A
Evolution of moisture transport properties in cement mortar under marine salt spray environmentBing Li1,2*, Saierjiang Halike1*, Xiyang Dai1, Simin He1, Roberto Giordano2,
Jean-Marc Tulliani3, Junsong Wang4, Qinglin Meng4
1School of Civil Engineering and Architecture, Xinjiang University, Urumqi 830047, China
2 Department of Architecture and Design, Politecnico di Torino, Torino 10125, Italy
3 Department of Applied Science and Technology, Politecnico di Torino, Torino 10129, Italy
4 State Key Laboratory of Subtropical Building and Urban Science, South China University of
Technology, Guangzhou 510641, China
A
Abstract
Chronic salt spray exposure induces microstructural evolution in historic building materials, critically altering their hygrothermal performance through modified pore networks and capillary dynamics. This study systematically investigates the humidity-regulated transition of the water vapor permeability coefficient (δv)—a key parameter governing moisture transfer in building envelopes—under marine aerosol conditions. Through accelerated salt spray cycling (35 cycles, 5% NaCl solution) combined with dry-cup/wet-cup measurements across contrasting humidity gradients (0→50% vs. 50→98% RH), cement mortar exhibited distinctly opposing δv responses: a 31.3% reduction under low humidity due to crystalline pore blockage, contrasting with a 239.4% enhancement at high humidity caused by deliquescence-induced brine migration. A predictive piecewise model (R² >0.95) based on salt influence factors and critical humidity thresholds was developed. Mercury intrusion porosimetry (MIP) and scanning electron microscopy (SEM) analyses revealed NaCl crystallization preferentially occupying 0.02–0.12 µm pores within the top 2 mm, reducing effective porosity by 42% while establishing two distinct transport modes: pore occlusion by crystals (dry state) and interconnected brine networks (humid state). These findings provide a mechanistic framework for predicting moisture dynamics in porous building materials under salt spray environments, enabling accurate hygrothermal simulations that improve resilience and durability predictions for coastal historic buildings.
Keywords:
Coastal Climate
Cement Mortar
Salt Crystallization
Water Vapor Permeability
Heritage Conservation
1. Introduction
An axiom in Chinese heritage conservation holds that "dry conditions preserve for millennia, constantly wet conditions for centuries, but alternating dry-wet conditions cause decay within years." In natural environments, building materials are subjected to long-term cyclic hygrothermal loads, during which moisture undergoes repeated migration, storage, and phase change within porous structures [1–3]. These processes often lead to structural damage—such as cracking, strength loss, surface weathering, color fading, and biological growth—in historic buildings [4–9]. Accurate prediction and quantification of moisture transport in porous materials are therefore essential for enhancing the resilience and durability of built heritage. Coastal regions, which cover approximately 18.43% of the world's land area, are home to about 52.8% of the global population [9]. Benefiting from well-developed transportation, communication, infrastructure, and economic activity [10–12], coastal and island cities also concentrate a significant portion of the world’s cultural heritage and historic architecture [13–16]. In these environments, long-term exposure to marine salt spray induces salt crystallization in conventional building materials such as brick, mortar, and plaster, fundamentally altering pore structures and disrupting moisture transport mechanisms [17–19]. Under cyclic moisture and salt action, coupled heat and moisture transport occurs within porous materials, significantly affecting their physical properties and functional performance, thereby posing considerable challenges to building envelopes [2, 17, 20–22]. A thorough investigation and systematic interpretation of these processes are essential for advancing conservation techniques and sustainable management strategies for coastal historic structures.
The deposition of salt crystals alters the original pore characteristics of porous materials by occupying their pore volume. Recent studies have elucidated the mechanisms of salt crystallization and growth within porous building materials. Systematic laboratory analyses of common construction materials – including brick [23], sandstone [24], limestone [3], plaster [25], and mortar [26] – confirm that salt deposition fundamentally alters porosity and pore architecture. The two most commonly used salts in experiments are Na₂SO₄ and NaCl. Mercury intrusion porosimetry (MIP) data indicate that Na₂SO₄ reduces porosity by 6–8% [19, 23, 27] whereas NaCl can lead to more substantial reductions, with reported values ranging from 4–10% [25, 28] up to 47–50% under specific conditions [19, 27]. Collectively, these studies reveal a general trend of porosity reduction due to salt crystallization. However, substantial inconsistencies persist across experimental results, which may arise from variable factors including salt speciation (e.g., crystalline polymorphism), solution molarity, salt exposure protocol, sampling location, and material-specific pore heterogeneity [3, 18, 19, 24, 25, 29]. The combined effects of saline environments and multiple factors result in distinct pore distribution characteristics in porous materials, thereby altering their inherent moisture transport behavior.
In porous building materials, salt crystallization and moisture transport constitute a dynamically coupled process. Salt crystallization within pores alters the hygric parameters of porous building materials—such as capillary absorption coefficients, capillary moisture retention capacity, and vapor permeability coefficients—through the dual mechanisms of pore occlusion and impaired pore connectivity [18, 19, 30, 31]. Conversely, as salt can only migrate in dissolved form through the pore network, its redistribution is intrinsically moisture-dependent, establishing a mutually causal relationship between moisture transport pathways and salt deposition patterns [29, 30, 32–34]. This bidirectional coupling introduces unique challenges in coastal saline environments, where dynamic pore-scale interactions between moisture flux and salt crystallization give rise to emergent system-level behavior. Further complicating the issue, methodological variations across studies—stemming from differing research objectives and experimental protocols—have hindered consensus on the evolution of hygric properties in salt-affected porous materials [3, 18, 19, 27, 29, 34–36]. Consequently, current understanding remains fragmentary, highlighting the necessity of a unified theoretical framework to reconcile experimental inconsistencies and reliably predict material performance in salt-rich environments.
