An Experimental Study on the Consolidation of Earthen Surfaces Using Nanoparticle-Based Products

Abstract

This paper presents the results of research conducted as part of a bilateral cooperation project between National Research Council (Italy) and Chinese Academy of Cultural Heritage (China) for the conservation of the earthen walls of Ancient Ulanbay City (Xinjiang, China). In 2007 and 2012, conservation interventions were carried out on the remains of the ancient walls, focusing on areas at risk of collapse. This involved the construction of new adobe masonry (sun-dried earthen bricks and mud mortar) to support the ancient rammed-earth walls, which required consolidation treatments due to their exposure to weathering. In order to support the site’s conservation efforts, several nanoproducts were selected for testing as consolidants for the adobe bricks. Nano-silica (NanoEstel) and nano-lime (Calosil E25), with and without ethyl silicate, and a nano-calcium oxalate-functionalized ethyl silicate (SurfaPore FX WB) were tested and compared with commonly used products for surface consolidation. Ethyl silicate was applied alone as a reference treatment. The mixtures tested in this research had not been previously explored, thus offering new opportunities to identify suitable solutions for the consolidation of earthen structures exposed to environmental conditions. In this study, adobe bricks were sampled from the archaeological site, and the effectiveness of each treatment was assessed based on changes in chromatic appearance, cohesion, and water behaviour. The results showed different behaviours of nanoproducts. Nano-silica, alone or especially in combination with ethyl silicate, is overall more effective than nano-lime for the consolidation of earthen materials, thanks to its greater compatibility with these materials.

1. Introduction

The synthesis of new products for the preservation of Cultural Heritage, to complement or replace those that have been in use for many years, is an evolving field of research and one of the most requested and funded topics at the European level. Nanotechnologies represent a new frontier of environmental conservation and have also been applied to the conservation of Cultural Heritage. Earth was the first material used by man for building, and earthen heritage architecture is spread all over the world. This cultural asset is at risk of extinction due to causes related to a lack of maintenance or incorrect and incompatible conservation interventions, including replacement with modern materials (i.e., cement), which can lead to further degradation. In recent decades, this material has received more attention as it is considered economic and sustainable. Often, it is preferred to substitute the degraded materials instead of consolidating them. Conserving such existing structures remains challenging. Widespread in many countries all over the world (in 2025 more than 200 of 1248 cultural properties in the World Heritage List had earthen structures [1]), earthen constructions not represent only a thousand-year tradition, but, above all, they constitute a testimony to the knowledge, skills and habits of civilisations that used raw earth for their buildings, particularly rural ones. In the early 1980s, one third of the world’s population was estimated to live in earthen buildings. For decades, earth has been considered the most used material for the construction of many historic villages [2,3,4,5,6]. The concept of modernity to which people aspire has led to the progressive abandoning of earthen construction techniques, producing rapid deterioration and leading to the extinction of much earthen heritage. These dwellings are mostly abandoned and are at risk of disappearing. One of the main problems of earthen architecture is its limited durability under the aggressive action exerted by natural atmospheric agents, such as wind and rain (for example, capillary rises in water), often aggravated by a lack of maintenance. Indeed, earthen constructions, although located in different climatic contexts, are affected by similar decay phenomena. Although specific measures have been adopted to ensure effective protection, constant maintenance is the most demanded. It is a widespread practice to apply a layer of render to guarantee protection, and several studies have been conducted to increase the durability of earth render [7,8,9,10,11]. It is also possible to consider other intervention treatments that can be applied directly on site; so far, these are mostly adopted in the archaeological field, where the most pressing conservation requirements include the need to ensure the original surface’s readability. In the field of earthen architecture, in situ conservation interventions have been little studied, and the research has mainly considered applications on decorated/painted surfaces [12]. The most frequently used product is probably ethyl silicate or, more recently, nano-silica [11,13,14,15,16,17], due to its compositional affinity with earth. However, some organic polymeric products (both synthetic and natural) have also been used [5,12]. Furthermore, some experimental treatments using nanoproducts (mainly calcium hydroxide nanoparticles) have been carried out, either alone [18,19,20,21] or in combination with TEOS [13]. Other studies have focused on developing treatment techniques or products that achieve consolidation through the formation of CSH phases with a cementing action [14,15,16,17,22,23,24]. Finally, bio-mediated calcium carbonate precipitation techniques (EICP and MICP) have been used to consolidate earthen structures [16,17,25].

In China, the approach to conserving earthen structures was very ‘conservative’ [26], with potassium silicate being the most frequently used product for surface treatment [27,28,29,30,31,32,33,34]. Conservators have only recently started to investigate new consolidant materials and techniques [35,36,37,38,39,40,41,42,43].

The present work was conducted as part of the bilateral project between National Research Council (CNR-Italy) and Chinese Academy of Cultural Heritage (CACH -China), focusing on the surface conservation of the case study of Ulanbay Ancient City through laboratory tests.

Ulanbay Ancient City dates to the Tang and Yuan dynasties (618–907) and is located about 17 kilometres from the centre of Urumqi, capital of the Xinjiang Uygur Autonomous Region. It is currently the oldest and largest ancient city site in Urumqi. Studies conducted by the Protection Association of Ulanbay Ancient City have revealed the significant role played by the ancient city of Tangluntai in the cultural history of the Western regions and the development of the Silk Road. It was approved as a National Key Cultural Relic Protection Unit on 25 June 2001. The city is at an altitude of 1100 m and is surrounded by a 2 km long wall. Most of these ancient rammed-earth walls range in height from 4 m to 7 m, and have square-based turrets along them. Almost all the structures within the walls were destroyed. Only the ruins of the walls remain, and conservation interventions were performed on them in 2007 and 2012 (Figure 1a). These interventions addressed areas at risk of collapse by reinforcing the rammed-earth walls with adobe masonry to support the deteriorating sections (Figure 1b). Above the adobe there is a thin layer of earth render, most likely due to the effects of rainfall, forming a slurry film. This should be prevented by adequate surface consolidation treatment. Otherwise, once dry, this layer may detach from the substrate due to wind erosion or the expansion and contraction induced by cyclic temperature variations or salt crystallisation (Figure 1c) [44]. The main factors affecting the walls and inducing degradation, for which previous interventions were required, are essentially climate-related. It rains infrequently but intensely in the area, mostly for two months in the summer when temperatures are very high, leading to rapid evaporation and the formation of cracks, exfoliation, and scaling off, ultimately resulting in the collapse of the foundations. During the winter, the climate is dry but windy, which favours strong surface erosion. The climate is mainly characterised by harsh and humid conditions. The average annual temperature is 7.5 °C. Rainfall levels in Ürümqi are noteworthy, as it rains even during the driest month and around 406 mm of rainfall is recorded each year. This climate is classified as Dfa (humid continental) according to the Köppen–Geiger climate classification [45].

