Devitrification-driven pore formation in the tight tuff from the Tiaohu formation in the Santanghu Basin, Northwest China

Abstract

The formation of oil-bearing pores in tight tuff has attracted considerable attention from petroleum geologists since the discovery of industrial oil.Devitrification may be an important cause for the formation of these pores; however, the relevant geological circumstance for devitrification still remains unclear. This study tries to decipher the formation of devitrification pores in the tight tuff in the Tiaohu Formation,the Santanghu Basin, Xinjiang, NW China. The result shows that the oil-bearing pore size in tuff is mainly in the range of micrometers to a few nanometers, and the porosity is mainly distributed between 0.10% and 26.71%; the permeability is mainly distributed between 0.17 and 1.20 mD. After high-temperature soaking, the oil-bearing tight tuff illustrated devitrification under both acidic and alkaline circumstances, with glassy tuff showing the greatest variation in porosity, followed by crystal pyroclast glassy tuff, while the mudstone tuff and silicified tuff show relatively small variations in porosity. The 140 °C threshold marks the optimal thermal window for devitrification-driven porosity: it coincides with the smectite–illite transition and the main hydrocarbon-generation stage in the Santanghu Basin. Porosity in all tuff varieties peaks at this temperature, recording a net gain of up to 16.31%; above 140 °C, porosity declines progressively. Devitrification proceeds in three successive stages: (1) neo-mineral nucleation, (2) metasomatic replacement, and (3) dissolution.

Introduction

The Tiaohu Formation within the Santanghu Basin has emerged as a critical research focus due to its substantial tight oil reserves hosted in volcaniclastic tuff lithologies. Distinct from conventional reservoir systems, the unique pore structures in these tight tuffs significantly influence their storage capacity and hydrocarbon accumulation potential^1–3. Current understanding of these pore systems highlights the importance of devitrification-related porosity, which constitutes fundamental components of lacustrine reservoir pore networks^4,5,5.

Despite these advances, a critical research gap persists: how devitrification pores actually develop under subsurface conditions of varying temperature, pressure, and fluid chemistry remains poorly constrained. While recent studies (e.g^5,6., have documented the abundance of devitrification pores in volcanic reservoirs, they predominantly characterize pre-existing pore architectures rather than quantifying the kinetic controls and geochemical pathways that govern pore genesis in situ. For example, Zheng et al.⁵ reported that devitrification pores contribute up to 70% of total pore volume in some tuff units, yet the specific temperature–pH window that maximizes pore creation, and the attendant mineral reactions, have not been systematically investigated under simulated burial conditions.

The petrogenetic (diagenetic) potential of devitrification manifests through two primary pathways: (1) direct crystallization of glassy phases into aluminum silicate minerals (including orthoclase, anorthite, and zeolites), and (2) subsequent dissolution of these crystalline products under acidic diagenetic fluids, ultimately generating secondary porosity^7,8. This dual-pathway process suggests that devitrification contributes differentially to porosity development depending on specific geochemical conditions and mineralogical transformations.Empirical evidence from volcanic reservoirs demonstrates the significant porosity contribution from devitrification processes. Field studies reveal that micropores formed through this mechanism may constitute up to 70% of total pore volume in some tuff units⁹, while petrophysical evaluations attribute 18% of total reservoir quality enhancement todevitrification-related processes⁸. Regional studiesfurther quantify these effects—the Permian Tiaohu Formation exhibits devitrification-derived porescontributing 14% of total porosity, and recent work in the Upper Triassic Ordos Basin has confirmedextensivedevitrification within tight tuff reservoirs that generates inter-microcrystalline pore networks⁶.

By integrating high-temperature–pressure reactive-flow experiments with micro-CT quantification, this study directly addresses the above gap, providing the first quantitative constraints on how temperature (100–180 °C), pH (3–9), and rock composition control the rate and magnitude of devitrification-driven pore formation in Tiaohu Formation tuffs.

Geologic setting

Situated in northeastern Xinjiang, the Santanghu Basin constitutes one of China’s major continental hydrocarbon systems, hosting conventional and unconventional resources¹⁰. The basin contains multiple hydrocarbon-bearing formations, with the Tiaohu Formation being particularly significant for its tight sandstone and tuffaceous reservoirs. Tectonically, this intracontinental superimposed basin formed through complex interactions between the Tianshan and Altai orogenic belts, overlying an Early Paleozoic collisional orogen^11,10 (Fig. 1). The unique tectono-stratigraphic architecture has created favorable conditions for multi-layer hydrocarbon accumulation.

Fig. 1.

Location and regional tectonic framework and litho-stratigraphic column chart of the Santanghu Basin, NW China (modified after¹², drawn using Coreldraw, version 21.2.0.706 https://www.coreldraw.com/cn/)

From a petroleum geological perspective, the basin exhibits remarkable resource diversity, containing significant accumulations of tight oil, shale oil, and shale gas¹⁰. This hydrocarbon diversity results from the interplay of structural deformation, stratigraphic trapping, and lithologic heterogeneity. The Tiaohu Formation, serving as a principal exploration target, exhibits complex diageneticevolution that control reservoir quality through multiple processes including compaction, cementation, and mineral alteration - particularly the devitrification-induced porosity formation discussed in previous studies.

The Santanghu Basin exhibits a complete sedimentary successionunconformably overlying Paleozoic basement, spanning from Carboniferous to Quaternary with a cumulative thickness exceeding 6500 m. The Carboniferous Karagang Formation records the basin’s peak volcanic activity, characterized by>70% basaltic volcaniclastic deposits through intense subaqueous eruptions^11,13.The overlying Permian Lucaogou Formation (150-300m thick) marks a significant paleoenvironmental transition to a hybrid depositional system combining siliciclastic-carbonate deposition dominated by organic-rich tuffaceousmicrites¹⁰. This transition reflects diminished volcanic output but persistent influence from peripheral magmatism, resulting in fine-grained ash-carbonate mixtures. The Lucaogou Formation exhibits a tripartite stratigraphic division: (1) basal coarse-grained tuffaceous sandstones, (2) middle interbedded volcanic ash/carbonate deposits (middle member, 150–300 m), and (3) upper carbonate-cemented siltstones. The middle member’s high organic content (TOC up to 3.5%) and volcanic ash-rich composition established critical precursors for subsequent devitrification-related porosity development in overlying Tiaohu Formation tuffs.The Permian Tiaohu Formation contains a set of volcanic rocks (mainly basalt) interbedded with clastic sedimentary rocks. According to the lithological characteristics, it can be further divided into three lithological Members from bottom to top: Member 1, Member 2, and Member 3 (Fig. 1).