Focusing on moisture transport mechanisms in porous building materials under coastal climates, our preliminary semi-immersion experiments revealed critical relationships between chloride migration and crystallization patterns in cement mortar [30]. The study demonstrated that surface salt crystallization morphology is governed by water-to-cement ratio (W/C)-controlled porosity, while chloride ion distribution along horizontal windward and vertical orientations is indirectly influenced by surface evaporation conditions. Notably, Salt crystallization deposition reduced the capillary absorption coefficient (Acap) by 46% and capillary moisture content (wcap) by 56% in specimens with W/C = 0.6 [30]. To simulate persistent coastal salt spray exposure, we conducted accelerated salt spray tests on cement mortar specimens. After 35 cycles, these tests showed a chloride concentration of 1.99%, an apparent density increase of 3.85%, reductions of 18.7% in both open porosity (P0) and saturation moisture content (wsat), as well as decreases of 30% in Acap and 17.4% in wcap [31]. Through parametric analysis of salinity effects on capillary absorption, we established modified computational equations for Acap and wcap in cement mortar under salt spray deposition [31]. These results provide comprehensive insights into the liquid water transport and storage behavior of cement mortar in coastal saline environments.
Building on previous research, this study systematically investigates the impact of salt deposition on water vapor transport in cement mortar through a multi-scale experimental framework. By combining accelerated salt spray tests with gravimetric analysis, we quantified the effect of NaCl deposition on vapor permeability under different humidity conditions using dry-cup and wet-cup methods. Through comparison with salt-free reference specimens, we established a normalized influence factor for the water vapor permeability coefficient, enabling mathematical modeling of its relationship with salt content. Microstructural analyses using mercury intrusion porosimetry (MIP) tracked pore evolution over 35 salt spray cycles, while scanning electron microscopy (SEM) with depth-resolved sampling revealed spatial salt crystallization patterns from surface to interior. This integrated multi-scale methodology bridges microscale salt-pore interactions with macroscale moisture transport phenomena, providing a fundamental basis for predicting the hygrothermal performance and structural resilience of historic buildings in coastal salt spray environments.
Nomenclature | |||
|---|---|---|---|
A | Exposed specimen surface area (m2) | Rsample | Water vapor resistance of the specimen (m2 s Pa/kg) |
Acap | Capillary absorption coefficient kg/(m2 s0.5) | Rtotal | Water vapor resistance of the sample and internal air layer of the device (m2 s Pa/kg) |
C | Salt content of specimens (kg/kg) | wcap | Capillary moisture content (kg/m3) |
dair | Thickness of air layer inside the device (m) | wsat | Vacuum-saturated moisture content (kg/m3) |
gv | Water vapor flux density (kg/m2/s) | δv | Water vapor permeability coefficient (kg/(mžsžPa)) |
Water vapor flux rate (kg/s) | δv,air | Water vapor permeability coefficient of the static air layer (kg/(mžsžPa)) | |
H | Specimen thickness (m) | δv,control | Water vapor permeability coefficient of control group specimens (kg/(mžsžPa)) |
Δpv | Water vapor partial pressure difference (Pa) | δv,s | Water vapor permeability coefficient of salt-laden specimens (kg/(mžsžPa)) |
P0 | Open porosity (%) | Influence factor of water vapor permeability coefficient | |
pv,sat | Saturated water vapor partial pressure (Pa, 2 808 Pa at 23℃) | φ1 | Relative humidity (higher-side) (%) |
Rair | Water vapor resistance of air layer inside the device (m2 s Pa/kg) | φ2 | Relative humidity (lower-side) (%) |
2. Materials and Methods
2.1 Specimen Preparation
Cement mortar was prepared using CEM II/B-LL 32.5R cement (Buzzi, Italy), CEN-standard sand (UNI EN 196-1:2011), and deionized water. The binder composition complied with EN 197-1, consisting of 65–79% clinker, 21–35% limestone, and 0–5% gypsum. The chemical composition of the cement, determined by X-ray fluorescence spectrometry (XRF; Rigaku NEX CG), is provided in Table 1, loss on ignition (LOI) was measured via thermogravimetric-differential thermal analysis (TG-DTA). Constituents were mixed at a mass ratio of 1:3:0.5 (cement : sand : water), homogenized, and cast into 70 × 70 × 70 mm³ molds under vibratory compaction. After initial curing at 23°C and 95% RH for 24 h, specimens were demolded and subsequently water-cured for 28 days at 23°C under laboratory ambient conditions.
For water vapor permeability evaluation, plate specimens measuring 70 × 70 × 20 mm³ were precision-cut from the cured cubes using a diamond saw (Fig. 1-1a). All surfaces were then cleaned sequentially with compressed air and deionized water to r remove residual particulates and ensure open pore networks. The specimens were then dried, labeled, and grouped for subsequent testing. The selected thickness of 20 mm conforms to the ISO 12572:2016 specification for homogeneous materials [37], achieving a balance between test duration and methodological validity.
Component | CaO | MgO | SO3 | Al2O3 | Fe2O3 | K2O | Na2O | TiO2 | SrO | P2O5 | LOI | Others |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Content | 60.2 | 3.56 | 3.46 | 3.12 | 2.19 | 1.12 | 1.05 | 0.185 | 0.147 | 0.104 | 12.8 | 0.19 |
2.2 Accelerated Salt Spray Cycling Test
Salt deposition and corrosion in building materials constitute a long-term process influenced by numerous uncontrollable climatic variables, making field experiments particularly challenging. To investigate moisture transport characteristics in porous building materials under coastal salt fog conditions, this study employed accelerated salt spray testing to simulate realistic exposure environments. While existing international salt spray test standards for building materials—such as BS EN 14147:2003(E) [38] and ASTM B117-11(E) [39]—are applicable to natural stone and primarily focus on durability assessment, this study referenced the fundamental test procedures specified in these standards and made corresponding adjustments to experimental parameters based on research objectives.
The experiment was conducted in a CEAST 5050 chamber (Italy), which generated salt fog using a 5 wt% NaCl solution (Fig. 1–2, Fig. 1–4). Specimens were mounted at a 20° inclination from the vertical with 44 mm spacing between units (Fig. 1–3). Each test cycle consisted of 8 h of spraying (35°C, 100% RH) followed by 16 h of drying (55°C), achieving a salt deposition rate of 2 mL/(80 cm²·h) in accordance with ASTM B117-11. This wet-dry cycling improves simulation accuracy by mimicking natural environmental fluctuations. A total of 24 specimens were divided into six groups (5O, 5A–5E) corresponding to 0, 7, 14, 21, and 35 exposure cycles, with four replicate specimens per group (e.g., 5A1–5A4). After exposure, all specimens were conditioned at 23°C and 50% RH until mass stabilization (ISO 12572:2016 [37]) to standardize moisture content before permeability testing..