The main aim of this research was to find a proper and compatible treatment for the conservation of the surfaces of this important site, conducting a laboratory selection between several nanoproducts commonly used for surface consolidation: nano-silica (NanoEstel), applied either alone or mixed with ethyl silicate, a nano-calcium oxalate-functionalized ethyl silicate (SurfaPore FX WB) and nano-lime (Calosil E25) mixed with ethyl silicate. Ethyl silicate was applied alone as a reference treatment. The mineralogical and granulometric characteristics of the earthen bricks were determined and the performance of each treatment was evaluated (cohesion, water behaviour, and chromatic variation), individualising the best treatment solution in terms of compatibility, the absence of drawbacks and consolidating efficacy.

2. Overview of Surface Consolidation Treatments on Earthen Substrates

The following paragraphs provide an in-depth analysis of the current bibliography on the consolidation of raw earth substrates, including historical structures and decorated or painted surfaces. While the bibliography is not exhaustive, it covers all the sources that the authors are aware of. Interest in this subject is evident from the increasing number of experiments conducted in recent years, including techniques that have not previously been applied to this type of substrate in the cultural heritage field. This growing interest is also linked to the significant heritage in the Middle East, which is currently undergoing large-scale restoration projects involving substantial financial investment.

2.1. In the World

One of the first works reporting an experiment for the surface consolidation of earthen structures was conducted by Giacomo Chiari in 1990 [46]. The study describes the results of an experimental campaign carried out between 1968 and 1971 in Iraq, where various consolidation products were tested, including acrylic resins, polyvinyl acetate, and ethyl silicate, for the consolidation of adobe. Ethyl silicate was found to be the most effective product in laboratory testing and was extensively applied in situ in the Seleucia and Hatra areas. However, twenty years after the treatment, a different behaviour of the treated surfaces was observed. This was not due to differences in the composition of the earthen material or local climatic conditions, but rather to the design of the building. The final conclusion indicated that ethyl silicate is not able to withstand twenty years in an outdoor environment exposed to rain. Chiari [46] specifies that no tested product has all the necessary characteristics for consolidating adobe, and that the choice of product should be based on the specific characteristics and conservation problem of each case. A few years after Chiari’s study [46], Beas conducted her own research in her thesis [12], where she created different types of earthen plasters using traditional recipes and applied various consolidating agents, including acrylic resin, ethyl silicate, and extract from cactus. Beas highlights that the mucilage admixture could be the best product even if it induces only a slight increase in abrasion resistance, as colour and water vapour transmission remain practically unchanged. On the other hand, the acrylic additive improved abrasion resistance and slightly increased water resistance, but significantly changed the colour and reduced water vapour transmission. Ethyl silicate was the most effective in terms of consolidation, with higher abrasion resistance and a good reduction in the rate of capillary water absorption, but it also caused a slight colour change and the greatest decrease in water vapour permeability. Ferron [47] examined the potential for consolidating Mesa Verde decorated earth surfaces, which presented significant disaggregation issues, using various alkoxysilanes and a gelatine-based solution [47,48]. The study evaluated several factors, such as cohesion, resistance to water absorption, and colour, and, similar to Chiari [46], emphasised the need for a compromise when selecting a consolidant in order to achieve the desired effect, for example, improved consolidation. Lohnas [18] was one of the first researchers to apply nano-products to earthen materials. In her thesis, she conducted laboratory tests to assess the effectiveness of a calcium hydroxide nanoparticle dispersion (CaLoSiL) in consolidating adobe blocks with painted clay plaster surfaces. The product was applied via spray, and its efficacy was evaluated through surface absorption, scotch tape, and colour change tests. The scotch tape tests showed an increase in the cohesion of the paint layer after consolidation. XRD analysis revealed the possible presence of plombierite (a calcium silicate hydrate) on the paint layers of the treated samples. This is consistent with previous studies that have reported the formation of calcium silicate hydrates through the pozzolanic reaction between clay minerals and calcium hydroxide [22]. Lohnas [18] also notes the importance of identifying the appropriate application technique to ensure a uniform distribution and control surface whitening. Lanzón et al. [19] conducted a study on the surface modification of stucco, adobe, and stone using a diluted solution of Ca(OH)₂ nanoparticles. The nanoparticles were applied in five consecutive coats and the efficacy of the treatment was evaluated through erosion tests and electron and optical microscopy. The results showed that the adobe surface had significantly improved resistance due to the formation of nanostructured CaCO₃ with binding action. Moreover, there was a moderate change in colour due to the low concentration of the suspensions. The penetration depth in adobe was approximately 3 mm. The low concentration of nanoparticles also made the suspensions stable and easy to apply with nebulizing devices. Lanzon [20] continued to experiment with Ca(OH)₂ nanoparticles on earthen materials, this time on the walls of the Alhambra. After analysing fragments of the walls, it was discovered that the external surface had likely been protected with air lime-based mortars. As a result, the researchers proposed testing the nanoparticles of calcium hydroxide on some of the analysed fragments and reconstructed ones. The Ca(OH)₂ nanoparticles were nebulized using a low-pressure laboratory compressor. The study showed that the nanoparticles increased hardness and resistance to erosion, making them suitable for reinforcing the earthen walls of the Alhambra. Garcia-Vera et al. [20] investigated the effectiveness of Ca(OH)₂ coatings in protecting earthen plaster surfaces. Earthen plaster samples were produced in the laboratory and the nano-calcium hydroxide product was synthesised and applied five times. The effect of nano-lime on surface consolidation, water absorption, and durability in the presence of acid rain was then evaluated. The treatment improves surface cohesion and reduces water absorption. However, its effectiveness in protecting the earthen surface from interaction with acid rain has not been demonstrated. García-Vera et al. [49] extend the study of the consolidation of earthen plasters by evaluating the effectiveness of ethyl silicate. The samples are produced in the laboratory using two types of clay, and a commercial ethyl silicate product containing 75% active ingredient is used. The ethyl silicate is applied to the surface of the samples using a nebuliser in a single application. This improves the surface cohesion of the earthen plasters, reduces water absorption and provides good protection against acid rain. The authors conclude that ethyl silicate can be used to consolidate plasters, suggesting that improvements in homogeneous distribution could perhaps be achieved using an alternative application method. Stazi et al. [11] investigated the effectiveness of several admixtures of earthen plasters and surface treatments based on silicon nanoparticles, titania nanoparticles, silica nanoparticles, silane–siloxane and beeswax. According to the results of the water-repellence and water erosion resistance tests, earthen plaster treated with the silane–siloxane product was found to be the most effective. The authors affirm that the choice of treatment depends on the desired properties. Salazar-Hernández et al. [13] further extend the study of the application of ethyl silicate, combining it with other products. The authors aimed to evaluate the impact of adding silica and PDMS nanoparticles on the effectiveness of ethyl silicate when applied to adobe. They used adobe from XVII and XVIII-century mining haciendas in Guanajuato, Mexico. Four consolidants were synthesised: TEOS; TEOS mixed with silica nanoparticles; PDMS-OH; and TEOS mixed with PDMS and silica nanoparticles. The consolidants were applied with a brush. The results showed that colloidal silica improved penetration into the porosity, as well as the hardness and cohesion of the earthy material. However, the best formulation appears to be one that incorporates both additives. In fact, PDMS not only improves penetration into the porosity of the earthy material, and therefore its hardness and cohesion, but also improves the flexibility of the TEOS gel, thereby limiting cracking. Adding both components to the TEOS gel provides an excellent consolidating treatment for adobe materials. Elert et al. [22] propose attacking the rammed earth walls of the medieval city of Alhambra with an alkaline solution as an alternative method of consolidation. This method involves dissolving clay minerals using highly alkaline solutions (NaOH or KOH), resulting in the formation of mineral phases such as zeolites. These minerals are not subject to swelling and also play an earth-cementing role. The positive results obtained on laboratory samples led the authors to conduct another experiment on adobe walls that were exposed outdoors for 18 months and to compare the alkaline treatment with ethyl silicate, which is commonly used [14,15]. The treatments were applied with a brush. After 18 months of exposure, the results show that the wall treated with the alkaline attack exhibits better mechanical resistance and water resistance. Furthermore, the alkaline treatment shows a four times greater penetration than ethyl silicate (20 mm compared to 5 mm). Both treatments cause a similar change in surface colour. Mineralogical and compositional analyses revealed the formation of an amorphous silica gel in the case of the wall treated with ethyl silicate (Estel 1000), and an amorphous phase, likely zeolite-type, with a cementing function for the clay particles in the case of the alkaline treatment. Elert et al. [16,17] extend the study by experimenting with additional products. They tested and compared ethyl silicate, calcium hydroxide and silica nanoparticles, and consolidation methods (alkaline activation and bacterial biomineralisation), both on samples made in the laboratory from soil from the Alhambra formation and directly in situ. The products are applied with a brush, except for bacterial biomineralisation, which is applied by spraying. The ethyl silicate consolidant produces the best results in terms of improving mechanical strength and weather resistance in laboratory tests. Nano-calcium hydroxide and nano-silica could be limited to substrates with powdery surfaces. Conversely, alkaline activation and biomineralisation treatments require protocol optimisation to improve their effectiveness when applied in situ. Parracha et al. [25] developed new bioproducts based on iron-enriched E. coli cultures for application to earthen plasters, with the aim of increasing their water resistance and superficial cohesion. These biotreatments were spray-applied to an earthen plaster surface in a laboratory setting. While all biotreatments increased resistance to water absorption, creating a slight waterproofing effect, none increased surface cohesion. The authors emphasise the need for further study to understand the effect on the microstructure of earth mortars. Camerini et al. [23] studied the possibility of consolidating a powdery earth substrate through the formation of calcium silicate hydrate (CSH) via an in situ pozzolanic reaction between silica and calcium hydroxide nanoparticles. Different formulations were tested to identify the one that induced the best performance in terms of resistance to peeling, abrasion and wet/dry cycles, as well as lower colour variation. XRD measurements evidenced the formation of CSH within the pores of the treated adobe samples. Tonelli et al. [24] also investigated the possibility of reinforcing adobe through the precipitation of a CSH binder gel phase from diluted aqueous suspensions of GGBS activated with calcium hydroxide. Samples for testing were prepared using handmade adobe bricks from a contemporary construction site in the state of Morelos (Mexico). The best formulation was selected from among the various formulations and applied to the samples. These investigations demonstrate the formation of CSH compounds within the porosity and cracks of the adobe and show that these compounds improve the cohesion of the earthen material.