Member 1 (P₂t¹) of the Tiaohu Formation exhibits extensive basin-wide distribution with remarkable lithological consistency.The depositional periodwas characterized by intense volcanic activities resulted in the accumulation of a thick basaltsequence (200–600 m) throughout the majority of the basin. In certain areas, this basalt was intruded by diabase dikes, forming critical hydrocarbon migration pathways. These geological characteristics collectively establish Member 1 as the principal oil-bearing intervalwithin the Tiaohu Formation.

Member 2 of the Tiaohu Formation (P₂t²) mainly comprises lacustrine succession characterized by dark gray to gray mudstones, tuffaceous mudstones, calcareous mudstones, tuffaceous sandstones, and gray-black carbonaceous mudstones, with a total thickness varying between 50–400 m. The depositional history records an evolutionary trend from initially vigorous volcanic activity that gradually waned through time. The lower portion contains a distinctive 10–30 m thick interval of shallow to semi-deep lacustrine tuff deposits, which currently represents an important tight oil reservoir unit. These transition upward into more stable tuffaceous mudstone and fine-grained siltstone deposits. The reservoir potential varies significantly among these lithofacies: tuff exhibits the highest oil saturation,tuffaceous siltstone displays minor hydrocarbon shows, while tuffaceous mudstone exhibits non-reservoir characteristics.

Member 3 of the Tiaohu Formation (P₂t³) records a resurgence of intense volcanic activity, dominated by volcaniclastic rocks and effusive lava flows, with sparse lacustrine or transitional sedimentary interbeds. Post-eruptive tectonic uplift induced significant erosion, truncating the upper stratigraphic record.

Petrologic and microscopic pore features of the tuff

Petrologic features of the tuff

The tuff in the Tiaohu Formation can be subdivided into glassy tuff (GT), crystal pyroclast glassy tuff (CPGT), crystal pyroclast tuff, mudstone tuff (MT), silicified tuff (ST), and diatomaceous tuff. In this study, water-rock reaction simulation experiments were conducted on GT, CPGT, MT, and ST (Fig. 2).GT constitutes the most oil-bearing tuff reservoir within the Tiaohu Formation, containing >50% glassy components. Microscopically, it exhibits a characteristic glassy texture, with volcaniclastic debris composed of both vitric and crystalline fragments arranged in a subordinate directional pattern. Following devitrification, secondary minerals typically include quartz or cryptocrystalline siliceous matrix with feldspar microcrystals. Localized occurrences of biotite are observed, though most have transformed into opaque minerals, retaining only foliated remnants of their original sheet-like morphology.

Fig. 2.

Tuff under microscopy. (a-b). GT, Ma56-15H, 2251.08m; (c-d).CPGT, Ma55, 2478.10m; (e-f). MT, Ma702,1539.01m; (g-h). ST, Lu104, 2123.25m (GT, glassy tuff; CPGT, crystal pyroclast glassy tuff; MT, mudstone tuff; ST, silicified tuff)

Compared with GT, CPGT contains relatively lower proportions of vitric fragments and higher proportions of crystalline fragments. However, total vitric content remains >50%, with approximately 10% crystalline fragment content. Additionally, crystalline fragments exhibit larger grain sizes, and clay mineral content exceeds that observed in GT.

The primary volcanic ash component of MT comprises vitric fragments (>50%) with minor fine-grained crystalline debris (<10%). Notably, MT contains significant terrigenous clay contributions (>20%), resulting in rock density enhancement. Its mineralogy dominantly consists of dark cryptocrystalline clay minerals and volcaniclastic debris composed predominantly of angular to subangular feldspar and quartz phenocrysts.

ST shares similar vitric/crystalline fragment compositions with GT and CPGT, but is distinguished by well-developed silicification characterized by continuous amorphous silica within the tuff matrix.

Pores in the tuffs

Following petrophysical classification scheme for unconventional reservoirs², the storage systemswithin theTiaohu Formation tuff reservoirs can be categorized into two principal types: pore and fracture (Table 1).

Table 1.

Reservoir space classification and characteristics of the Tiaohu Formation.

Type		Subtypes	Size	Features
pore	primary	Intergranular pores	nm-μm	The debris particles are in point or line contact with each other, and most of them have larger pore sizes
		Intercrystal pores	nm-μm	The main pores are between feldspar, quartz, and clay minerals (mainly illite and chlorite), with a large number of pores and relatively small pore sizes compared to intergranular pores
		Organic pores	nm-μm	Residual pores inside the thermal evolution of organic matter.
	secondary	Devitrified pores	nm-μm	Mainly ranging from micrometers to nanometers, but in large quantities.
		Dissolution pores	nm-μm	Pores formed by dissolution of particles
fissure			μm-mm-	Extension and width instability

A total of 41 tuff samples from the Tiaohu Formation were selected for porosity and permeability analysis (Table 2). The results show a positive correlation between porosity and permeability (Fig. 3). The porosity of tuff primarily ranges between 5% and 20%, with significant variations among different tuff types of tuff. GT exhibits the highest porosity, indicating that the devitrification process has a significant positive impact on the quantity of pores. However, the porosity of ST is relatively low, indicating that siliceous cementation has a significant constraint on the pore development in tuff. The measured permeability of the tuff in the Tiaohu Formation is mostly exceeds 0.1mD, mainly distributed between 0.1mD and 4mD, reflecting lowpermeability.

Table 2.

Gas-measured porosity and permeability in the Tiaohu Formation.

Well No.	Depth/m	Lithology	Porosity/%	Permeability/mD
Ma56	2 145.07	GT	7.62	0.42
Ma56-15H	2 251.08	GT	4.22	0.54
Ma56-12H	2 121.09	GT	8.48	0.35
Ma56-12H	2 124.21	GT	5.14	0.50
Ma56	2 141.80	GT	0.95	0.27
Ma7	1 789.67	GT	5.50	0.39
Ma56-12H	2 119.72	GT	13.79	0.55
Tiao27	2 850.40	GT	15.22	0.47
Ma7	1 788.92	MT	1.32	0.48
Lu104H	2 142.86	ST	4.37	0.27
Lu 104H	2 123.35	ST	0.01	0.24
Lu 104H	2 124.51	ST	0.32	0.12
Lu 104H	2 147.79	ST	0.08	0.31
Lu 102H	2 861.80	ST	1.51	0.59
Ma55	2 268.50	CPGT	0.29	0.47
Ma55	2 271.25	CPGT	1.89	0.71
Ma56	2 144.62	CPGT	5.50	0.22
Ma56-12H	2 122.69	CPGT	9.64	0.33

Fig. 3.