2.3 Water Vapor Permeability Measurement
The water vapor permeability of cement mortar was evaluated following ISO 12572: 2016(E) [37] using a custom-designed apparatus (Fig. 1(4 − 1)) to establish controlled vapor pressure gradients across specimens. Specimens (70 × 70 × 20 mm³) were edge-sealed with self-adhesive aluminum tape (Fig. 1( 3 − 1)) to prevent sealant penetration into surface pores, then bonded to modified plastic container lids (72 × 72 mm² openings) using Bostik liquid rubber sealant (Italy). This ensured mid-plane alignment (10 mm exposed above/below the lid) and void-free interfacial contact (Fig. 1(3 − 2)). Airtightness was verified by submerging specimens in deionized water (2–3 mm below the top surface) for 6 h (Fig. 1(3–3)), any leaking specimens were resealed until integrity was confirmed.
A
Two test methods were employed to assess moisture transport under varying humidity conditions [37, 40]. The dry-cup method simulated low-humidity environments (0 ± 2% RH vs. 50 ± 2% RH, Fig. 1(Table)) using silica gel desiccant, whereas the wet-cup method evaluated high-humidity behavior (50 ± 2% RH vs. 98 ± 2% RH, Fig. 1(Table)) with saturated K₂SO₄ solution maintained with excess salt crystals. Key setup parameters included a 20 mm air layer between specimens and desiccant (dry-cup) and a 50 mm air layer above the solution (wet-cup), with solution depths standardized at 15 mm. All assemblies were preconditioned at 23 ± 0.5°C and 50 ± 2% RH until mass stabilization (variation < 5% over three consecutive days, Fig. 1(4 − 2)) to ensure uniform moisture distribution prior to permeability testing.Preconditioned assemblies were sealed onto containers with equal-mass of desiccant or salt solution (Fig. 1(4 − 3)) and transferred to a climate-controlled chamber (23 ± 2°C, 50 ± 2% RH) (verified usingTesto 174H sensors, Fig. 1( 4–4)) with ≥ 100 mm/s airflow. Mass measurements commenced after 28 days, conducted at 48 h intervals under temperature-controlled conditions (23 ± 2°C) until five consecutive readings showed less than 5% deviation from the mean, confirming steady-state vapor flow. This protocol follows ISO 12572 requiements for airflow and equilibration while incorporating refinements such as air-tightness control and mass-equalized solutions to minimize variability.
The water vapor flux rate Gv (kg/s) through cement mortar specimens was determined via linear regression of the total mass of dry/wet cup assemblies against weighing time, with a coefficient of determination (R2) ≥ 0.99. The water vapor flow density gv (kg·m− 2∙s− 1) was calculated as:
gv =
(1)
Where A is the exposed surface area of specimen (m2). The vapor pressure difference Δpv (Pa) across the specimen was derived using:
∆pv = pv,sat |φ1 – φ2 | (2)
Where pv,sat = 2.808 Pa (saturation vapor pressure at 23°C), and ϕ1, ϕ2 are relative humidity values on the tow sides of the spcimen. The total vapor transfer resistance Rtotal (m2·s·Pa·kg− 1) was calculated as:
Rtotal =
(3)
The air layer resistance Rair (m2·s·Pa·kg− 1) within the apparatus was determined by:
Rair =
(4)
Where dair is the thickness of the air layer (m), and δv,air is the water vapor permeability coefficient of static air. The specimen’s water vapor permeability coefficient δv (kg·m− 1·s− 1·Pa−1) were obtained via:
Rsample = Rtotal-Rair (5)
δv =
(6)
Where Rsample is water vapor resistance of specimen (m2 s Pa/kg). H is the specimen thickness (m).
Fig. 1
Schematic of the experimental setup and protocol for water vapor permeability testing
2.4 Microstructural Characterization and Analysis
The chloride ion content in cement mortar subjected to accelerated salt spray cycles was quantified using a C-CL-3000 electrochemical analyzer (James Instruments, USA). Prior to analysis, the system was calibrated with five Cl⁻ standard solutions (0.005%, 0.01%, 0.05%, 0.1%, and 0.3%). Full-thickness specimens were were ground into powder to assess the depth-dependent distribution of salts. A 3 g aliquot of homogenized powder was dissolved in 20 mL acidic solution, vortexed, and allowed to stabilize for 2 min. After degassing (2–4 min), Cl⁻ concentrations were determined via potentiometric voltage differentials, with results displayed digitally. All measurements were performed at 23°C on desiccated specimens, with calibrated batches analyzed within 2 h to minimize instrumental drift.
To establish microstructure-property relationships, mercury intrusion porosimetry (MIP; Autopore V, Micromeritics, USA) and scanning electron microscopy (SEM; S4000, Hitachi, Japan) were utilized (Figs. <link rid="fig1">1</link>–2 and 1–3). SEM samples were prepared as cross-sectional profiles from surface to mid-depth region, dried and sputter-coated with an Au/Pd (SPI Supplies, USA). For MIP, samples were dried at 50°C to constant mass (Fig. 1–2); then, 3–4 g of each sample was analyzed under a maximum pressure of 2,000 MPa, enabling the detection of pores larger than 6 nm in diameter based on the Washburn theory. Two replicates per group ensured statistical validity. All procedures complied with ISO 21466:2019 [41] for SEM and ISO 15901-1:2016 [42] for MIP, and specimens were kept in anhydrous state throughout pretreatment and handling.
3. Results and Discussion
3.1 Dry-cup and Wet-Cup Water Vapor Permeability Coefficient
The mass variations of cement mortar specimens under dry-cup (RH 0/50%) and wet-cup (RH 50%/98%) conditions are illustrated in Figs. 2(a) and 2(b), respectively. In dry-cup tests (Fig. 2a), the assembly mass increased linearly over time as water vapor was absorbed by the internal silica gel desiccant. Conversely, wet-cup assemblies (Fig. 2b) lost mass linearly due to vapor diffusion from the saturated K₂SO₄ solution to the drier ambient environment. Although minor initial mass differences occurred among the control and salt-sprayed groups (7–35 cycles) resulting from material heterogeneity and slight thickness variations during specimen preparation, they did not affect the validity of mass change rate analyses. Measurement errorss across datasets ranged between 0.01%–0.03%, mainly caused by heterogeneous pore structures and salt crystallization patterns, with negligible influence on the overall results.