2.2. In China

In China, Li’s group [27,28,29,30] studied the surface consolidation of several historical structures located along the Silk Road using potassium silicate. The potassium silicate was applied using various methods (infiltration or spraying, sometimes multiple times) and at different concentrations depending on the conservation state of the surfaces, which were mainly affected by degradation caused by wind and rain. Experiments indicated that applying low concentrations of potassium silicate multiple times was the most effective approach for conserving rammed earth structures because it maximises deep infiltration and minimises the concentration gradient. This avoids the formation of a hard crust that could detach from the substrate over time. As potassium silicate has been shown to increase resistance to wind and rain erosion, new studies and applications on earthen structures along the Silk Road are ongoing [31]. A high-molar-ratio potassium silicate solution was found to improve resistance to wind and rain erosion while maintaining acceptable water vapour permeability [32]. Zhang et al. [33] continued to study the effectiveness of potassium silicate in protecting historic sites in China from wind erosion. Zhang et al. [33] tested the effects of the treatment on samples using the infiltration method at up to 85% saturation, with different concentrations and numbers of applications. Evaluation was conducted in a wind tunnel by comparing treated samples with original, untreated material fragments from historic sites. The optimal treatment concentration was identified by observing that three applications significantly improved wind erosion resistance. Recently, Zhang et al. [34] tested the high-molar-ratio potassium silicate solution previously used by Li et al. [32] on laboratory-made soil samples from the Andier site in the Taklimakan Desert hinterland, Xinjiang Autonomous Region, China. This confirmed an increase in resistance to both wind and rain erosion.