Correlation diagram of porosity and permeability of tuff in Tiaohu Formation (GT, glassy tuff; CPGT, crystal pyroclast glassy tuff; MT, mudstone tuff; ST, silicified tuff)

The micro CT scan of different types of tuff in the Tiaohu Formation indicates that the maximum pore throat radius is predominantly distributed between 5 μm and 10 μm. The obtained ST and MT have pore throat coordination numbers mostly less than 3, indicating overall poor pore connectivity. The coordination number of pore throats in GT and CPGT is relatively large, with the highest coordination number of pore throats reaching 13 and 6, respectively. These findings indicate that the pore connectivity of these two types of tuff is relatively better than that of ST and MT.

Methods

Sample preparation

The main reagents used in the experiment include NaOH (99%, Macklin), oxalic acid (C₂H₆O₆; 99.8%, Rhawn), and deionized water. All chemical reagents are of GR grade (Guaranteed Reagent) purity. The experiment simulates an organic acid environment (pH=3) and an alkaline fluid environment (pH=9), under three temperature conditions of 100 °C, 140 °C, and 180 °C. Organic acidic fluids are applied to react with deep core samples, to replicate the actual burial diageneticconditions of deep reservoirs in terms of action mode and water-rock ratio. Oxalic acid was selected because it is (i) one of the most abundant low-molecular-weight organic acids generated during kerogen catagenesis in the Tiaohu Formation^11,14, (ii) the strongest monoprotic acid among natural carboxylic acids, ensuring a stable pH ≈ 3 under experimental conditions, and (iii) less prone to thermal decarboxylation below 200 °C than longer-chain acids, thus maintaining the desired acidity throughout the seven-day run⁶.

The experimental tuff samples (Table 3) are all from the Tiaohu Formation in the Santanghu Basin, including GT, CPGT, MT, ST, tuffaceous sandstone, and tuffaceous siltstone. The first four rock types are the specifically targets of this experiment.

Table 3.

Experimental conditions for tuff samples from the Tiaohu Formation.

Samples No.	Well No.	Depth (m)	Lithology	pH	Temperature/°C	Pressure/Mpa
xj045-3	Ma56	2142.9	GT	3	100	3
xj041-3	Ma56-133H	2650.56	GT	3	100	3
xj075-2	Ma56-15H	2260.33	GT	3	100	3
xj045-2	Ma56	2142.9	GT	3	140	3
xj041-2	Ma56-133H	2650.56	GT	3	140	3
xj070	Ma56-15H	2251.08	GT	3	140	3
xj095-3	Ma7	1790.87	GT	3	140	3
xj044	Ma56	2141.8	GT	3	180	3
xj110	Ma56-12H	2121.09	GT	3	180	3
xj008-2	Lu104H	2125.2	CPGT	3	100	3
xj066-7	Ma56-15H	2247.5	CPGT	3	100	3
xj101-2	Ma7	1887.18	CPGT	3	100	3
xj008-5	Lu104H	2125.2	CPGT	3	140	3
xj101-3	Ma7	1887.18	CPGT	3	140	3
xj111	Ma56-12H	2122.69	CPGT	3	180	3
xj001	Lu104H	2123.25	CPGT	3	180	3
xj047	Ma56	2144.62	CPGT	3	180	3
xj135	Tiao27	2850.4	MT	3	140	3
xj093	Ma7	1788.92	MT	3	180	3
xj100-5	Ma7	1885.3	CPGT	9	100	3
xj066-3	Ma56-15H	2247.5	CPGT	9	140	3
xj031	Ma55	2268.5	CPGT	9	180	3
xj035	Ma55	2271.25	CPGT	9	180	3
xj133	Tiao27	2848.47	ST	9	140	3
xj097	Ma7	1793.1	ST	9	180	3
xj016	Lu104H	2142.86	ST	9	180	3
xj041-6	Ma56-133H	2650.56	GT	9	100	3
xj076-6	Ma56-15H	2262.33	GT	9	100	3
xj095-6	Ma7	1790.87	GT	9	100	3
xj076-2	Ma56-15H	2262.33	GT	9	140	3
xj094	Ma7	1789.67	GT	9	180	3
xj109	Ma56-12H	2119.72	GT	9	180	3
xj113	Ma56-12H	2124.21	GT	9	180	3
xj048	Ma56	2145.07	GT	9	180	3

(GT, glassy tuff; CPGT, crystal pyroclast glassy tuff; MT, mudstone tuff; ST, silicified tuff)

Equipment

The high-temperature, high-pressure reactor (model PPLKH) features acid/alkali resistance and a 50 mL capacity. The external chamber measures 65 mm (width) × 138 mm (height), while the inner liner dimensions are 39.8 mm (inner diameter) × 78 mm (height). The vacuum oven supports chemical reactions under controlled conditions (pressure 3 MPa; temperature 280 °C). Critical components in contact with reaction solutions—including pistons, cylinders, plugs, reactor bodies, and samplers—are constructed from 304 stainless steel to ensure corrosion resistance. The system includes programmable temperature/pressure presets and overload protection mechanisms.

CT scanning was conducted on water-rock samples before and after reactions to analyze pore structure evolution. The methodology employs X-ray attenuation imaging through 360° sample rotation to generate three-dimensional models. Scanning parameters utilize cone-shaped X-rays with variable magnification lenses to achieve precise pore structure and bulk density characterization. The GE-produced micro-nano dual-tube core CT system (V|Tom|XS180&240) features: Micro-tube: 0–240 kV, 2–122 μm pixel size, Nano-tube: 0–180 kV, 0.6–20 μm pixel size, Power output: 1–100 W. In this study, the micro-CT scans were acquired at 2 µm voxel resolution using a 180 kV nano-tube. A global grey-value threshold of 8,000 (on a 0–65,535 16-bit scale) was applied after beam-hardening correction to segment pore space from mineral matrix; this threshold was validated by matching helium-porosity measurements on the same plugs (±0.5% absolute difference).

Experimental process

Sample Preparation: An appropriate quantity of solid reagent was added to a clean beaker, followed by deionized water. The mixture was agitated vigorously with a glass rod to expedite dissolution. A portion of the high-concentration solution was extracted using a disposable syringe for subsequent use. A magnet was positioned within the beaker, which was then placed at the center of a magnetic stirrer. Stirring was maintained at calibrated speeds for >15 minutes to ensure homogeneity, with pH monitored throughout using a calibrated pH meter. Adjustments to pH were made by adding organic acid solutions or deionized water as required. Prior to measurements, the pH meter was calibrated with standard buffer solutions.

