na2CO3	25
AWA1	11.15	0.1 Na2CO3	80	71.6	336
AWA2	10.33	0.1 Na2CO3	150	94.4	336
AWA3	10.12	0.1 Na2CO3	200	110.2	336
AWA4	10.07	0.5 Na2CO3	250	150.4	336
AWB1	10.92	0.5 Na2CO3	80	70.6	336
AWB2	10.6	0.5 Na2CO3	150	95.2	336
AWB3	10.45	0.5 Na2CO3	200	113.3	336
AWB4	10.09	0.5 Na2CO3	250	148.5	336

Experiments were carried out in 500 mL reactor vessels containing 200 mL of solution, corresponding to a fluid–rock ratio of 10:1. The initial pressure was approximately 44 bar at room temperature; pressure increased progressively with temperature, and the final pressures attained during each experiment are reported in Table 1. All experiments were conducted under strictly closed-system conditions. Upon completion, the reactors were allowed to cool to room temperature (25 °C), after which final pH measurements were recorded (Table 1). Initial and final pH values of the solutions are provided in Table 1.

Results

Pre-reacted sandstone composition

The Al Wajh sample utilized in the study is a medium-grained and well sorted sandstone, with a mean grain size of 0.43 mm. Petrographic point-count analysis indicates that the framework is dominated by detrital monocrystalline quartz (40.5 vol%), microcline (4 vol%), orthoclase (10.1 vol%), and plagioclase (5 vol%). Mica is present as minor biotite (1 vol%) and muscovite (1 vol%). Lithic fragments occur as felsic plutonic (1 vol%) and siltstone (1 vol%) rock fragments.

The pore-filling clay matrix occurs in trace amounts (0.5 vol%), whereas grain-coating clay reaches up to 7 vol%; both are predominantly smectitic in composition based on SEM–EDX analysis. Kaolinite is present mainly as a grain-replacive cement (3.5 vol% of total rock volume), replacing detrital feldspar and felsic plutonic rock fragments.

According to the classification scheme of Folk (1980), the pre-reacted sandstone is compositionally classified as arkosic, with an average framework composition of Q₆₅F₃₂R₃ (where Q = quartz; F = feldspar; and R = rock fragments).

Bulk XRD analysis of the pre-reacted Al Wajh sandstone indicates that the dominant framework minerals are quartz (67 wt%), orthoclase (11 wt%), microcline (9 wt%), and albite (11 wt%). Accessory phases include muscovite (1 wt%), chlorite (1 wt%), kaolinite (1 wt%), and dickite (trace amounts).

Post-reacted sandstone composition

The mineralogical composition of the post-reacted Al Wajh sandstone was established using both bulk and clay-fraction XRD analyses (Fig. 2A–D). Bulk XRD results indicate that the framework minerals of the treated samples are dominated by quartz, albite, and orthoclase (Fig. 2A–D). Clay minerals consist mainly of mixed-layer illite–smectite, kaolinite, illite, and chlorite (Fig. 2C, D).

Fig. 2.

Bulk and clay-fraction XRD results for pre- and post-treatment Al Wajh sandstone samples. (A) Bulk XRD patterns of samples treated with 0.1 M Na₂CO₃, showing detrital quartz, albite, orthoclase, and the development of authigenic analcime. (B) Bulk XRD patterns of samples treated with 0.5 M Na₂CO₃, also indicating the presence of quartz, albite, orthoclase, and authigenic analcime. (C) Clay-fraction XRD of samples treated with 0.1 M Na₂CO₃, showing kaolinite, montmorillonite, illite, chlorite, and the formation of analcime. Note the disappearance of the primary kaolinite peak and the emergence of analcime peaks at 150, 200, and 250 °C. (D) Clay-fraction XRD of samples treated with 0.5 M Na₂CO₃, showing similar mineral assemblages and changes. The kaolinite peak disappears and analcime peaks appear at the same elevated temperatures.

During experiments conducted at 150, 200, and 250 °C, mixed-layer illite–smectite, chlorite, and kaolinite were progressively consumed, as evidenced by the reduction or disappearance of characteristic kaolinite and chlorite reflections in the clay-fraction diffractograms (Fig. 2C, D).