Linear regression of mass-time profiles (Figs. 2a, 2b) provided the water vapor flux rates (Gv) for each assembly (Tables 2 and 3). Dry-cup data showed coefficient of determination (R2) > 0.98 and residual sum of squares (RSS) < 6.00 × 10⁻⁷. Notably with the control group exhibiting the highest Gv, which decreased progressively with salt spray cycles (Table 2). Wet-cup results demonstrated superior linearity (R2 > 0.99, RSS < 5.89 × 10⁻⁵)), with the control group showing the lowest Gv— a value that increased systematically with more exposure cycles (Table 3).
Fig. 2
Mass variation of dry- and wet-cup specimens during water vapor permeability testing
Salt Spray Cycling | Dry Cup Assembly | Gv (kg/s) | R2 | RSS |
|---|---|---|---|---|
Reference | 5O1 ~ 5O4 | 2.06783 × 10 − 9 | 0.990 | 6.00332 × 10− 7 |
7 Cycles | 5A1 ~ 5A4 | 1.82618× 10 − 9 | 0.992 | 3.49465 × 10− 7 |
14 Cycles | 5B1 ~ 5B4 | 1.79315× 10 − 9 | 0.988 | 5.3501 × 10− 7 |
21Cycles | 5C1 ~ 5C4 | 1.72146× 10 − 9 | 0.993 | 2.75475 × 10− 7 |
28 Cycles | 5D1 ~ 5D4 | 1.49543× 10 − 9 | 0.981 | 5.83101 × 10− 7 |
35 Cycles | 5D1 ~ 5D4 | 1.32698× 10 − 9 | 0.993 | 1.55235 × 10− 7 |
Salt Spray Cycling | Dry Cup Assembly | Gv (kg/s) | R2 | RSS |
|---|---|---|---|---|
Reference | 5O1 ~ 5O4 | -2.34299 × 10− 9 | 0.999 | 6.85676 × 10− 7 |
7 Cycles | 5A1 ~ 5A4 | -3.94800 × 10− 9 | 0.997 | 1.04771 × 10− 5 |
14 Cycles | 5B1 ~ 5B4 | -4.75102 × 10− 9 | 0.997 | 1.42988 × 10− 5 |
21Cycles | 5C1 ~ 5C4 | -5.09070 × 10− 9 | 0.998 | 1.47054 × 10− 5 |
28 Cycles | 5D1 ~ 5D4 | -5.66033 × 10− 9 | 0.998 | 1.75064 × 10− 5 |
35 Cycles | 5D1 ~ 5D4 | -6.83262 × 10− 9 | 0.995 | 5.88554 × 10− 5 |
The water vapor permeability coefficient (δv) and associated process parameters for all cement mortar groups were determined using Equations (1)–(6), with results summarized in Tables 4 and 5. The specimen surface area (A) used for calculating vapor flux density (gv) was consistent across groups due to mold-controlled dimensions with negligible length/width deviations. To compensate for thickness (H) variations introduced during the cutting process, the reported thickness values represent the average of six measurements taken at different locations using a vernier caliper.
As shown in Table 4 (dry-cup method, 0–50% RH), the control group exhibited the highest vapor permeability coefficient δv (5.95×10⁻¹² kg/(m·s·Pa)). This value decreased progressively with increasing salt spray exposure, reaching its lowest level after 35 cycles. In contrast, Table 5 (wet-cup method, 50–98% RH) revealed an opposite trend: the control group displayed a lower baseline δv (8.25×10⁻¹² kg/(m·s·Pa)) which increased systematically over 35 exposure cycles. This dichotomy underscores the dual-phase influence of salt crystallization : under low humidity, pore-blocking effects dominates, reducing δv, whereas under high humidity, deliquescence-induced brine formation enhances moisture transport through capillary action.
Salt Spray Cycling | Gv (kg/s) | [kg/(m2žs)] | (Pa) | (m2žsžPa/kg) | (m2žsžPa/kg) | (m2žsžPa/kg) | [kg/(mžsžPa)] |
|---|---|---|---|---|---|---|---|
Reference | 2.07 × 10− 9 | 4.22 × 10− 7 | 1404 | 3.33 × 10− 9 | 1× 108 | 3.23 × 109 | 5.95 × 10–12 |
7 Cycles | 1.83 × 10− 9 | 3.73 × 10− 7 | 1404 | 3.77 × 10− 9 | 1× 108 | 3.67 × 109 | 5.78 × 10–12 |
14 Cycles | 1.79 × 10− 9 | 3.66 × 10− 7 | 1404 | 3.84 × 10− 9 | 1× 108 | 3.74 × 109 | 5.66 × 10–12 |
21Cycles | 1.72 × 10− 9 | 3.51 × 10− 7 | 1404 | 4.00 × 10− 9 | 1× 108 | 3.90 × 109 | 5.17 × 10–12 |
28 Cycles | 1.50× 10− 9 | 3.05 × 10− 7 | 1404 | 4.60 × 10− 9 | 1× 108 | 4.50 × 109 | 4.54 × 10–12 |
35 Cycles | 1.33× 10− 9 | 2.71 × 10− 7 | 1404 | 5.18 × 10− 9 | 1× 108 | 5.08 × 109 | 4.09 × 10–12 |
Salt Spray Cycling | Gv (kg/s) | [kg/(m2žs)] | (Pa) | (m2žsžPa/kg) | (m2žsžPa/kg) | (m2žsžPa/kg) | [kg/(mžsžPa)] |
|---|---|---|---|---|---|---|---|
Reference | -2.34 × 10− 9 | 4.78 × 10− 7 | 1347.84 | 2.82 × 10− 9 | 2.5× 108 | 2.57 × 109 | 8.25 × 10–12 |
7 Cycles | -3.95 × 10− 9 | 8.06 × 10− 7 | 1347.84 | 1.67 × 10− 9 | 2.5× 108 | 1.42 × 109 | 1.340 × 10–11 |
14 Cycles | -4.75 × 10− 9 | 9.70 × 10− 7 | 1347.84 | 1.39 × 10− 9 | 2.5× 108 | 1.14 × 109 | 1.88 × 10–11 |
21Cycles | -5.28 × 10− 9 | 1.08 × 10− 6 | 1347.84 | 1.25 × 10− 9 | 2.5× 108 | 1.00 × 109 | 2.10 × 10–11 |
28 Cycles | -5.66 × 10− 9 | 1.16 × 10− 6 | 1347.84 | 1.17 × 10− 9 | 2.5× 108 | 9.17 × 108 | 2.38 × 10–11 |
35 Cycles | -6.83× 10− 9 | 1.39 × 10− 6 | 1347.84 | 9.67 × 10− 9 | 2.5× 108 | 7.17× 108 | 2.80 × 10–11 |
Xie et al. [43] reported a water vapor permeability coefficient (δv) in the range of 2.66×10-12 ~ 4.02×10།12 kg/(m·s·Pa) for cement mortar under 43%–97.6% RH, which is slightly lower than the values obtained in this study. This discrepancy likely stems from differences in cement type and mix proportions. Notably, their findings are consistent with the present results: salt deposition suppresses vapor transport under low humidity/high salinity conditions but enhances it under high humidity/elevated salinity environments (Tables 4–5, Fig. 3).