Experimental Execution: Each experimental group contained 10 samples receiving 40 mL of reaction solution per replicate. Samples underwent programmed heating (0.5 °C/min) with periodic pressure adjustments. After reaching the target temperature, pressure was incrementally increased to the set point, maintaining constant temperature and pressure conditions for seven days. After reaction, the reactor was naturally cooled to ambient temperature before sample retrieval and rinsing with deionized water. SEM analysis was conducted on pre- and post-reaction block samples, while columnar specimens were subjected to porosity and permeability assessments for comparative evaluation.

Post-Reaction Analysis: Post-reaction solutions were extracted via disposable syringes and filtered through 0.22 μm organic membranes prior to cation concentration analysis. Major cations (Al³⁺, Ca²⁺, K⁺, Mg²⁺, Na⁺, Fe²⁺, Fe³⁺, Mn²⁺, Si⁴⁺) were quantified using an Agilent 5100 ICP-OES system (manufactured in the USA) at the State Key Laboratory of East China University of Technology. Analytical conditions included 21.5°C ambient temperature, humidity of 47% RH, and cross-comparative methodology.

Results

The findings indicate that porosity and permeability of the Tiaohu Formation tuffs generally increase after water-rock reactions, while a minor subset exhibits decreases in these properties. Notably, the correlation between porosity and permeability improves significantly after reactions (Table 4 and Fig. 4).

Table 4.

Porosity and permeability of tuff before and after water-rock reaction in the Tiaohu Formation.

Sample No.	Well No.	Depth (m)	Lithology	Porosity before reaction /%	Porosity after reaction /%	Permeability before reaction /mD	Permeability after reaction /mD	T/℃	pH	P /MPa
XJ048	Ma56	2145.07	GT	7.62	7.32	0.42	0.32	180	9	3
XJ070	Ma56-15H	2251.08	GT	4.22	20.53	0.54	0.49	140	3	3
XJ110	Ma56-12H	2121.09	GT	8.48	13.62	0.35	0.42	180	3	3
XJ113	Ma56-12H	2124.21	GT	5.14	10.61	0.5	0.65	180	9	3
XJ109	Ma56-12H	2119.72	GT	13.79	15.16	0.55	1.07	180	9	3
XJ135	Tiao27	2850.40	GT	15.22	26.71	0.47	0.42	140	3	3
XJ093	Ma7	1788.92	MT	1.32	3.88	0.48	0.49	180	3	3
XJ124	Ma62H	2379.00	MT	5.38	5.41	0.82	0.83	100	3	3
XJ033	Ma55	2270.10	MT	4.53	4.54	1.13	1.20	100	3	3
XJ016	Lu104H	2142.86	ST	4.37	3.22	0.27	0.23	180	9	3
XJ047	Ma56	2144.62	CPGT	5.50	4.54	0.22	0.17	180	3	3
XJ111	Ma56-12H	2122.69	CPGT	9.64	15.26	0.33	0.89	180	3	3

(GT, glassy tuff; CPGT, crystal pyroclast glassy tuff; MT, mudstone tuff; ST, silicified tuff).

Fig. 4.

The pore-permeability correlation diagram before and after the water-rock reaction of tuff in the Tiaohu Formation

Porosity and permeability

Acidic Conditions (100°C): Mudstone tuff (MT) samples XJ124 and XJ033 exhibited initial porosities of 5.38% and 4.53%, respectively. After water-rock reactions, marginal increases to 5.41% (Δ+0.03%) and 4.54% (Δ+0.01%) were observed.

Elevated Temperature (140°C): Glassy tuff (GT) samples XJ070 and XJ135 demonstrated significant porosity enhancements. XJ070 increased from 4.22% to 15.22% (Δ+11.0%), while XJ135 increased from 20.53% to 26.71% (Δ+6.18%). Scanning electron microscopy (SEM) confirmed near-doubling of GT porosity (from ~10% to ~20%) after reaction (Fig. 5).

Fig. 5.

GT under scanning microscope. (a)Before water-rock reaction experiment(Ma56-15H, 2 262.33 m); (b)After water-rock reaction experiment(Ma56-15H, 2262.33 m).

Highest Temperature (180 °C): GT sample XJ110 showed a porosity increase from 8.48% to 13.62% (Δ+5.14%). Crystal pyroclast glassy tuff (CPGT) samples exhibited contrasting trends: XJ047 decreased from 5.50% to 4.54% (Δ−0.96%), whereas XJ111 increased from 9.64% to 15.26% (Δ+5.62%). Tuff sample XJ093 displayed a substantial rise from 1.32% to 3.88% (Δ+2.56%).

Alkaline Conditions (180°C): GT samples XJ113 and XJ109 showed porosity increases (XJ113: 5.14%→10.61%, Δ+5.47%; XJ109: 13.79%→15.16%, Δ+1.37%), while XJ048 exhibited a minor decrease (7.62%→7.32%, Δ−0.30%). Silicified tuff (ST) sample XJ016 displayed porosity reduction (4.37%→3.22%, Δ−1.15%) (Fig. 6).

Fig. 6.

Cross-plot of porosity and permeability (left) and variation of porosity after water-rock reaction of tuff (right)

The analysis of experimental data reveals significant variations in pore enlargement effects among samples of the same tuff type under identical physicochemical conditions. Under acidic fluid conditions, porosity increases were observed in GT samples XJ070 (Well Ma56-15H, 2251.08 m) and XJ135 (Well Tiao27, 2850.40 m) after water-rock reactions. Specifically, XJ070 exhibited a 16.31% porosity increase, while XJ135 showed an 11.49% rise, demonstrating a 4.82% porosity difference between the two samples. Conversely, under alkaline fluid conditions, GT sample XJ048 (Well Ma56, 2145.07 m) displayed a minor 0.30% porosity decrease after reaction, whereas GT sample XJ109 (Well Ma56-12H, 2119.72 m) exhibited a 1.37% porosity increase.

Pore structure

CT scanning experiments examined six samples (Table 5) representing four tuff types: glassy tuff (GT), crystal pyroclast glassy tuff (CPGT), mudstone tuff (MT), and silicified tuff (ST). The samples were subjected to two fluid conditions (temperature 180 °C; pH: 3 and 9) with three experimental replicates per condition (Table 5).

Table 5.

CT scanning samples for water-rock reaction.