Analcime represents the principal authigenic zeolite phase formed at temperatures between 150 and 250 °C (Fig. 2A–D).

Crystal shape and size of analcime

Four distinct crystal morphologies were observed in the synthesized analcime: (1) spherical (Fig. 3A); (2) cubic (Fig. 3B); (3) trapezohedral (Fig. 3C, D); and (4) interpenetration twins (aggregates; Fig. 3E, F) crystals. The spherical analcime crystals (Fig. 3A) occur mainly as grain coatings and range in diameters from 9.6 to 14.7 µm (average 11.1). These crystals are most well developed at 200 ^○C. In contrast, cubic analcime crystals are smaller, with diameters ranging from 3.4 to 5.6 µm (average 4.4 µm). The trapezohedral crystals are typically euhedral and occur as discrete individuals (Fig. 3C, D). The diameters across the trapezohedral crystal tops range from 3.1 to 5.5 µm (average 4.3), while individual crystal faces measure between 1.2 to 3.9 µm (average 2.3 µm). This morphology was observed only in experiments conducted at 200 °C. Aggregate analcime crystals commonly consist of intergrown cubic and polygonal forms, often displaying interpenetration twinning (Fig. 3E, F). The cubic aggregates range from 1.6 to 4.5 µm in diameter (average 2.8 µm), whereas individual cubic crystals within these aggregates measure 0.4 to 1.0 µm (average 0.7 µm). Polygonal crystal aggregates (Fig. 3F) are larger, with diameters ranging from 21.6 to 23.9 µm (average 22.9 µm).

Fig. 3.

SEM images showing the various shapes of analcime crystals, formed during the experiments. (A) Spherical analcime crystals occurring mainly as grain-coating cement. (B) Well-developed cubic analcime crystals. (C, D) Trapezohedral analcime crystals. (E) An aggregate of cubic analcime crystals. (F) An aggregate of polygonal analcime crystals.

Mode of occurrence of analcime

The studied analcime crystals occur in 4 main types: (1) grain-replacive; (2) clay-replacive; (3) grain-coating; and (4) pore-filling cements.

Grain-replacive analcime

Crystals of analcime often occur as diagenetic replacements of detrital grains, including K-feldspar and albite (Fig. 4A, B). The replacement commonly occurs along the edges of partly to pervasively dissolved feldspar grains (Fig. 4C) or within their interior parts (Fig. 4A). SEM analysis has shown that both the replacement of albite and K-feldspar grains by analcime seems to occur in a dissolution–precipitation fashion (Fig. 4A–E). However, SEM observations reveal that the dissolution of albite resulted in the formation of amorphous aluminosilica gel, around and upon which analcime precipitated (Fig. 4E). Other minor zeolites formed include mordenite and chabazite (Fig. 4F). Occasionally, a rod-like and fibrous mordenite nucleated on a silica gel (Fig. 4F). Additionally, the spherical silica gel develops on severely etched feldspar grain (Fig. 4G), and, in some other instances, authigenic smectite (montmorillonite) grows perpendicular to the etched feldspar surface (Fig. 4H).

Fig. 4.

Grain-replacive analcime, analcime, aluminosilica gel, and smectite. (A) Trapezohedral analcime crystal replacing the centre of a pervasively dissolved detrital K-feldspar grain. (B) Incomplete replacement of albite grain by analcime. (C) Well-developed, cubic analcime crystals replacing the edges of an etched K-feldspar grain. (D) Replacement of albite by analcime (E) Replacement of albite by analcime and formation of silica gel from albite dissolution. (F) Development of silica gel from feldspar dissolution and formation of mordenite around the gel. (G) Precipitation of abundant spherical silica gel on a dissolved feldspar gain. (H) Formation of authigenic smectite around pervasively dissolved K-feldspar grain.

Clay-replacive analcime

Diagenetic analcime replaces kaolinite, illite, and smectite-rich clays (Fig. 5A–F). For instance, cubic, spherical, and cubic aggregate of analcime crystals developed in the vicinity of smectitic and illitic clays (Fig. 5A). Additionally, authigenic analcime commonly engulfs the clay they replace (e.g., kaolinite and smectite; Fig. 5B, C). Furthermore, SEM observations indicate that, at higher temperature, smooth, aggregate analcime crystals significantly consume kaolinite (Fig. 5B), smectite (Fig. 5C), and illite (Fig. 5D, E, F), with the illite crystals almost completely consumed during the replacement (Fig. 5E, F).