The opposing effects of salt deposition on δv under diffrent humidity conditions are governed by the deliquescence behavior of NaCl. Below 75% RH (deliquescence threshold at 23°C [44]), accumulated salt crystals reduce porosity and pore connectivity, increasing vapor transport resistance [19, 28, 45–47]. Above 75% RH, deliquesced NaCl migrates as brine via capillary action and evaporation-driven flows [19, 29, 30, 48]. In the wet-cup configuration, where the low-humidity side is maintained at 50% RH, the brine releases vapor until recrystallization occurs [18, 49], inducing intense liquid-phase mass transport that elevates δv. Morover, the intrinsic hygroscopicity of NaCl futher enhaces moisture migration as salt content increases.
Fig. 3
Water Vapor Permeability Coefficients versus Salt Content in Cement Mortar
3.2 Salt-Induced Permeability Model
To quantify the effect of salt content on the water vapor permeability coefficient (δv) of cement mortar under salt spray exposure, the correction factor
was calculated for both dry-cup (ϕ ≤ 0.75) and wet-cup (ϕ > 0.75) methods using Eq. (7). This parameter integrates the mechanistic influence of NaCl crystallization below 75% RH and deliquescence above 75% RH on vapor transport properties. Polynomial regression was used to express
as a function of salt content in the low-humidity (ϕ ≤ 0.75) and high-humidity (ϕ > 0.75), yielding coefficients of determination (R2) greater than 0.96 and 0.95, respectively (Figs. 4a and 4b). This resulting dual-phase model reflects the crystallographic and hygroscopic transitions of NaCl, confirming its role in regulating vapor transport across humidity gradients.
7
where
and
are the water vapor permeability coefficients kg/(mžsžPa) of salt-containing specimens and salt-free specimens, respectively.
Fig. 4
Influence factor (
for Water Vapor Permeability as a Function of Salt Content a) φ1 ≤ 0.75 b) φ1 > 0.75
The water vapor permeability coefficient for salt-free porous materials (
) was derived from Equations (1)–(6) and expressed in Eq. (8). All symbols and units in this expression align with those defined in Section 2.3. To incorporate the influence of salt deposition,
was then modified by applying the polynomial fit for the influence factor
(Fig. 4), yielding the adjusted permeability coefficient
for NaCl- contaminated cement mortar, as expressed in Eq. (9):
8
9
The water vapor permeability coefficient (δv) is a key parameter for evaluating hygrothermal transfer in building envelopes and estimating energy consumption. Accurately determining δv for salt-laden porous building materials in coastal salt spray environments is of considerable practical relevance. Conventional measurement of δv across diverse materials is bothe labor-intensive and time-consuming. By contrast, the correction factor approach (Eq. 9) enables efficient adjustment of previously measured δv values for defferent salt contents, substantially reducing the need for repeated testing.
Cement mortar, as a widely used porous material in historic building envelopes, exhibits no chemical reactivity with NaCl [19, 30, 31]. Its salt and moisture transport behavior, along with associated physical transformations, are primarily controlled by pore structure characteristics. Thus, the findings presented here offer valuable insights for other historic building materials—such as manufactured bricks and natural stone—where salt-induced damage occurs mainly through physical mechanisms. It should be noted that this study focuses exclusively on NaCl, the predominant salt in seawater; the effects of other marine salts (e.g., MgCl₂, CaSO₄) and salt mixtures on δv remain unexamined and warrant further investigation to advance predictive models and conservation strategies for coastal historic structures.
3.3 Pore Structure and microscopic morphology
The water vapor permeability of porous building materials is primarily controlled by their pore structure. In cement mortar, salt crystallization within pores represents the key factor altering the permeability coefficient. To investigate the microstructural origins of these changes in salt-containing specimens, this study employed mercury intrusion porosimetry (MIP) to quantify pore parameters, combined with scanning electron microscopy (SEM) for morphological characterization.
3.3.1 MIP Porosimetry
To examine microstructural changes in cement mortar under salt spray exposure, pore size distributions of specimens subjected to 7–35 salt spray cycles were systematically compared with those of reference specimens (Figs. 5a–e). The dominant pore volume peak in cement mortar was observed at approximately 0.1 µm. Under cyclic salt spray conditions, pore volumes within the 0.02–0.12 µm range exhibited a progressive decline (16–42%) with increasing exposure cycles, accompanied by a distinct shift of the pore volume peak toward larger pore sizes. These observations confirm that NaCl crystallization preferentially occupies smaller pores (< 0.1 µm), consistent with earlier reports on NaCl deposition in cement mortar (0.01–0.08 µm) [27] and limestone (0.01–0.12 µm) [50]. Since gas diffusion primarily occurs in pores > 10⁻³ µm [51], the accumulation of salt crystals progressively impedes vapor transport in dry-cup tests, resulting in a systematic reduction of permeability coefficients (Table 4, Fig. 3). In contrast, wet-cup tests revealed an opposing mechanism: deliquescence of salt crystals under high humidity alleviates pore blockage, while concurrent brine-vapor migration through interconnected pores enhances moisture transport efficiency, thereby increasing water vapor permeability coefficients (Table 5, Fig. 3). This humidity-dependent dual behavior underscores the two roles of NaCl—pore occlusion by crystallization under low humidity (dry-cup) versus transport enhancement through deliquescence under high humidity (wet-cup).