Alongside the experimentation with potassium silicate for conserving historical earthen structures on the Silk Road, various treatments, including acrylic resins, have been tested over time [35,36,37]. Zhou et al. [35] modified an aqueous dispersion of acrylic and silicone polymers to create a non-aqueous dispersion in organic solvents for conserving a number of Chinese archaeological sites. The new consolidant formulation identified as the most effective through laboratory testing was applied by spraying on-site. In all in situ applications, the product demonstrated good penetration and slight colour change, and the surface remained sufficiently hard, even three years after treatment, without losing earthen material when scratched. Wan et al. synthesised an organic–inorganic epoxy–silica–acrylate hybrid material for consolidating the Jinsha archaeological site in Chengdu, China. The best laboratory-tested formulation of the hybrid product was applied by spraying to a wall of the Jinsha archaeological site in Chengdu, where ethyl silicate and potassium silicate were also tested for comparison. The new hybrid product proved superior, as it was rapidly absorbed and showed good penetration depth and scratch resistance, even at low concentrations. Zhao et al. [37] studied the consolidation effect of a silicone-modified acrylic resin for conserving archaeological sites made of adobe and rammed earth, such as the Gaochang ruins in Xinjiang. When applied to laboratory specimens by spraying, the product showed improved resistance to wind erosion and water. It also withstood freeze–thaw cycles and did not form a surface crust. Furthermore, it demonstrated good adhesion strength with the internal parts, despite significantly reducing permeability to water vapour. Furthermore, it maintained the capacity for moisture exchange between the inside and outside. Therefore, the modified acrylic resin is suitable for consolidating archaeological sites in arid regions of north-western China. Recently, Du et al. synthesised a new consolidation material for earthen sites: a silicone-modified acrylic emulsion. The product was tested on a sample in a laboratory. The results showed that the new formula has increased resistance to water ingress, erosion and salt attack, with acceptable colour changes and water vapour permeability. Additionally, the use of nanotechnologies for the conservation of historic earthen surfaces and sites in China has recently been investigated. Peize Han et al. [39] evaluated the effectiveness and compatibility of a nano-lime treatment for consolidating the Dunhuang frescoes painted on a mural earthen wall in the Mogao Caves, a World Heritage Site in China. The treatment was applied to a mural earthen wall model in the laboratory and increased the cohesion and surface hardness, albeit to varying degrees depending on the colour. The colours remained within an acceptable range. Chen et al. [40] studied micro-lime and ethyl silicate, both alone and in succession in different orders, for a total of five treatments. Samples made with soil from the Great Wall of Yongchang in China’s Gansu Province were used for the laboratory tests. The products were applied to all sides of the samples using a pipette. Better reinforcement effects were obtained by first applying ethyl silicate, followed by micro-lime. This sequence enables the ethyl silicate to penetrate deep into the porous structure of the earthen material and perform a consolidating action, while the micro-lime remains more on the surface and plays a protective role against external environmental agents. Zeng and Schwantes [41] recently studied the relationship between the consolidating effect of Ca(OH)₂ micro-lime and the humidity present in earthen structures and their environment. Their research focused on understanding the limitations of this product when applied to historic earth structures located in arid climates, such as those found in China’s cultural heritage sites. Three different concentrations of micro-lime dispersion were applied to earthen samples made using natural clay soil from the vicinity of the Yungang Grottoes, which are situated in the Datong region of Shanxi Province, China. The product was dripped onto all six surfaces of the samples. When testing different UR% conditions, an increase in both consolidating power and carbonation rate was observed when the relative humidity increased. Other recent Chinese investigations into the conservation of earthen structures concern the use of a TEOS and barium hydroxide mix [42], as well as enzyme-mediated (EICP) and microbial-mediated (MICP) calcium carbonate precipitation [43]. Chen et al. [42] explore the use of a TEOS and barium hydroxide mixture for consolidating rammed earth walls, comparing it with a TEOS and ethanol mixture. These two treatments were tested in a laboratory on earthen samples and on the rammed earth wall of the south gate in Qinzhen, Xi’an City, in situ. The treatments were applied via surface drip infiltration. The results showed that the TEOS and barium hydroxide mixture was more effective than TEOS alone, providing a greater increase in mechanical strength as well as resistance to salt attack and freeze–thaw cycles. Furthermore, the TEOS and barium hydroxide mixture does not significantly alter the colour and leaves water vapour permeability virtually unchanged. Li et al. [43] investigated the effect of soil density on EICP and MICP techniques for soil stabilisation by examining historical earthen sites in the ancient city of Suoyang in Gansu Province, China. The experiments used cylindrical samples of three different densities, created using soil from near the Suoyang site. The cementing solutions were applied with a pipette by dripping them from above. The results indicated greater calcium carbonate production for the higher soil densities, which had a positive effect on the soil’s mechanical strength. Samples treated with MICP showed greater increases in calcium carbonate production than those treated with EICP, suggesting that MICP is a superior technique.

3. Materials and Methods

3.1. Preparation of Earthen Material Specimens

Three adobe bricks, with an average size of 23 cm × 17 cm × 10 cm, were taken from the adobe walls of the ancient city of Ulanbay for testing. A total of 42 specimens measuring 5 cm × 5 cm × 2 cm were obtained from the adobe bricks sampled. The adobe bricks were cut by removing the outermost layer from all sides to a thickness of approximately 1 cm. Then, slices 2 cm thick were cut parallel to the side, measuring 23 cm × 10 cm. Parallelepiped-shaped samples with a height of 2 cm and an area of 5 cm × 5 cm were then obtained from these slices. The residues from cutting the samples were used for characterising the earthen material. A total of 100 g was used to determine the particle size distribution; 1 g each was used for the calcimetric analysis and the analysis of the main mineralogical composition; and 5 g was used for the analysis of the mineralogical composition of the clay fraction.

A set of six specimens was used for each treatment and relative performance evaluation. Six specimens were left as untreated references.

3.2. Treatments

Considering the state of the art in the field of earth materials consolidation and in agreement with Chinese colleagues, it was decided to test several products and mixtures on the adobe bricks from Ulanbay, with a total of six different treatments, as indicated in

Table 1. Products and their application by brush.

Treatment	ID	Product Concentration (w/w)	Applied Mixture Concentration (w/w)	Amount of Treatment Applied (mg/cm2)
Ethyl Silicate	ES	100%	50% in ethanol	39 ± 8
Modified Ethyl Silicate	MoES	as supplied	as supplied	22 ± 5
Ethyl Silicate + Nano-Silica (1:1 v/v)	NSES	100% + 30% in water	3% in ethanol + 3.75% in water	5 ± 1
Ethyl Silicate + Nano-Lime (1:1 v/v)	NLES	100% + 3% in ethanol	5% in ethanol + 3% in ethanol	4 ± 1
Nano-Silica	NS	30% in water	30% in water	21 ± 5
Nano-Silica 3.75%	diNS	30% in water	3.75% in water	3 ± 1

.