Fig. 5.

Clay-replacive analcime crystals. (A) Spherical and cubic analcime crystals replacing smectite and illite. (B) Smooth, well-developed aggregate of polygonal analcime crystals engulfing and replacing kaolinite. (C) An aggregate of analcime crystals rimming and replacing smectite. (D) Analcime aggregate replacing and engulfing illite. (E, F) Pervasive replacement of illite by aggregates of trapezohedral analcime crystals.

Grain-coating and pore-filling analcime

Spherical crystals of analcime grow on and around detrital quartz and feldspars (e.g., Fig. 3A). The crystals range in diameter from 9.6 to 14.7 µm, averaging 11.1. They provide partial to complete cover around the detrital grains. Occasionally, the spherical grain-coating crystals aggregate grow into the intergranular pore, thereby developing into a pore-filling cement (Fig. 3A). SEM observations reveal that the grain-coating analcime crystals replace smectitic, grain-coating clays (Fig. 3A).

Analcime occurs as a pore-filling cement. Pore-filling analcime consists of spherical, trapezohedral, and interpenetration crystals, filling and reducing primary intergranular porosity. Blocky, pore-filling aggregates of analcime crystals over 10 µm and often block pore throats. It has been observed that the pore-filling analcime crystals are more developed at high temperatures 200 and 250 °C.

Other minor zeolites synthesized

SEM–EDX analysis revealed the development of two additional, minor zeolites phases (mordenite and chabazite), particularly in experiments conducted at elevated temperatures between 200 and 250 °C (Fig. 6A–F; Table 2). Mordenite appears as acicular, needle-like crystals around smectitic clays and detrital feldspars (Fig. 6A, B), while chabazite forms euhedral, pseudohexagonal platelets arranged in distinctive stacked rosette-like aggregates in smectite vicinity (Fig. 6C, D). Individual mordenite fibres range in length from 10.7 to 44.8 µm (av. 23.3 µm) and in width from 0.3 to 0.6 µm wide (av. 0.4 µm). Both mordenite and chabazite developed in close association with detrital feldspar and clay minerals, often forming as replacive phases or and pore-filling cements. Chabazite platelets range in diameter from 7.8 to 36.4 µm (av. 18.9 µm), while the entire rosettes structures vary in thickness from 7.2 to 47.3 µm (av. 27.8 µm). The rosette morphology of chabazite coupled with its elemental composition are typical of those described by Welton¹⁶. Additionally, the chabazite is characterized by abundant intercrystalline porosity (Fig. 6C, D), with pore diameters ranging in size from 0.6 to 5.3 µm, averaging 2.6. While majority of the synthesized mordenite and chabazite occur as replacements of feldspar and clay minerals, mordenite and analcime grew on aluminosilica gels (Fig. 6E, F), formed from the dissolution of both feldspars and clay minerals.

Fig. 6.

SEM images for other minor zeolites synthesized. (A) SEM image showing the replacement of smectite by rod-like, fibrous mordenite. (B) Replacement of detrital K-feldspar by mordenite. (C, D) SEM images showing the formation of chabazite and mordenite via replacement of smectite. (E) Development of analcime and mordenite around silica gel. (F) Formation of mordenite around silica gel.

Table 2.

Semi-quantitative major oxide composition of the synthesized zeolites determined by SEM–EDX analysis.

Oxide	Analcime [AWA2; n = 19]		Analcime [AWA3; n = 16]		Analcime [AWA4; n = 3]		Analcime [AWB2; n = 6]		Analcime [AWB3; n = 4]
	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD
SiO2	52–70	63 ± 4	63–71	66 ± 2	63–64	64 ± 0	63–68	65 ± 2	63–70	67 ± 3
TiO2	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0
Al2O3	19–27	23 ± 3	18–23	21 ± 1	19–20	19 ± 0	18–21	20 ± 1	15–19	17 ± 2
FeO	0–4	1 ± 1	0–2	1 ± 1	2–2	2 ± 0	1–3	2 ± 0	2–3	3 ± 1
MgO	0–2	0 ± 1	0–1	0 ± 0	0–0	0 ± 0	0–1	0 ± 0	0–1	0 ± 0
CaO	0–6	1 ± 2	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0	0–3	1 ± 1
Na2O	6–18	12 ± 3	7–14	12 ± 2	14–14	14 ± 0	8–14	12 ± 2	6–15	10 ± 3
K2O	0–1	0 ± 0	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0	0–0	0 ± 0