Sample No.	Well No.	Depth/m	Lithology	Temperature/℃	pH
XJ044	Ma56	2141.80	GT	180℃	3
XJ111	Ma56-12H	2122.69	CPGT	180℃	3
XJ093	Ma7	1788.92	MT	180℃	3
XJ094	Ma7	1789.67	GT	180℃	9
XJ097	Ma7	1793.10	ST	180℃	9
XJ109	Ma56-12H	2119.72	GT	180℃	9

pH=3 and T=180 ℃

CT scanning analysis of GT sample XJ044 (Well Ma56-15H, 2141.80 m) reveals well-developed micropores with no significant heterogeneity (Fig. 7). Comparative analysis of binary segmentation images shows an exponential increase in pore quantity after reaction. Three-dimensional pore network models highlight expanded white regions (representing connected pore spaces) between two or more pores in reacted samples. Red lines connecting red dots denote pore throats—channels linking adjacent pores—with throat quantity directly correlating to pore connectivity strength.

Fig. 7.

CT 3D image of GT sample XJ044 (Well Ma 56, 2141.80 m).

CT scanning analysis reveals that pre-reaction pore radii in GT samples (XJ044) range between 5–10 μm, exhibiting a multimodal distribution pattern(Fig. 8). Post-reaction pore radii expand to 5–15 μm, showing increased modal frequency and expanded pore connectivity. Pore throat radii demonstrate similar trends: pre-reaction multimodal distributions shift to concentrated post-reaction ranges (5–15 μm), with throat radii increasing significantly (2–12 μm post-reaction). Notably, coordination number distribution displays no significant post-reaction increase.

Fig. 8.

Pore-throat structure characteristics of GT sample XJ044 (Well Ma56, 2141.80m) based on micro CT scanning.

Analysis of CPGT sample XJ111 (Well Ma56-12H, 2122.69 m depth) confirms widespread micropore development with no significant heterogeneity. Binary segmentation comparisons show exponential pore proliferation after the reaction. Three-dimensional pore network models reveal marked increases in white pore throats (interconnecting channels), indicating enhanced throat connectivity (Fig. 9).

Fig. 9.

CT Analysis of CPGT sample XJ111 (Well Ma56-12H, 2122.69 m)

Pore-throat structure analysis (Fig. 10) on CPGT reveals pre-reaction pore radii ranging from 5 to 22 μm. Post-reaction pores expand to 5–26 μm, showing increased modal frequency despite unchanged distribution patterns. Pore throat radii increase (5–26 μm post-reaction) similarly, while their distribution frequency remains stable. Both pore and throat radii exhibit multimodal variation. Post-reaction throat radii (2–16 μm) demonstrate reduced diameter ranges compared to pre-reaction conditions, maintaining multimodal distributions with diminished peak frequencies. Notably, coordination numbers increase after reaction.

Fig. 10.

Pore-throat structure characteristics of CPGT sample XJ111 (Well Ma 56-12H, 2122.69 m) based on micro CT Scanning

CT core analysis of MT sample XJ093 (Well Ma7-15H, 1788.92 m depth)reveals extensively developed micropores with no significant heterogeneity (Fig. 11). Comparative analysis of binary segmentation images demonstrates an exponential increase in pore quantity within the tuff following reaction. Three-dimensional pore network models confirm substantial pore throat proliferation in the reacted sample.

Fig. 11.

Digital core analysis of MT sample XJ093 (Well Ma7, 1788.92 m) based on micro-CT scanning

Pore-throat structure analysis of MT sample (Fig. 12) reveals pre-reaction pore radii ranging from 2 to 12 μm. Post-reaction distributions exhibit bimodal patterns, with pore radii concentrated in 2–12 μm and 30–60 μm ranges. The 2–12 μm pore frequency decreases following reaction, accompanied by increased prevalence of larger pores (40–60 μm). Similar trends are observed in throat radii, though 2–12 μm throat frequencies show no significant increase. Both pore and throat radii display multimodal variations. Pre-reaction throat radii exhibit bimodal distributions (2–8 μm and 20–35 μm), transitioning to unimodal post-reaction distributions (2–8 μm) with elevated peak frequencies. Notably, coordination numbers and their distribution frequencies significantly increase following water-rock reactions.

Fig. 12.

Pore-throat structure characteristics of MT sample XJ093 (Well Ma 7, 1788.92 m) based on micro-CT scanning

pH=9 and T=180 ℃

CT core analysis of GT sample XJ094 (Well Ma7-15H 1789.67 m depth) under alkaline conditions (pH=9) and high-temperature conditions (180°C) (Fig. 13) reveals well-developed micropores and observable fracture networks. Comparative analysis of binary segmentation images demonstrates a significant pore proliferation following reaction. Three-dimensional pore models indicate thickened fracture surface throat apertures and increased throat quantity following water-rock reactions.

Fig. 13.

Digital core analysis of GT sample XJ094 (Well Ma7, 1789.67m) based on micro CT scanning

Pore-throat analysis of GT sample (Fig. 14) reveals pre-reaction pore radii ranging from 3 to 10 μm. Post-reaction pore radii expand to 3–14 μm, with peak frequencies showing slight decreases. Throat radii distributions mirror pre-reaction patterns but exhibit increased throat radii (2–10 μm post-reaction) accompanied by marked declines in peak frequencies. Pre-reaction throat radii range from 2 to 8 μm, shifting to wider post-reaction distributions (2–10 μm), indicating overall throat expansion. Notably, coordination numbers display no significant post-reaction increase.

Fig. 15.

Digital core analysis of ST sample XJ097 (Well Ma7, 1793.10 m) based on micro-CT scanning

Fig. 14.

Pore-throat structure characteristics of GT sample XJ094 (Well Ma 7, 1789.67 m) based on micro CT scanning. CT digital core analysis of ST sample XJ097 (Well Ma7-15H, 1793.10 m depth) under alkaline conditions (pH=9) and high-temperature conditions (180 °C) (Fig. 15) reveals extensively developed micropores with no significant heterogeneity. Comparative analysis of binary segmentation images demonstrates significant pore proliferation following reaction. Three-dimensional pore network models confirm exponential increases in throat quantity within the reacted sample.

Pore-throat structure analysis of ST sample (Fig. 16) reveals pre-reaction pore radii ranging from 5 to 10 μm. Post-reaction distributions show reduced pore radii (5–9 μm) accompanied by elevated peak frequencies. Throat radius changes mirror pore trends, with pre-reaction throat radii (5–7 μm) expanding to 4–8 μm post-reaction, though peak frequencies decline significantly. Notably, coordination numbers increase following reaction despite decreased distribution frequencies.

Fig. 17.

Digital core analysis of GT sample XJ109 (Well Ma 56-12H, 2119.72 m) based on micro-CT scanning

Fig. 16.