Ethyl silicate is one of the most common and longest-used products for the conservation of earthen materials due to its compositional affinity with the earth [12,13,16,17,21,46,47,48]. In this work, it was tested on its own (hereinafter ES) and in a mixture with nanoparticles, both nano-silica (hereinafter NSES) and nano-lime (hereinafter NLES). In general, the addition of nanoparticles to ethyl silicate would lead to an increase in performance in terms of effectiveness (consolidating power) and reduction in negative effects (cracking due to the shrinkage of silica gel during curing) [13,23]. Furthermore, the mix composed of nano-lime + ethyl silicate may form hydraulic compounds through the reaction between free calcium ions and silicic acid (Calcium Silicate Hydrate gel—CSH) [23,50], and nano-lime and clay minerals through the pozzolanic reaction, with considerable consolidating action [51,52,53,54,55,56,57,58,59,60].

Moreover, the drawback of ES based materials, which exhibit cracks after condensation as a result of the sol–gel process, can be overcome by the template formed with nano-calcium oxalate, which modified the pore system and the solvent evaporation velocity [40]. Furthermore, the incorporation of nano-calcium oxalate into the silica matrix resulted in advanced chemical affinity between the consolidant and carbonate species [50]. This property facilitates the adhesion of the consolidant in mixed matrixes containing both silicate and carbonate compounds, which served as the adobe substrates in our experiments. Thus, an innovative nanocomposite based on TEOS modified with nano-calcium oxalate was also tested (hereinafter MoES).

Finally, nano-silica was tested on its own, in two different concentrations (hereinafter NS and diNS), to verify if nano-sized particles have a better consolidating effect even if applied in an aqueous environment. The ethyl silicate was applied at 50% w/w, the best concentration according to Ferron and Matero [48].

The mixtures were prepared following the procedures indicated in the work of Al-Dosari et al. [61] and Borsoi et al. [57]. Al-Dosari et al. [61] prepared the mixture through the direct mixing of ethyl silicate and silica nanoparticles. To the solution of ethyl silicate (3% w/w in ethanol), placed to mix on a magnetic stirrer, the aqueous solution containing the nano-silica particles was added at a concentration of 3.75% w/w (which was shown to be better in Borsoi et al.’s [57] experiment). The mixture was then stirred vigorously for 45 min to homogenise it.

Borsoi et al. [57] primarily homogenised nanoparticles of nano-lime (3% w/w in ethanol) by ultrasonic treatment. They added ethyl silicate (5% w/w in ethanol) to nano-lime and mixed through magnetic stirring for 5 min to obtain a homogeneous consolidation product [57]. In the experiments of Al-Dosari et al. [61] and Borsoi et al. [57], there are no indications regarding the relationship between nano-particles and ethyl silicate; therefore, it was decided to create the mixtures with a 1:1 v/v ratio.

All the formulates were applied to the samples by brushing. The applications were performed in a conditioned room, at 20 ± 2 °C and 50% RH, wet-on-wet until rejection (when the surface remained wet for more than one minute). After treatment, all the test specimens were allowed to dry in a climatic chamber at 60 ± 5% RH and 20 ± 2 °C, until they reached a constant weight. This was achieved after 45 days; the amount of deposited material in the stone specimens, in terms of mg/cm² of dry matter, was then calculated.

The following commercial products used: ethyl silicate (Wacker SILRES BS OH 100, Wacker Chemie, Munich, Germany) [62], ethyl silicate functionalized with nano-calcium oxalate (Surfapore FX WB, Nanophos, Athens, Greece) [63], nano-silica (Nano Estel, by CTS s.r.l., Florence, Italy) [64] and nano-lime (Calosil EP 25, IBZ-Salzchemie GmbH & Co. KG, Halsbrücke, Germany) [65]. All the Technical Leaflets are available in reference [62,63,64,65].

3.3. Earthen Materials Characterizations

3.3.1. Mineralogical Composition

The mineralogical composition of the earthen bricks was determined through X-ray diffraction (XRD) (X’Pert PRO diffractometer equipped with X’Celerator detector and HighScore software, version 3.0d (PANalytical B.V., Lelyweg, The Netherlands) for the acquisition and interpretation of data according to the following operative conditions: CuK α1 = 1.545 Å radiation; 40 KV; 30 mA, 2θ = 3–70°. Analysis of the clay minerals was performed following the Cipriani methodology [66,67] and utilising a Philips PW 1729 diffractometer (Eindhoven, The Netherlands) under the following operative conditions: CuK α1 = 1545 Å radiation; 40 KV; 20 mA; 2θ = 3–20°). According to this methodology, the samples are first gently ground and put in a sedimentation-graduated cylinder. After an interval calculated according to Stokes law (1 h), a fraction <4 p was collected. This fraction, properly concentrated in suspension, was then placed on a small glass plate, where it dried and formed an oriented powder sample that will be used in the XRD scanning. To recognise the different clay minerals, the oriented powder sample undergoes the following treatments before each XRD scanning: not treated; treated with ethylene glycol; and thermal treatment. These treatments cause the shifting or disappearance of the basal reflections of some clay minerals and allow for their identification. A semi-quantitative analysis was performed by comparing the elongation of the clay mineral peaks in the acquired diffractograms [68].

3.3.2. Calcimetry

The amount of calcium carbonate was determined through the Gasometric technique using a Dietrich Früling calcimeter [69].

3.3.3. Grain Size Distribution

The grain size distribution of the earthen bricks was determined in order to separate the following fractions: sand (Ø > 63 µm), silt (4 µm < Ø < 63 µm) and clay (Ø < 4 µm). These were according to the standard practice ASTM D2217-85 [70] and ASTM D422-63 [71].

3.4. Performance Evaluation

3.4.1. Consolidation Effectiveness

The possible consolidating action induced by the various treatments was assessed by measuring the drilling resistance with the Drilling Resistance Measurement System-DRMS (SINT Technology, Calenzano, Italy) [72,73,74]. The measurements were taken under the following operating conditions: 200 rpm speed of rotation, 10 mm/min forward speed of the bit, and 10 mm depth of the hole. A Diamond drill bit with a 5 mm diameter was used. Three measurements were taken for each sample.