Oxide	Analcime [AWB4; n = 4]		Mordenite [AWA4; n = 6]		Mordenite [AWB3; n = 4]		Mordenite [AWB4; n = 4]		Chabazite [AWB4; n = 5]
	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD
SiO2	65–68	67 ± 1	62–66	64 ± 2	36–70	52 ± 12	53–58	56 ± 2	45–60	50 ± 5
TiO2	0–0	0 ± 0	0–2	1 ± 1	0–1	0 ± 0	0–6	2 ± 2	0–1	0 ± 0
Al2O3	15–20	19 ± 2	3–15	9 ± 4	4–15	10 ± 4	5–13	10 ± 3	17–21	18 ± 2
FeO	1–5	2 ± 1	3–12	8 ± 3	4–8	6 ± 1	3–15	8 ± 4	3–9	6 ± 2
MgO	0–2	0 ± 1	2–12	7 ± 4	2–8	4 ± 2	4–11	7 ± 3	4–8	5 ± 1
CaO	0–1	0 ± 0	0–2	1 ± 1	0–1	0 ± 1	0–1	1 ± 0	2–19	12 ± 6
Na2O	8–13	10 ± 2	7–10	8 ± 1	15–44	25 ± 12	11–18	14 ± 3	4–6	5 ± 1
K2O	0–1	0 ± 0	1–6	2 ± 2	1–1	1 ± 0	1–2	1 ± 0	2–4	3 ± 1

Chemical composition of analcime

Major oxides

The major oxide compositions of the synthesized zeolites were determined using SEM–EDX analysis (Table 2). The major oxide data show clear compositional contrasts among analcime, mordenite, and chabazite. SiO₂ is the dominant oxide in all samples. Analcime displays consistently high and relatively uniform SiO₂ contents (mean 63–67 wt.%). Mordenite shows broader variation, particularly in sample AWB3 (36–70 wt.%, 52 ± 12 wt.%). Chabazite contains lower SiO₂ (50 ± 5 wt.%).

Al₂O₃ contents in analcime range from approximately 19 to 23 wt.% and are similarly elevated in chabazite (18 ± 2 wt.%). Mordenite exhibits lower and more variable Al₂O₃, with means around 9–10 wt.% in AWB3 and AWB4 (Table 2). TiO₂ is negligible in nearly all samples.

FeO and MgO contents distinguish the mineral groups. Analcime contains low FeO (generally 1–3 wt.%) and negligible MgO. Mordenite shows higher FeO (up to 8 ± 4 wt.%) and MgO (4–7 wt.%). Chabazite has intermediate FeO (6 ± 2 wt.%) and moderate MgO (5 ± 1 wt.%).

Marked differences are observed in Na₂O and CaO. Analcime is consistently Na-rich (10–14 wt.% Na₂O) and Ca-poor. Mordenite, especially AWB3, contains very high and variable Na₂O (25 ± 12 wt.%). Chabazite is characterized by high CaO (12 ± 6 wt.%) and comparatively low Na₂O (5 ± 1 wt.%). K₂O is generally low in analcime and slightly higher in mordenite and chabazite.

Porosity

Porosity within the synthesized analcime crystals is predominantly intercrystalline. This intercrystalline porosity is generally abundant and becomes more pronounced with increasing experimental temperature. In contrast, intracrystalline porosity, typically associated with crystal dissolution, is largely absent in the studied samples, as the analcime crystals underwent growth rather than dissolution under the experimental conditions. The intercrystalline porosity ranges in diameter from 0.2 to up to 9.7 µm, averaging 2.0. Aggregate cubic analcime crystals contain the least intercrystalline porosity, whereas the trapezohedral crystal shapes form the best intercrystalline porosity.