For comparative analysis with immersion-based methods [35, 43], specimens immersed in 2.39 wt% NaCl solution (simulating seawater salinity) for 48 h were also examined (Fig. 5f). The reduction in pore volume within the 0.02–0.12 μm range due to immersion was similar to that observed in specimens subjected to 7/14 salt spray cycles (Fig. 5a), suggesting comparable pore-filling mechanisms between salt spray deposition and solution immersion. This similarity persists despite differences in salt delivery (surface deposition vs. bulk infiltration), supporting the general applicability of NaCl-induced pore structure modifications in porous matrices.
Fig. 5
Pore Size Distribution of Cement Mortar with Salt Deposition
(a–e: 7–35 Salt Spray Cycles; f: 48 h Salt Solution Immersion)
Specific surface area, defined as the total surface area per unit mass of a material, encompasses both external and internal surfaces in porous systems, with its magnitude governed by pore morphology and surface characteristics. Together with pore volume distribution, it fundamentally defines a material’s microstructure and governs its macroscopic properties [52, 53]. To track pore structure evolution in cement mortar under salt spray exposure, cumulative pore volume and specific surface area were measured using mercury intrusion porosimetry (MIP) for reference specimens and specimens exposed to 7–35 salt spray cycles (Fig. 6). Reference specimens showed the highest values, with a cumulative pore volume of 0.069 m³/g and a specific surface area of 5.218 m²/g. Both parameters decreased progressively as exposure cycles increased—a trend attributed to the substantial filling of 0.02–0.12 µm pores by deposited NaCl crystals (Fig. 5), as corroborated by earlier analysis.
Fig. 6
Cumulative pore volume and specific surface area of cement mortar subjected to
varying salt spray cycles
The reduction in specific surface area observed in salt-containing specimens arises from two principal mechanisms. First, partial filling of micropores by salt crystals converts them into denser solid structures, thereby diminishing the internal surface area. Second, unlike Na₂SO₄ crystals that typically form along pore walls, NaCl nucleates at liquid–gas interfaces and grows toward the pore walls, without covering the internal pore surface until contact is made [19, 28, 50, 54]. Although crystallization increases the total internal surface area in absolute terms, it also raises the specimen mass. Given that dry NaCl crystals have a density of 2,165 kg/m³ at 23°C under standard pressure [19], —higher than the measured density of cement mortar (2,133 kg/m³)—and considering the non-porous nature of NaCl relative to the porous mortar matrix, salt deposition ultimately reduces the specific surface area of the composite material.
It is noteworthy that although NaCl deposition reduces specific surface area (m²/g), it increases the absolute external surface area per unit volume (m²/m³) within pore space. This extended interface enhances gas-solid contact, thereby raising flow resistance. As a result, the observed decrease in water vapor permeability with increasing salt content below 75% RH (Fig. 2, Table 2, Table 4), can be explained by the presence of undissolved salt crystals, which impede vapor transmission through surface adsorption mechanisms.
3.3.2 SEM Morphology Analysis
To visualize microstructural evolution under cyclic salt spray exposure, scanning electron microscopy (SEM) observations were performed on cement mortar specimens across all salt spray cycles (Fig. 7). After 7 cycles, cubic NaCl crystals densely covered the surface (Fig. 7 7a, 7b), resulting in noticeable surface roughening (Fig. 6, 7c). By 14 cycles, crystal density increased, with particles exhibiting more irregular shapes and smaller sizes, forming thicker coatings (Fig. 7 14a–14c). At 21 cycles, granular crystals transitioned to interconnected flake-like layers (Fig. 7 21a–21c). After 28 cycles, clusters crystal aggregates developed (Fig. 7 28a–28c), and by 35 cycles, salt had diffused inward into subsurface pores, with fine crystals adhering to larger molten clusters (Fig. 7 35a–35c).
Salt crystallization morphology and distribution evolve dynamically through repeated deposition, crystallization, dissolution, and redistribution during wet-dry cycling. Since water vapor transport occurs primarily through open surface pores, changes in near-surface pore density and connectivity critically influence vapor permeability. The observed progression—from sparse to dense coverage and from thin films to thick crystalline layers—demonstrates how NaCl deposition progressively occludes surface pores, reducing open porosity by 18–32% and thereby impeding vapor transport [19, 30]. These microstructural transformations directly corroborate the 22–45% reduction in water vapor permeability coefficients measured via dry-cup tests (Table 4, Fig. 3), establishing a causal relationship between morphological evolution and macroscopic transport properties.
Fig. 7
Surface salt crystallization morphology of cement mortar under varying salt spray cycles (SEM)
A
A
A
A
A
To comprehensively evaluate microstructural changes in cement mortar induced by salt spray exposure, cross-sectional morphologies at different depths were examined and compared with reference specimens (Fig. 8). Reference samples (Ra–Rd) exhibited characteristic cementitious phases, including C-S-H gel networks, needle-like ettringite (3CaO.Al₂O₃.3CaSO₄.32H₂O), and rod/plate-shaped portlandite. After 7–14 salt spray cycles, localized but dense NaCl crystals formed within interfacial voids (Fig. 8 7a–14c). With further cycling (21–35 cycles), salt clusters proliferated, occluding fine pores while generating larger macropores with smoother surfaces (Fig. 8 28c–35c). These observations confirm that salt deposition modifies both surface and subsurface pore networks. NaCl migration operates through humidity-driven dissolution-recrystallization dynamics: deliquescence under humid conditions enables brine infiltration into deep pores via capillary action [29, 30], while drying cycles drive internal brine toward shallow regions, culminating in recrystallization and pore architecture modification (Fig. 5a-f). The progressive accumulation of salt crystals corresponds to the persistent decrease in 0.02–0.12 µm pore volumes observed during salt spray testing (Fig. 5). Moreover, higher salt content amplifies hygroscopic moisture uptake and liquid-phase transport under elevated humidity. Repeated crystallization and dissolution cycles continuously reshape the pore network, providing a mechanistic explanation for the contrasting water vapor permeability behaviors measured in dry- and wet-cup tests (Tables 4–5).Fig. 8
Internal Morphology of Cement Mortar under Varying Salt Spray Cycles (SEM)
4. Conclusions
This study systematically investigated the evolution of water vapor permeability in in cement mortar, a representative historic building material, under coastal salt spray environments via integrated accelerated testing (35 cycles, 5% NaCl solution), dry-cup/wet-cup measurements, and microstructural characterization. The major findings are summarized as follows:
(1) Our results establish a critical humidity-regulated duality in salt-permeability interactions: below NaCl's 75% RH deliquescence threshold, increasing salt content reduces water vapor permeability coefficients (δv) by 3–31% through crystalline pore occlusion, whereas above this threshold, δv increases 62–239% via deliquescence-induced brine films that facilitate molecular diffusion through connected liquid pathways. This phase-transition-driven reversal fundamentally redefines moisture transport prediction in saline environments.