Microdurometer (Rockwell surface hardness) was used in support of DRMS to better highlight the hardening effect of the treatments in the more superficial layers. The surface micro-hardness tests were performed with a Galileo model DG 201 hardness tester (office Galielo S.p.a., Florence, Italy). The instrument measures the penetration depth of a penetrator loaded with a known load. The penetrator used for this test is a steel ball with a diameter of 5 mm and a load of 5 kg. The result is expressed in microns of the penetration depth of the tip in the material. For each sample, 10 measurements were taken (30 measurements in total for treated references, and the same for untreated references).

3.4.2. Absence of Drawbacks of Applied Treatments

Three other analyses were performed to evaluate the compatibility of the applied treatments in terms of undesirable changes in aesthetic appearance and behaviour towards water: A. colour measurements; B. water absorption by capillarity; and C. water vapour permeability. These tests are essential to ensure one of the main conservation principles is respected, related to the fact that the chosen treatment should not alter the intrinsic characteristics of the original material

A. The colour variations induced on the earthen plasters with and without additives comprise a purely aesthetic parameter that may be interesting to verify in case of the restoration and integration of missing parts. The Konica Minolta spectrophotometer CM 700 d (Konica Minolta Sensing Europe B.V., Cinisello Balsamo, Italy), adopting the CIE L* a* b* method [75], was used for the calculation of the colour parameters. According to this method, the colour of a surface is described by three parameters: L* (0 to 100) represents the lightness, a* is related to the impulse of red–green colour and b* is related to the impulse of yellow–blue colour [76].

The total change in colour is summarised by the parameter ΔE*, calculated with the following equation:

ΔE* = [(ΔL*)² + (Δa*)² + (Δb*)²]^1/2

where ΔL* = (L* not treated − L* treated); Δa* = (a* not treated − a* treated); Δb* = (b* not treated − b* treated).

Colour variations that determine a ΔE* greater than 5 are perceptible to the naked eye. The measurements were performed with diffuse illumination (D65 standard source) on an area of 8 mm in diameter, with specular components both included (SCI) and excluded (SCE). Nine measurements were taken for each sample.

B. The water absorption by capillarity was evaluated by adapting a method from LNEC [77] that consists of placing the sample in a layer of wet sand. In this test, a paper filter was also used between the sand and the specimen to keep the surface clean and flat (Figure 2).

The curve for weight gain over time was determined using the following steps: 0, 5, 10, 15, 20, 25, 30, 45, 60, 90, and 120 min.

The 120 min time (end of the test) corresponds to the rise in the water up to the top face of the untreated sample, avoiding disaggregation. The following parameters were then determined:

-

Amount of water absorbed at 120 min per unit area (Qf);

-

Capillary absorption coefficient (CA), given by the initial slope of the regression line of the water absorption curves;

-

Protective efficacy (EP), calculated as follows:

EP% = [(Qfnot treated − Qftreated)/Qfnot treated] ∗ 100

Qfnot treated = amount of water absorbed at the end of the test by the untreated sample.

Qftreated = amount of water absorbed at the end of the test by the treated sample.

C. The water vapour permeability test was conducted according to the procedure and instrumentation described in the EN 15803 [78], measuring the mass of vapour flowing every 24 h for up to 10 days, when equilibrium was reached (i.e., when the difference between two successive weightings at an interval of 24 h is not greater than 0.1% of the mass of the specimen). For assembling the wet cup test, 10 mL of distilled water was placed in the cup, leaving a small gap of air between it and the material. The test chamber was maintained at a constant temperature of 23 °C and a relative humidity of 50%. The barometric pressure was recorded every day.

The tests were performed on three specimens for each type of treated and untreated brick sample until a steady state was reached (10 days after the beginning of the test). The test results are expressed as µ (water vapour diffusion resistance) [78]. The water vapour diffusion resistance value indicates to what extent the transport resistance of the water vapour is higher in a material than in air.

4. Results

4.1. Earthen Materials Characterisation

The mineralogical composition of the different adobe bricks is similar, with the coarsest fraction characterised by the presence of quartz and feldspars, as well as a quantity of carbonates (calcite) equal to 11–15%, as determined by calcimetry (

Table 2. Principal mineralogical composition of adobe bricks by XRD.

Sample	Quartz	Feldspars	Calcite *	Phyllosilicate + Accessory Minerals
Adobe 1	x	X	15	x
Adobe 2	x	X	11	x
Adobe 3	x	X	13	x

* CaCO₃ through calcimetry.

). Figure 3 shows an example of an XRD diagram relating to the main mineralogical composition of the adobe brick.

The finer fraction is mainly constituted of clay minerals, with kaolinite (25–30%), illite (35–50%), chlorite (15–20%), and smectite (10–20%) (

Table 3. Clay mineral composition of adobe bricks by XRD.

Sample	Kaolinite *	Illite *	Chlorite *	Smectite *
Adobe 1	25	35	20	20
Adobe 2	30	40	15	15
Adobe 3	25	50	15	10

* Semiquantitative data.

).

The grain size data are reported in

Table 4. Grain size distribution of adobe bricks.

Sample	Sand%	Silt%	Clay%
	Ø > 63 µm	4 µm < Ø < 63 µm	Ø < 4 µm
Adobe 1	20	49	31
Adobe 2	17	54	36
Adobe 3	23	55	32

and plotted in Figure 4’s ternary diagram, which shows that the earthen material of the bricks falls between the sand–silt–clay and the clayey silt zones.

This analytical data allows this material to be defined as lean earth. With regard to a comparison with the optimal granulometric curve reported by CRATerre for earthen bricks [79,80], it is observed that the granulometry of Ulanbay adobe bricks falls within the zone of acceptability.

4.2. Performance Evaluation Outcomes

4.2.1. Drilling Resistance and Microdurometer

The results of the drilling resistance tests are shown in

Table 5. Drilling resistance results.