Discussion

Mechanisms of formation of zeolites

The formation of zeolites and feldspars in siliciclastic rocks, rather than clay minerals, is commonly attributed to elevated silica activity and high alkali-to-hydrogen ion (Na⁺/H⁺) ratios, and these conditions are typically favored by the presence of entrapped saline-alkaline water^1,6,12,61,65. Our experimental findings support this interpretation, indicating that zeolite formation from arkosic sandstone cemented by kaolinite and smectitic clays is strongly influenced by both temperature and the alkalinity of the reacting solutions.

Several mechanisms have been proposed to account for the origin of authigenic zeolites in sandstones, including (1) alteration of precursor minerals such as zeolites or volcanic glass; (2) crystallization from amorphous aluminosilicate gels, which form during the dissolution of detrital silicates; (3) direct precipitation from slightly to moderately saline, alkaline pore fluids; and (4) water–rock interaction reactions involving detrital silicates (e.g., feldspars, kaolinite, smectite, or illite) and saline-alkaline pore water^6,66,67. However, two primary mechanisms were responsible for the formation of zeolites observed in our study. The predominant pathway involved the dissolution of detrital feldspars and clay minerals (Figs. 5, 6), which released the essential elements such as Si and Al into the solution and resulted in subsequent precipitation of zeolites. While some of the Na⁺ required for zeolite formation may have been derived from the dissolution of albite, the majority of the Na⁺ was likely supplied by the Na-rich experimental fluid. A secondary mechanism involved the crystallization of zeolites from intermediate aluminosilicate gels (Fig. 6E, F), which formed during early stages of precursor breakdown under elevated temperatures and alkaline conditions.

Our study indicates that the formation of analcime occurred primarily through a dissolution–precipitation mechanism, rather than by pseudomorphic replacement. This is attributed to the structural incompatibility between the framework analcime and that of its precursor minerals. Detrital feldspars, including both potassium feldspar and plagioclase (albite), exhibited varying degrees of dissolution under experimental conditions (Fig. 4A–E), thereby providing the essential Si, Al, and Na required for analcime formation. Analcime with different crystal morphologies subsequently precipitated in close proximity to these etched feldspar grains (Fig. 4A–E), indicating the spatial association between feldspar dissolution and zeolite nucleation. In addition, analcime crystals were observed to grow at the expense of partially to completely dissolved kaolinite, smectite, and illite (Fig. 5A–F), suggesting that clay minerals served as important precursors by supplying essential framework elements such as Si, Al, and Na. Evidence from SEM (Fig. 5B) and XRD analyses (Fig. 2C, D) have shown that kaolinite is almost completely consumed in the formation of analcime. This is particularly evident in the XRD patterns, where the primary kaolinite peak disappears entirely at experimental temperatures of 150, 200, and 250 °C, which coincides with the formation of analcime at the same temperatures (Fig. 2C, D). Although the formation of analcime is strongly influenced by the dissolution of both feldspar grains and clay minerals, the formation of minor zeolite phases (i.e., mordernite and chabazite) appears to be more closely linked to the dissolution of smectite and illite (Fig. 6A–D). The relatively high concentrations of Mg, K, and Ca detected in the mordernite and chabazite (Table 2) further corroborate this interpretation, suggesting that the breakdown of smectite and illite contributed these cations to the fluid, thereby promoting the formation of the minor zeolite phases under suitable conditions. This highlights a compositional control in zeolite mineralogy, where the type and abundance of the precursor clays strongly influence the distribution and chemistry of the resulting zeolite minerals. Additionally, the occurrence of mordenite and chabazite exclusively at higher experimental temperatures (i.e., ≥ 200 °C) suggests that their formation is favoured by high silica activity. However, the continued formation of feldspar-replacing smectite, despite the consumption of pre-existing smectite suggests slight to moderate alkaline formation that partly favours the formation of the smectite.

Although less common, a secondary pathway for the formation of analcime, mordenite, and chabazite in this study involved the intermediate development of amorphous aluminosilicate gels (Fig. 6E, F). These gels likely originated from the dissolution of silicate minerals, especially feldspars and clay phases (Fig. 4E, F). The gels subsequently underwent pervasive dissolution, which facilitated the nucleation of analcime, mordenite, and chabazite (Fig. 6E, F). This interpretation is consistent with previous studies, which propose that zeolites can crystallize from aluminosilicate gels that initially precipitate from solution, indicating a two-step process involving gel formation followed by zeolite growth^8,12,43.