(2) To quantify these mechanisms, we developed a generalized piecewise correction framework using polynomial regression. The model incorporates humidity-specific influence factors —ηδv = (-2983.71C2 + 59.12C + 0.68) for φ ≤ 0.75 and ηδv = (6035.16C2-28.26C + 1.51) for φ > 0.75 (where C is salt content in kg/kg)—which scale baseline permeability (δv, control) to predict saline mortar behavior. Validation against experimental data confirmed its predictive accuracy (R2 ≥ 0.95), enabling direct implementation in hygrothermal simulation tools.
(3) Microstructural analyses (SEM/EDS) revealed progressive salt crystallization evolution: initial surface-deposited cubic crystals (7–14 cycles) transition to interconnected subsurface clusters (21–35 cycles), preferentially occupying 0.02–0.12 µm pores and reducing near-surface porosity by 16–42%. This microstructural transformation directly explains the δv suppression observed in dry-cup tests. Simultaneously, deliquescence-generated capillary brine networks above 75% RH critically enhance vapor diffusion in wet-cup configurations, with salt redistribution depths exceeding 2 mm after prolonged exposure.
These findings provide the first mechanistic framework quantifying the coupling between salt content, humidity, and permeability in porous building materials, enabling accurate prediction of moisture-related parameters and physical properties for coastal built heritage. While validated for NaCl (constituting approximately 69% of seawater ions), the physical basis of the model suggests applicability to other porous heritage construction materials where salt interactions are predominantly physical rather than chemical. Future work will extend this framework to multi-salt systems using synchrotron tomography and implement the correction factors in WUFI-Pro for whole-building hygrothermal simulation.
Declaration of Competing Interest
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.
A
Funding
This work was supported by the National Natural Science Foundation of China (51938006, 52468004), the State Key Laboratory of Subtropical Building and Urban Science (2022ZC02), the China Scholarship Council (202206150001), and the Natural Science Foundation of Xinjiang Uygur Autonomous Region (No. 2025D01C12).
A
Acknowledgement
The authors gratefully acknowledge the SISCON Laboratory at Politecnico di Torino for supporting the experimental activities. Special thanks are extended to Prof. Raffaella Sesana for providing access to the salt spray chamber, as well as to Dr. Leonardo Iannucci and Prof. Sabrina Grassini for their technical assistance in fabricating the 3D-printed specimen holders.
A
Author Contribution
Bing Li: Writing – Original Draft, Investigation, Formal AnalysisSaierjiang Halike: Writing – Review & EditingXiyang Dai: Data Curation,VisualizationSimin He: Methodology, ValidationRoberto Giordano: Conceptualization, SupervisionJean-Marc Tulliani: Resources, InvestigationJunsong Wang: Project AdministrationQinglin Meng: Funding Acquisition, Supervision
References
1.
Zhao J, Feng S, Kou Z, et al. Characterization of hygrothermal properties of two wood species- the impact of anisotropy on their thermal and moisture behaviors. Constr Build Mater. 2023;398:132375.
2.
d'Ambrosio FR, Alfano BI, Palella G, Riccio. Moisture in historical buildings from causes to the application of specific diagnostic methodologies. J Cult Herit. 2023;61:150–9.
3.
Koronthalyova O. Moisture storage capacity and microstructure of ceramic brick and autoclaved aerated concrete. Constr Build Mater. 2011;25(2):879–85.
4.
Pinheiro VRF, Fontenele R, Magalhães A, et al. Evaluation of the influence of climatic changes on the degradation of the historic buildings. Energy Build. 2024;323:114813.
5.
Karoglou M, Moropoulou A, Mouzakis C, et al. Monitoring of a historic building's masonry moisture content. J Building Eng. 2025;104:112313.
6.
Sheng L, Xie H, Ma Y, et al. Assessing the deterioration risk of polychrome clay sculptures based on the hygrothermal environment: A case study of Baosheng temple, China. Case Stud Constr Mater. 2022;17:e01287.
7.
Gaylarde CC, Gutarowska B, Beech IB, et al. Metagenomic and morphological analysis of biodeteriogenic biofilms on historic stone buildings in the tropical north-east of Brazil. Int Biodeterior Biodegrad. 2025;204:106143.
8.
Blumberga A, Freimanis R, Muizniece I, et al. Trilemma of historic buildings: Smart district heating systems, bioeconomy and energy efficiency. Energy. 2019;186:115741.
9.
Lubelli B, Moerman J, Esposito R et al. Influence of brick and mortar properties on bioreceptivity of masonry – Results from experimental research. Constr Build Mater (2020).
10.
Jin X, Luan W, Yang J, et al. From the coast to the interior: global economic evolution patterns and mechanisms. Humanit Social Sci Commun. 2023;10(1):723.
11.
He F, Yang J, Zhang Y, et al. Offshore Island Connection Line: A new perspective of coastal urban development boundary simulation and multi-scenario prediction. GIScience Remote Sens. 2022;59(1):801–21.
12.
Cosby AG, Lebakula V, Smith CN, et al. Accelerating growth of human coastal populations at the global and continent levels: 2000–2018. Sci Rep. 2024;14(1):22489.
13.
Wei G, Han G-S, Lang X. Cultural heritage site selection characteristics and the impact of the natural environment in Jinan city. China Sci Rep. 2024;14(1):31038.
14.
Hu H. Climate and environmental dynamics: deciphering the distribution and vulnerability of world heritage sites in Europe. J Environ Manage. 2025;392:126693.
15.
Morillas H, Maguregui M, Gallego-Cartagena E, et al. The influence of marine environment on the conservation state of Built Heritage: An overview study. Sci Total Environ. 2020;745:140899.
16.
Morillas H, Vazquez P, Maguregui M, et al. Composition and porosity study of original and restoration materials included in a coastal historical construction. Constr Build Mater. 2018;178:384–92.
17.
Morillas H, de Mendonca Filho FF, Derluyn H, et al. Decay processes in buildings close to the sea induced by marine aerosol: Salt depositions inside construction materials. Sci Total Environ. 2020;721:137687.