Treatment	Depth Range (mm)	Drilling Resistance [N]
		Sample 1	Sample 2	Sample 3	Average
Not treated	0–1	1.39 ± 0.25	1.79 ± 0.22	0.96 ± 0.13	1.38 ± 0.17
	1–10	1.31 ± 0.33	1.24 ± 0.62	1.02 ± 0.24	1.19 ± 0.45
	0–10	1.30 ± 0.35	1.28 ± 0.62	1.00 ± 0.25	1.20 ± 0.46
ES	0–1	4.45 ± 1.43	4.01 ± 1.27	4.57 ± 1.54	4.34 ± 1.32
	1–10	3.56 ± 0.68	3.65 ± 0.80	3.21 ± 0.48	3.47 ± 0.69
	0–10	3.60 ± 0.89	3.64 ± 0.93	3.31 ± 0.84	3.51 ± 0.89
MoES	0–1	5.25 ± 1.89	4.66 ± 1.78	4.93 ± 1.44	4.95 ± 1.65
	1–10	2.60 ± 1.12	2.39 ± 1.01	2.30 ± 0.73	2.43 ± 0.97
	0–10	2.82 ± 1.47	2.58 ± 1.31	2.52 ± 1.16	2.64 ± 1.32
NSES	0–1	2.81 ± 0.66	2.65 ± 0.57	3.28 ± 0.53	2.91 ± 0.56
	1–10	2.59 ± 0.65	2.06 ± 0.67	2.32 ± 0.84	2.32 ± 0.75
	0–10	2.58 ± 0.70	2.09 ± 0.71	2.39 ± 0.89	2.35 ± 0.80
NLES	0–1	1.24 ± 0.25	1.12 ± 0.34	1.09 ± 0.26	1.15 ± 0.20
	1–10	1.53 ± 0.23	1.41 ± 0.28	1.27 ± 0.27	1.41 ± 0.28
	0–10	1.49 ± 0.29	1.37 ± 0.33	1.25 ± 0.30	1.37 ± 0.32
NS	0–1	1.76 ± 1.35	2.96 ± 1.37	2.04 ± 1.50	2.25 ± 1.17
	1–10	1.11 ± 0.17	0.58 ± 0.46	0.89 ± 0.36	0.86 ± 0.41
	0–10	1.16 ± 0.49	0.80 ± 0.93	1.00 ± 0.67	0.99 ± 0.73
diNS	0–1	1.39 ± 0.16	0.86 ± 0.37	1.65 ± 0.46	1.30 ± 0.25
	1–10	1.92 ± 0.65	1.18 ± 0.38	1.82 ± 0.65	1.64 ± 0.66
	0–10	1.86 ± 0.66	1.14 ± 0.40	1.78 ± 0.66	1.59 ± 0.67

, which presents the average values calculated for a single sample (three holes) and for each treatment (three samples) over specific depth ranges (0–1 mm, 1–10 mm and 0–10 mm). Figure 5 shows the average perforation profiles for each treatment (the average of three samples), plotted as drilling resistance versus total hole depth (10 mm).

For the calculation of the drilling resistance averages, three depth ranges were considered: 0–1 mm, 1–10 mm and 0–10 mm. The 0–1 mm and 1–10 mm drilling resistance intervals were calculated to quantify and distinguish the contribution, over the entire test period, shown by the presence, for some treatments, of a surface crust. Observations related to the results obtained for each treatment are reported below:

-

MoES: The presence of a hard surface crust, about 1 mm thick, with drilling resistance values of about 5.0 N (±1.5 N). Subsequently, the drilling resistance values remain around 3.0 N (±0.5 N) up to a depth of 6 mm and then fall to values comparable with those of the untreated 1.4 N (±0.4 N) (Figure 5 and Table 5).

-

ES: The presence of a hard surface crust about 1 mm thick with drilling resistance values of about 4.3 N (±1.3 N). In the range between 1 mm and 10 mm, the values of resistance to drilling remain constant at around 3.5 N (±0.7 N), never lowering to the values of the untreated reference (Figure 5 and Table 5).

-

NSES: There is no surface crust; the product, up to 4 mm, gives the adobe brick a resistance to drilling equal to about 3.0 N (±0.6 N). Then, in the following interval of 4–10 mm, the values of the resistance profile to the drill are lowered to those of the untreated reference (1.8 N ± 0.4 N) (Figure 5 and Table 5).

-

NS: The presence of a hard surface crust, about 1 mm thick, with resistance to drilling values of about 2.0 N (±1.0 N). In the following interval the profile of resistance to drilling stabilises at values comparable with those of the untreated reference (Figure 5 and Table 5).

-

diNS and NLES: There is no increase in the parameter of resistance to drilling, with results comparable with the values of the untreated reference for the entire examined range (Figure 5 and Table 5).

Comparing the results, from a conservative perspective, the NSES treatment achieves good results in terms of increased cohesion and penetration depth (approximately 4 mm from the surface), without forming a superficial crust. On the other hand, the ES and MoES treatments provide penetration depths of up to 10 mm and 6 mm, respectively, from the surface, but induce the formation of a hard surface crust (approximately 1 mm thick for both treatments). This crust is also observed in the NS treatment, with a thickness of around 0.7 mm, coinciding with the treatment’s maximum penetration. The diNS and NLES products do not provide appreciable increases in resistance to drilling.

The micro-surface hardness data are consistent with those of the DRMS, in which a “hard” surface corresponds to low tip penetration values (i.e., a surface with a tip penetration value of 55 μm is softer than one with a value of 23 μm) (

Table 6. Micro-hardness test results.

Treatment	Micro-Hardness (μm)
Not treated	55 ± 5
ES	23 ± 5
MoES	20 ± 5
NSES	34 ± 5
NLES	52 ± 5
NS	32 ± 5
diNS	45 ± 5

).

As can be seen from the DRMS curves, the best results for MoES and ES are actually due to the presence of a surface crust. NS and NSES are the next best ones according to the DRMS test, with the presence of the crust being observed only in the case of NS. In this specific case, the micro-hardness test does not contribute with additional indications regarding the DRMS test.

As microhardness testing is a surface technique, in some cases it may lead to incorrect interpretations of the results. The thickness of the crusts identified in this study ranges from 700 μm to 1 mm, and the maximum depth measured by the micro-hardness test was 55 μm, falling within the thickness of the crust. Consequently, in the case of MoES, ES and NS, the measured hardness is not significant.

4.2.2. Colour Measurements

The values of the SCE (specular component excluded) and SCI (specular component included) colour measurements are very similar, so only the SCI ones are reported.

Table 7. SCI colour variation (Δ*) ^# results and ΔE* parameter.