Pore throat structure characteristics of ST sample XJ097 (Well Ma7, 1793.10 m) based on micro CT scanning. CT digital core analysis of GT sample XJ109 (Well Ma56-12H, 2119.72 m depth) under alkaline conditions (pH=9) and high-temperature conditions (180°C) (Fig. 17) reveals extensively developed micropores with no significant heterogeneity. Comparative analysis of binary segmentation images demonstrates exponential pore proliferation following reaction. Three-dimensional pore network models confirm exponential increases in throat quantity within the reacted sample.

Pore-throat structure analysis of ST sample (Fig. 18) reveals pre- and post-reaction pore radii predominantly distributed between 5–10 μm. Upper bounds of pore/throat radius distributions decrease following reaction, accompanied by reduced peak frequencies. Throat radii expand from pre-reaction ranges (4–8 μm) to post-reaction distributions (4–9 μm), while maintaining stable peak frequencies. Notably, coordination numbers increase following reaction, with corresponding elevation in coordination value magnitudes.

Fig. 18.

Pore-throat structure characteristics of GT sampleXJ109 (WellMa 56-12H, 2119.72m) based on micro-CT scanning

Comprehensive digital core analysis and CT-based pore-throat structure characterization reveal: (1) Under comparable fluid conditions, GT demonstrates the most intense water-rock reactions, followed by CPGT. MT exhibits stronger reactions than ST, which shows the weakest response. (2) Comparative analysis of binary segmentation images and 3D pore network models across fluid conditions demonstrates that acidic environments enhance tuff reactions, promote more thorough devitrification, and achieve superior pore enlargement. (3) Coordination numbers indicate superior pore connectivity in GT, followed by CPGT. MT displays moderate connectivity, whereas ST exhibits the poorest interconnectivity.

Changes in fluid and mineral composition

The water-rock reaction experiment comprised 34 samples (19 columnar tuffs and 15 blocky tuffs). Thirty-six samples were subjected to organic acid (pH=3) and alkaline solutions (pH=9), with three temperature gradients (100 °C, 140 °C, and 180 °C) maintained under pressure 3 MPa across four experimental groups (Table 6).

Table 6.

Ion concentration test results after water-rock reaction of the Tiaohu Formation tuff.

Sample No.	Well No.	Lithology	Depthp/m	pH	Temperature/°C	Al3+(mg/L)	Ca2+(mg/L)	K+(mg/L)	Mg2+(mg/L)	Na+(mg/L)	Fe2+、3+(mg/L)	Mn2+(mg/L)	Si4+(mg/L)
xj045-3	Ma56	GT	2142.9	3	100	5.7843	0	0.3919	1.3959	22.1235	1.3814	0.2535	12.383
xj041-3	Ma56-133H	GT	2650.56	3	100	0.0696	18.9037	0.4576	0.8261	14.8281	0	0.1001	6.6211
xj075-2	Ma56-15H	GT	2260.33	3	100	0.4431	0	0.1514	0	24.8139	0	0	4.8657
xj045-2	Ma56	GT	2142.9	3	140	1.0092	0	0.9648	0.0902	22.1035	0.1217	0.0379	20.8179
xj041-2	Ma56-133H	GT	2650.56	3	140	0.4414	13.8978	0.9322	0	17.6167	0.0232	0.038	19.2649
xj070	Ma56-15H	GT	2251.08	3	140	0.2301	1.1842	0.8811	0.0424	136.1462	0.0889	0.027	28.0713
xj095-3	Ma7	GT	1790.87	3	140	3.6816	0.1621	2.6106	1.1641	9.8703	1.4952	0.2278	24.8622
xj044	Ma56	GT	2141.8	3	180	0.1379	0	0.4602	0	38.3626	0	0	22.6526
xj110	Ma56-12H	GT	2121.09	3	180	0.0328	0	0.5771	0	29.1964	0	0	0.5622
xj008-2	Lu104H	CPGT	2125.2	3	100	0.2344	1.3238	1.2761	0	72.2698	0	0.0026	3.0823
xj066-7	Ma56-15H	CPGT	2247.5	3	100	0.4542	0	0.1516	0	25.0783	0	0	4.8528
xj101-2	Ma7	CPGT	1887.18	3	100	6.5545	0	0.3874	1.8795	38.0816	4.6274	0.2438	16.2051
xj008-5	Lu104H	CPGT	2125.2	3	140	1.9182	0.3036	1.5024	0.4129	5.767	0.0496	0.0478	5.1979
xj101-3	Ma7	CPGT	1887.18	3	140	3.0996	2.653	1.6831	1.7011	7.7725	1.824	0.2227	30.6246
xj111	Ma56-12H	CPGT	2122.69	3	180	0.1314	0.3834	0.788	0.1713	41.4468	0	0.0662	24.5201
xj001	Lu104H	CPGT	2123.25	3	180	0.0064	0	0.2503	0	26.476	0	0	16.7787
xj047	Ma56	CPGT	2144.62	3	180	0.2124	0	0.4364	0	43.1181	0.035	0	19.0441
xj135	Tiao27	MT	2850.4	3	140	0	19.3266	2.3126	0.1215	55.9212	0	0	52.3371
xj093	Ma7	MT	1788.92	3	180	0.0792	0	0.4027	0	66.2793	0	0	11.7442
xj100-5	Ma7	CPGT	1885.3	9	100	0.1788	0	0.2291	0	18.0142	0	0	6.9625
xj066-3	Ma56-15H	CPGT	2247.5	9	140	0	0	0.1758	0	7.1609	0	0	3.9473
xj031	Ma55	CPGT	2268.5	9	180	0	0	0	0	0	0	0	0.6449
xj035	Ma55	CPGT	2271.25	9	180	0.5333	0	0.198	0	38.5382	0	0	28.5019
xj133	Tiao27	ST	2848.47	9	140	0.0456	6.8835	3.9016	0.0561	117.1883	0.041	0.0024	68.2754
xj097	Ma7	ST	1793.1	9	180	0.4555	0	1.215	0	5.3093	0	0	21.0253
xj016	Lu104H	ST	2142.86	9	180	0.766	0	0.6903	0	21.2341	0	0	29.2518
xj041-6	Ma56-133H	GT	2650.56	9	100	0.5637	5.0712	0.3995	0	16.8932	0	0.0033	5.8239
xj076-6	Ma56-15H	GT	2262.33	9	100	0.5283	1.2529	0.1849	0	22.7381	0	0	5.2623
xj095-6	Ma7	GT	1790.87	9	100	0.5973	0	0.1821	0	28.765	0	0	4.397
xj076-2	Ma56-15H	GT	2262.33	9	140	0.1184	4.5459	0.144	0	6.7027	0	0.0004	4.346
xj094	Ma7	GT	1789.67	9	180	0.0637	0	0.5801	0	78.3039	0	0	25.7643
xj109	Ma56-12H	GT	2119.72	9	180	0.0946	0.8387	0.8278	0	19.7352	0	0	26.8261
xj113	Ma56-12H	GT	2124.21	9	180	0.1125	2.6225	0.9645	0	52.9412	0	0	22.1944
xj048	Ma56	GT	2145.07	9	180	0.1155	0	0.4917	0	30.379	0	0	28.4366
pH=9	Water rock reaction liquor	/	/	9	/	0	0	0	0	9.1661	0	0	0
pH=3		/	/	3	/	0	0	0	0	0	0	0	0