Condition of formation of analcime

The Si/Al ratio for zeolites is one of the key parameters that establish the conditions under which they formed^1,12. Results of the measurements of Si/Al ratio for analcime in lacustrine samples from Bohai Bay Basin (China) and Jiuxi Basin (China) reveal that it ranges between 2.04 and 2.08, respectively, suggesting that the analcime probably formed under alkaline water conditions, with likely precursors consisting of detrital clays and other materials^6,68. However, the Si/Al ratio for analcime from the lacustrine Junggar and Erlian Basins (China) show that it varies from 2.36 to 2.64, suggesting that they were formed in saline-alkaline pore water. In addition, while analcime with low Si/Al ratio are characterized by weak thermal stability, those with high Si/Al ratio commonly have moderate to high thermal stability⁶. In the present study, although preliminary Si/Al estimates were obtained from SEM–EDX analyses, these data are inherently semi-quantitative and do not provide sufficient precision for reliable structural formula calculations. Accordingly, definitive Si/Al ratios are not reported. Instead, the experimental conditions themselves provide strong constraints on the formation environment. All hydrothermal experiments were conducted under strongly alkaline conditions (initial pH ~ 10–11.5 at 25 °C) using 0.1 M and 0.5 M Na₂CO3 solutions (Table 1). These alkaline conditions closely resemble those reported for natural lacustrine and saline–alkaline systems in which analcime commonly forms.

Therefore, while precise compositional ratios cannot be robustly constrained in this study, the controlled alkaline chemistry of the experiments supports the interpretation that analcime crystallization was promoted by elevated pH and Na-rich fluids, comparable to natural alkaline diagenetic environments.

Implications for conventional and unconventional reservoirs

Zeolites can have significant impacts on the quality and petrophysical properties of clastic reservoirs^4,6,18. As discussed, zeolites (e.g., analcime) preferentially form under highly alkaline conditions and, hence, are chemically unstable under acidic conditions^4,18. During maturation of organic matter in hydrocarbon source rocks, carboxylic acids are generated and, as a result of pressure buildup and compaction, these acids (along with water) are expelled from the source rocks and migrate into arenitic sandstone beds up-dip^69,70. This results in the dissolution of feldspar and analcime, thereby creating intragranular and intracrystalline porosities, respectively. For instance, Chen¹⁷ reported that the analcime-cemented, Middle Permian Lower-Wuerhe Formation, Junggar Basin, (Northwest China) has 9% porosity, more than 82% of which is secondary porosity related to the dissolution of analcime. However, although our study did not investigate the dissolution of analcime, the synthesized analcime has the potential of secondary porosity generation, should exposure to acidic water occur.

Our study has also revealed that the formation of analcime consumes smectitic and illitic pore-filling matrix (Fig. 5A–E), thereby strengthening the framework grains and preventing the impact of mechanical compaction. In addition, the analcime crystals exhibit abundant, well-developed intercrystalline porosity that would potentially contribute to fluid storage and withdrawal. Therefore, the intercrystalline, micropores of analcime are highly significant in hydrocarbon reservoirs^4,10,71.

Lacustrine, deep to semi-deep fine-grained rocks (e.g., siltstone, mudstone, shale etc.) are often well-cemented by analcime, mainly as pore-filling, fracture-filling, and grain-replacive cements^6,13. Like other brittle minerals such as quartz, feldspar, and carbonate cement, analcime is characterized by high Young’s Modulus and low Poisson’s ratio, thereby favouring their fracturing and enhancing the unconventional reservoir characteristics of the host rocks^5,13.

Limitations of the study and future work

Our study has shown that the synthesized analcime crystals are morphologically comparable to those reported in ancient natural systems^6,7,18, suggesting that both experimental and naturally occurring analcime require appropriate precursor minerals and alkaline pore-fluid compositions for crystallization. However, the chemical characterization of the synthesized zeolites in this study was based on semi-quantitative SEM–EDX analyses, which do not provide the level of precision required for robust determination of mineral chemistry or key compositional parameters such as Si/Al ratios. Future studies should incorporate wavelength-dispersive spectrometry (WDS) electron microprobe analyses to accurately establish the major-element chemistry of the synthesized zeolites and to reliably constrain important ratios, including Si/Al, that are critical for interpreting formation conditions and thermal stability.