18.
Koronthalyova O, Bagel L, Kuliffayova M, et al. Effect of Presence of Salt on Hygric Performance of Ceramic Bricks. Transp Porous Media. 2015;107(3):667–82.
19.
Todorović J, Janssen H. The impact of salt pore clogging on the hygric properties of bricks. Constr Build Mater. 2018;164:850–63.
20.
Huang D, Niu D, Su L et al. Diffusion behavior of chloride in coral aggregate concrete in marine salt-spray environment. Constr Build Mater 316 (2022).
21.
Delgado JMPQ, Guimarães AS, de Freitas VP, et al. Salt Damage and Rising Damp Treatment in Building Structures. Adv Mater Sci Eng. 2016;2016:1–13.
22.
Pozo-Antonio JS. Evolution of mechanical properties and drying shrinkage in lime-based and lime cement-based mortars with pure limestone aggregate. Constr Build Mater. 2015;77:472–8.
23.
Hendrickx R, Clercq HCD, Roels S et al. Experimental investigation of the influence of precipitated salts on the liquid transport properties of brick using an organic fluid, 2011.
24.
Angeli M, Benavente D, Bigas J-P, et al. Modification of the porous network by salt crystallization in experimentally weathered sedimentary stones. Mater Struct. 2008;41(6):1091–108.
25.
Lubelli B, van Hees RPJ, Groot CJWP. Sodium chloride crystallization in a salt transporting restoration plaster. Cem Concr Res. 2006;36(8):1467–74.
26.
Koniorczyk M. Salt transport and crystallization in non-isothermal, partially saturated porous materials considering ions interaction model. Int J Heat Mass Tran. 2012;55(4):665–79.
27.
Koniorczyk M, Gawin D. Heat and Moisture Transport in Porous Building Materials Containing Salt. J Building Phys. 2008;31(4):279–300.
28.
Espinosa-Marzal RM, Scherer GW. Impact of in-pore salt crystallization on transport properties. Environ Earth Sci. 2013;69(8):2657–69.
29.
Pel L, Huinink H, Kopinga K. Salt transport and crystallization in porous building materials. Magn Reson Imaging. 2003;21(3–4):317–20.
30.
Li B, Meng Q, Tulliani J-M et al. Salt migration and capillary absorption characteristics of cement mortar partially immersed in NaCl solution. J Building Eng 64 (2023).
31.
Li B, Meng Q, Giordano R et al. The hygric properties of cement mortar with salt spray deposition. Constr Build Mater 449 (2024b).
32.
Su L, Niu D, Huang D et al. Chloride diffusion behavior and microstructure of basalt-polypropylene hybrid fiber reinforced concrete in salt spray environment. Constr Build Mater 324 (2022).
33.
Franzoni E, Gentilini C, Santandrea M, et al. Effects of rising damp and salt crystallization cycles in FRCM-masonry interfacial debonding: Towards an accelerated laboratory test method. Constr Build Mater. 2018;175:225–38.
34.
Pel L, Huinink H, Kopinga K. Ion transport and crystallization in inorganic building materials as studied by nuclear magnetic resonance. Appl Phys Lett. 2002;81(15):2893–5.
35.
Bai L, Xie J, Liu J et al. Effect of salt on hygroscopic properties of cement mortar. Constr Build Mater 305 (2021).
36.
Xie Y, Xie J, Bai L et al. Experimental study on the effect of salt on the water absorption characteristic of cement mortar. J Building Eng 73 (2023).
37.
ISO 12572. 2016(E) Hygrothermal performance of building materials and products — Determination of water vapour transmission properties — Cup method 2016.
38.
BS EN 14147. 2003(E) Natural stone test methods. Determination of resistance to ageing by salt mist.
39.
B117–2011(E). Standard Practice for Operating Salt Spray (Fog) Apparatus1.
40.
ISO 15148. 2002(E) Hygrothermal performance of building materials and products. Determination of water absorption coefficient by partial immersion 2002.
41.
ISO 21466. 2019(E) Microbeam analysis — Scanning electron microscopy — Method for evaluating critical dimensions by CD-SEM 2019.
42.
ISO 15901-1. 2005(E) Evaluation of Pore Size Distribution and Porosimetry of Solid Materials by Mercury Porosimetry and Gas Adsorption 2005.
43.
Xie Y, Xie J-c, Bai L. Experimental Study on the Effect of Salt on Vapor Migration in Cement Mortar, Building Energy Efficiency (2021).
44.
Méndez-Bermúdez JG, Dominguez H, Pusztai L, et al. Composition and temperature dependence of the dielectric constant of 1-propanol/water mixtures: Experiment and molecular dynamics simulations. J Mol Liq. 2016;219:354–8.
45.
Franke L, Grabau J. The influence of salt content on the drying behavior of bricks. Conserv Hist Brick Struct (1998) 59–68.
46.
Hendrickx R, Clercq HD, Vanhellemont Y. Experimental Determination of Liquid Transport Properties on Salt-Contaminated Porous Stone.
47.
Charola AE. Salts in the Deterioration of Porous Materials: An Overview. J Am Inst Conserv. 2000;39(3):327.
48.
Rovnanikova P. Environmental pollution effects on other building materials. Trends Genet Tig. 2007;10(7):225–6.
49.
Lubelli B, van Hees RPJ, Brocken HJP. Experimental research on hygroscopic behaviour of porous specimens contaminated with salts. Constr Build Mater. 2004;18(5):339–48.
50.
Rodriguez-Navarro C, Doehne E. Influence of Evaporation Rate, Supersaturation and Crystallization Pattern. Earth Surf Proc Land. 1999;24:191–209.
51.
Lubelli B. Sodium chloride damage to porous building materials, Architecture, 2006.
52.
Sibiryakov B, Leite LWB, Sibiriakov E. Porosity, specific surface area and permeability in porous media. J Appl Geophys. 2021;186:104261.
53.
Ghasemi Y, Emborg M, Cwirzen A. Exploring the relation between the flow of mortar and specific surface area of its constituents. Constr Build Mater. 2019;211:492–501.
54.
Benavente D et al. M.A.G. del Cura, x, Role of pore structure in salt crystallisation in unsaturated porous stone, Journal of Crystal Growth 260(3) (2004) 532–544.