Major dissolved ions include Na⁺,Si⁴⁺,K⁺, and Al³⁺, with Na⁺ exhibiting the highest concentration (peak value: 136.15 mg/L), followed by Si⁴⁺ (68.28 mg/L), K⁺, and Al³⁺. Negligible dissolution was observed for Ca²⁺,Fe²⁺,Fe³⁺,Mn²⁺, and Mg²⁺, attributed to differential dissolution capacities of fluids on silicate versus carbonate minerals¹⁵.

Ion concentration hierarchies correlate with rock types: GT demonstrates the greatest concentration variability, succeeded by CPGT, while ST shows minimal variation. This trend aligns with GT’s superior physical properties, CPGT’s intermediate properties, and ST’s inferior performance.

Comparative analysis reveals higher Na⁺,Si⁴⁺,K⁺, and Al³⁺ concentrations under acidic conditions versus alkaline environments (Fig. 19), indicating more thorough water-rock reactions in acidic fluids.

Fig. 19.

Ion characteristics of tuff under different fluid properties.

Discussion

Mineral and elemental controls on the 140 °C porosity maximum

Micro-CT data show that the net porosity gain peaks at +16.31 % when experimental temperature reaches 140 °C. Within the Tiaohu Formation this temperature coincides with the kinetic optimum of the smectite-to-illite (S-I) reaction (120–150 °C)^6,13 and with the main stage of hydrocarbon expulsion (Easy%Ro ≈ 0.7–0.9 %). Using the present-day Malang Sag gradient of 29 °C km⁻¹ (surface temperature 10 °C), 140 °C equates to 4.2–4.7 km burial—identical to the cored depth of the most oil-productive tuff interval (4.4–4.6 km). At this level, carboxylic-acid-rich pore waters lower pH to 3–4, accelerating glass hydrolysis and feldspar dissolution while the S-I collapse creates 5–15 µm dissolution voids¹⁴. When experimental T is raised to 180 °C the gain is erased (−2 to −5 %) probably because (i) illitization of kaolinite + K-feldspar reduces solid volume by 15–20 % but the neo-formed illite fibres (≤2 µm) bridge pores and throttle throats^16,(ii) amorphous silica liberated at 140 °C reprecipitates as quartz overgrowths that line and ultimately occlude throats^9,(iii) faster Al³⁺/Si⁴⁺ diffusion accelerates cementation and heals micro-fractures that had briefly enhanced permeability at 140 °C.

The primary mineral assemblage of tuffs comprises quartz, plagioclase, potassium feldspar, and clay minerals (Fig. 20). Clay mineral contents are approximately 10% in GT, 9% in CPGT, and 8% in ST. So the glassy matrix and feldspar dominate reaction kinetics; high SiO₂/Al₂O₃ molar ratios (3.7–6.9) favour quartz and feldspar crystallisation that creates inter-crystalline micropores¹⁷. Conversely, mudstone tuff (MT) carries 38 % ductile clay that compacts under stress and suppresses devitrification, yielding persistently low porosity (<5 %)^18,19. Elevated zeolite content observed in oil-bearing tuffs from wells Ma36-16, L102H, and L104H, which likely reflects early alkaline diagenesis and volcanic material alteration but contribute little additional porosity once the system is flushed by acidic fluids at 140 °C.

Fig. 20.

Mineral composition statistics of GT, CPGT, and ST

A strong positive correlation exists between whole-rock SiO₂ content and porosity in GT. The Ma56-12H interval (84.9% SiO₂, 35% plagioclase) registers 13.79% gas-measured porosity, confirming that Si-rich, clay-poor glassy tuffs are the prime targets for devitrification-driven sweet-spots.

Water-rock reaction process

Building on the 140 °C inflection point documented above, we here unravel the step-by-step mechanisms that convert glassy tuff into an interconnected micro-pore network and why this progression is arrested at 180 °C. The devitrification process in water-rock reactions occurs within an open system^20–22, encompassing geochemical processes such as recrystallization^23,24, dissolution-precipitation, metal ion migration/transformation, and volume reduction during new mineral formation. These mechanisms collectively generate micropores^25–28. Our reactive-flow experiments couple with effluent chemistry (Table 6) and micro-CT 3-D pore networks allow these stages to be quantified for the first time in tight tuff.

Aluminosilicate minerals formed through glass devitrification undergo acidic fluid dissolution, producing dissolution micropores that constitute primary reservoir space in tuffs^8,7. The chemical reactions of quartz and feldspar—primary products of intermediate-acidic GT devitrification—can be summarized as:

As the dominant tuff type in the Tiaohu Formation, GT comprises predominantly glassy phases¹. Vitreous phases exhibit thermodynamic instability, driving devitrification toward crystalline phases. During devitrification, volume contraction generates micropores, while acidic fluids dissolve soluble minerals to form dissolution pores. Due to challenges in distinguishing dissolution from devitrification pores, both are collectively termed “devitrification pores” within the tuff matrix.

Based on water rock simulation experiments and geochemical analyses, six primary reactions govern Tiaohu Formation tuff behavior under fluid influence:

SiO₂+H₂O → Quartz+H₂O

Quartz, as the main product of the devitrification process, has a higher content of Si-O tetrahedra, an increase in the number of shared oxygen angles, and a decrease in the effective static charge of oxygen in volcanic glass. As a result, its ability to attract cations decreases, and the Si-O and Al-O structures containing oxygen are more likely to detach from the original glassy material, especially when the Mg and Fe content in the glassy material is low. When the Si, Na, and K content are high, it is conducive to the formation of quartz and feldspar (Fig. 21). The specific reaction is:

1

2

Fig. 21.