In addition, thermodynamic stability modeling using geochemical software (e.g., Geochemist’s Workbench, ACT2 module) would provide quantitative constraints on phase stability fields and reaction pathways under varying temperature, pH, and fluid compositions. Such modeling would strengthen the mechanistic interpretation of zeolite transformation sequences and allow better comparison between experimental results and natural diagenetic systems.

Furthermore, the experimental conditions cannot fully replicate natural diagenesis due to the relatively short duration of laboratory experiments compared to geological timescales. Although analcime in natural systems can form over a wide range of temperatures (including higher-temperature settings associated with igneous environments), the selected experimental temperatures may exceed those typical of sedimentary diagenetic systems. In addition, while the imposed alkaline conditions were sufficient to precipitate analcime, the simplified Na₂CO₃ solutions used in the experiments do not fully capture the chemical complexity of natural formation waters.

Although recent studies have advanced our understanding of the origin, occurrence, and distribution of analcime in clastic sedimentary rocks, further work is needed to better constrain its dissolution mechanisms, the development of secondary intracrystalline porosity, and the effectiveness of intercrystalline pore networks in facilitating fluid flow. In addition, geomechanical testing of analcime-cemented sandstones would help quantify its role in enhancing framework stability and resisting mechanical compaction. Additional experimental investigations are also required to determine the minimum pH threshold necessary for analcime formation, which would provide improved constraints on the alkalinity conditions required for its precipitation. Replicating similar hydrothermal experiments using modern, unconsolidated sand-dune or interdune sediments with variable feldspar and clay contents could further elucidate the early stages of alteration leading to analcime formation, particularly since the samples used in this study had already undergone shallow burial diagenesis prior to experimentation. The oxygen isotope composition of the synthesized analcime is vital for establishing their crystallization temperatures. However, if the isotope data appeared unreliable or challenging to measure, the oxygen isotopic analysis of other minerals commonly associated with analcime, such as chlorite or illite can be measured.

Conclusion

• The Tertiary, shallow-buried sandstone sample of the Al Wajh Formation (northwest Saudi Arabia) was utilized for experimental formation of various shapes of analcime crystals (including cubic, spherical, trapezohedral, and aggregates) at different experimental temperatures (80, 150, 200, 250 °C) for 336 h each (using a 0.1 and 0.5 M Na₂CO₃ synthetic solutions).

• The experimental results have provided evidence that that feldspar and clay minerals (e.g., kaolinite, smectite, and illite) were consumed to produce analcime and trace amounts of other zeolites (chabazite and mordenite).

• The synthesized analcime formed under strongly alkaline experimental conditions comparable to those reported for ancient, highly alkaline lacustrine environments. The synthesized analcime crystals were characterized by abundant, interconnected, and intercrystalline porosity (up to 9 µm in diameter). These crystals have the potentials to strengthen the sandstone’s framework components, prevent mechanical compaction, reduce pore-filling clay matrix, and generate secondary intracrystalline porosity due to dissolution, thereby significantly improving the overall reservoir quality during deep burial.

Author contributions

A.M. Bello: Conceptualization, Data Curation, Experiments, Formal Analysis, Sampling, Investigation, Methodology, Software, Visualization, Writing—original draft, Writing—review & editing. A.M. Salisu: Data Curation, Experiments, Validation, Visualization, Writing—review & editing. A.O. Amao: Resources, Supervision, Validation, Visualization, Writing—review & editing. O.S. Alade: Experiments, Resources, Validation, Writing– review & editing. M. Mahmoud: Resources, Supervision, Validation, Writing– review & editing. K. Al-Ramadan: Supervision, Resources, Funding Acquisition, Validation, Visualization, Writing—review & editing.

Funding

This study was funded by the College of Petroleum Engineering & Geosciences (Grant number: SF24004), King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia.

Data availability

All data supporting the findings of this study are available from the corresponding authors upon reasonable request. Formal data requests may be directed to abdulwahab.bello@kfupm.edu.sa or ramadank@kfupm.edu.sa.

Declarations

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.