Formation of feldspar and quartz through devitrification. (a). Localized devitrification of glass debris to form potassium feldspar (GT from Ma56-133, 2650.56 m); (b). Sodium feldspar (GT from Ma56, 2142.9 m) is formed through devitrification; (c). The devitrified microcrystalline quartz is in a mosaic-like contact (CPGT from Lu104H, 2125.2 m); (d). Devitrified microcrystalline quartz shows interstitial pores (CPGT from Ma56-15H, 2260.33 m).

Formation of clay minerals

The tuffs of the Tiaohu Formation contain two primary authigenic clay minerals: authigenic illite and authigenic chlorite³. Authigenicillite predominantly occurs as flaky aggregates, formed through devitrification combined with K-rich fluid interactions. Devitrified potassium feldspar may also transform into illite under specific conditions. Elevated Mg/Fe ratios in volcanic glass promote chlorite formation, particularly under alkaline diagenetic conditions, favoring flaky to acicular chlorite textures (Fig. 22). Representative reaction pathways include:

3

4

Fig. 22.

Formation of clay minerals through devitrification (a). Obviou sillitization at the edge of potassium feldspar (GT, Well Ma56-133, 2650.56 m); (b). Obviousillitization at the edge of sodium feldspar (CPGT, Well Ma56-15H, 2247.5 m); (c). Self generated sheet-like chlorite aggregate (GT, Well MA56-15H, 2650.56 m; (d). needle-shaped self generated chlorite aggregate (GT, Well MA56-133, 2650.56 m).

Transformation of plagioclase into potassium feldspar

During the devitrification process of the Tiaohu Formation tuff, plagioclase transforms significantly into potassium feldspar (potassium mineralization), which is manifested by the transformation of sodium feldspar into potassium feldspar under metasomatism and hydrolysis^3,12,14. That is, potassium feldspar is formed at the edge of sodium feldspar particles, while illite and quartz can also be formed. A large number of devitrified pores are developed at the edge of potassium feldspar (Fig. 23). The specific reaction is as follows:

5

6

Fig. 23.

Formation of potassium feldspar and quartz during devitrification process, with developed pores at the edges of feldspar (a). During the process of devitrification, sodium feldspar undergoes potassium feldspar mineralization, and the edge forms potassium feldspar (CPGT, Well Ma7, 1885.3 m); (b). Sodium feldspar hydrolyzes to form a combination of embedded quartz and sheet-like illite (GT, Well Ma56-15H, 2260.33 m); (c). During the process of devitrification, sodium feldspar undergoes potassium feldspar mineralization, and potassium feldspar is formed at the edge. Pores are developed and filled with organic matter (GT, Well Lu1, 2546.86m; d. Potassium feldspar is yielded at the edge of plagioclase (GT, Well Ma7, 1887.18 m).

Main controlling factors of devitrification

Temperature effects

Experimental data show that porosity gain is not linear with temperature: GT achieves its maximum increase (+16.31%) at 140 °C, whereas runs at 100 °C yield <3% and those at 180 °C lose 2–5% (Fig. 24). This non-monotonic trend mirrors the smectite-to-illite reaction rate that peaks at 140 °C under typical burial heating rates¹³. Consequently, we propose that the 140 °C isotherm—corresponding to 4.2–4.7 km burial in the Malang Sag—defines the upper limit of the “dissolution-dominated” regime in the Tiaohu tight tuff.

Fig. 24.

Pore variation law of GT during heating process

Geochemical composition of tuff

GT exhibits higher SiO₂ content and superior pore permeability compared to CPGT. Elevated SiO₂ levels correlate with increased Si-O tetrahedral connectivity in magma melts, promoting higher Si⁴⁺ polymerization and enhanced devitrification. The SiO₂/Al₂O₃ molar ratio ranges from 3.66–6.89 in GT versus 3.14–3.75 in CPGT. Higher molar ratios favor the formation of highly polymerized minerals (e.g., quartz and feldspar) during devitrification, resulting in more devitrification-related pores and improved reservoir quality^29,30.

Fluid properties

Both acidic (pH=3) and alkaline (pH=9) environments promote glass devitrification, with stronger effects observed under acidic conditions. Post-reaction solution analyses reveal decreased Ca²⁺, Mg²⁺, and Fe²⁺ concentrations, alongside significant increases in Al³⁺, Na⁺, and Si⁴⁺. GT demonstrates the greatest ionic variation, followed by CPGT, while ST shows minimal changes, consistent with its superior physical properties (GT > CPGT > ST). Mechanistically, GT undergoes recrystallization first, where felsic tuff transforms into quartz and feldspar. Subsequent hydrolysis of feldspar generates dissolution pores, with Al³⁺,Na⁺, and Si⁴⁺ enrichment in solutions:

Prolonged reaction converts kaolinite to opal (SiO₂·nH₂O), forming characteristic siliceous bands in tuffs.

Influence of potassium-rich fluids

Inflow of K⁺ (up to 136 mg L⁻¹ in effluent; Table 6) drives the replacement of plagioclase by K-feldspar plus illite (reactions 5-6), generating additional 2–8 µm pores at feldspar margins (Fig. 23). This K-metasomatism is most effective at 140 °C where K-feldspar nucleation rate exceeds illite growth rate, further amplifying the porosity peak.

Conclusion

The water-rock reaction simulations demonstrate that tuffs undergo devitrification under both acidic and alkaline conditions, with glassy tuff exhibiting the most pronounced changes in porosity and permeability, followed by crystal pyroclast glassy tuff. Mudstone tuff and silicified tuff display minimal porosity/permeability variations. Devitrification is primarily controlled by temperature, tuff chemistry, fluid properties, and K⁺ concentrations. Across tested conditions (100 °C, 140 °C, 180 °C; pH=3 and 9), glassy tuff achieved maximum porosity enhancement (16.31%) at 140 °C, with porosity progressively declining beyond this threshold. The evolution exhibits two distinct stages: a heating-induced porosity enhancement phase (≤140 °C) followed by a heating-induced porosity reduction phase (>140 °C).

Author contributions

Bin Bai, Chaocheng Dai and Guangrong Li wrote the main manuscript text. Ruojing Dong, Zhijun Zhu and Long Xiang prepared figures and tables. Yuanquan Zhou refined the language. All authors reviewed the manuscript.

Funding

This study was funded by the National Natural Science Foundation of China (U24B6004), PetroChina’s Science and Technology Program (2025DJ102, 2022KT0302-1) and PetroChina Research Institute of Petroleum Exploration & Development (RIPED) Institutional-Level Project (Advanced Basic Research) (2022yjcq03).

Data availability

Data is provided within the manuscript.

Declarations

Competing interests

The authors declare no competing interests.