Treatment	ΔL*	Δa*	Δb*	ΔE*
Not treated	0.23 ± 0.10	−0.03 ± 0.07	−0.03 ± 0.62	0.54 ± 0.27
ES	−1.24 ± 0.87	0.09 ± 0.13	−0.10 ± 0.37	1.30 ± 0.82
MoES	8.64 ± 0.52	−0.64 ± 0.18	3.67 ± 0.84	9.44 ± 0.25
NSES	−3.14 ± 0.53	1.33 ± 0.13	4.27 ± 0.56	5.46 ± 0.32
NLES	−8.46 ± 0.65	2.56 ± 0.16	11.69 ± 0.48	14.65 ± 0.27
NS	3.71 ± 0.67	−0.27 ± 0.14	−1.11 ± 0.79	3.97 ± 0.39
diNS	1.85 ± 0.98	−0.26 ± 0.18	−0.46 ± 0.16	1.94 ± 0.96

shows the difference values (Δ*) of L*, a*, b* parameters and the colour difference ΔE*, before and after treatment, while Figure 5 displays the diagrams of the same parameters.

The results pointed out that MoES and NLES have much higher values of ΔE* than the limit of 5 [81]. NSES has a slightly higher value of ΔE*, while the other products are below this (Table 7 and Figure 6). The most important colour variation was due to the variation in the L* parameter (brightness), with a whitening for NLES and a darkening for Moes. Moreover, for the same products, an increase towards the blue component of the b* parameter (yellow–blue axis) was observed.

4.2.3. Water Absorption by Capillarity

The results of the capillary absorption tests are reported in

Table 8. Water absorption according to capillarity results.

Treatment	Qf (g/cm2)	CA (g/cm2s1/2)	EP (%)
Not treated	0.457 ± 0.023	0.0038 ± 0.0005	-
ES	0.206 ± 0.015	0.0014 ± 0.0002	55 ± 1
MoES	0.020 ± 0.005	0.0002 ± 0.0001	95 ± 1
NSES	0.189 ± 0.013	0.0022 ± 0.0003	58 ± 1
NLES	0.154 ± 0.011	0.0018 ± 0.0003	66 ± 1
NS	0.250 ± 0.018	0.0014 ± 0.0002	45 ± 1
diNS	0.397 ± 0.021	0.0037 ± 0.0005	13 ± 1

Qf = amount of water absorbed at the end of the test; CA = coefficient of absorption; EP = protective efficacy.

as the final quantity of water absorbed (Qf), the capillary absorption coefficient (CA), and the protective efficacy (EP%), and are displayed in Figure 7 as an average curve for each treatment.

The results of the water absorption test by capillarity showed the following:

-

MoES: The amount of water absorbed at the end of the test is very low compared to the untreated reference; in fact, the protective efficacy is 95% (Table 8). The capillary absorption coefficient is very low and remains constant until the end of the test (Figure 7).

-

ES and NS: Both show similar behaviour. The amount of water absorbed at the end of the test is about half that of the untreated reference and the protective efficacy values are 55% and 45% respectively (Table 8). The capillary absorption coefficient is low and can be observed from the graph, which, after about 30 min, begins to increase with a speed (slope of the curve) like that of the untreated reference for NS, and like the untreated reference but slightly lower for ES (Figure 7).

-

diNS: Shows similar behaviour to the untreated reference (Table 8 and Figure 7).

-

NSES and NLES: Both show similar behaviour. The two capillary absorption curves are practically parallel, although the initial absorption coefficient is slightly higher for the NSES (Figure 7). Protective efficacy is good, at around 60% for both (Table 8). It should be noted that both curves have an intermediate plateau between 10 and 60 min, after which the absorption coefficient starts to increase with an absorption speed approximately equal to that at the beginning of the test (Figure 7).

4.2.4. Water Vapour Permeability

The results of the water vapour permeability test, expressed as water vapour resistance factor μ (-), are reported in

Table 9. Water vapour permeability results.

Treatment	Water Vapour Resistance Factor μ (-)
Not treated	8.3 ± 0.4
ES	9.7 ± 0.5
MoES	9.3 ± 0.5
NSES	8.7 ± 0.4
NLES	8.9 ± 0.5
NS	8.4 ± 0.4
diNS	8.1 ± 0.3

.

NS and diNS show water vapour resistance values similar to those of the untreated samples. NSES and NLES show slightly higher water vapour resistance values than untreated samples. MoES and ES show higher water vapour resistance values than untreated samples. The treatments seem to have a modified pore size distribution, increasing the micropore fraction [82].

5. Discussion

The nano-products’ conservative treatments are now consolidated on different types of stone materials, but represent a new frontier for the earthen materials. Among the six selected treatments tested in our work, Wacker OH 100 + Nano Estel (NSES) was shown to be the better one, in agreement with the results obtained by Salazar-Hernández et al. [22], even if they do not provide any indication of the formation of a surface crust, since this aspect was investigated.

Samples treated with MoES, ES and NS developed a hard surface crust, which is detrimental to preservation (Figure 5). The crust reduces water absorption and increases the coefficient of resistance to evaporation (see Figure 7 and Table 8 and Table 9).

In the case of NS, there was no change in the coefficient of resistance to evaporation compared to the untreated samples. The MoES and ES treatments appear to reduce the capillary porosity, but not the water vapour permeability.

With regard to the colour changes resulting from the treatments, the MoES treatment shows a rather significant variation (see Table 7). This is probably because the nanoparticles were deposited on the surface of the material rather than penetrating it, causing it to whiten (a positive change in the L* parameter). This finding is consistent with the formation of a surface crust.

The ES and NS treatments do not induce significant colour changes as they are below the threshold (see Table 7).

Treatments with NSES and NLES exhibit similar behaviour: a decrease in water absorption and an increase in evaporation resistance compared to untreated specimens. The former treatment shows increased cohesion and greater penetration depth, while the latter shows no variation in cohesion compared to the untreated specimen (Figure 5). The colour variation in the NSES treatment is slightly over the limit, while that of the NLES treatment is significantly higher (Table 7) due to whitening associated with the superficial precipitation of nano-lime (Calosil E25), which can be mitigated by applying the product with interposed Japanese paper. In the case of NSES treatment, negative effects can be considered acceptable, such as a slight (5%) increase in the evaporation resistance coefficient compared to untreated specimens and a colour variation close to the limit level (ΔE = 5).

The treatment with diNS does not induce any changes (positive or negative) with respect to the untreated specimen; this is probably because the product was applied at a low concentration.

In terms of comparing the results obtained in this study with those of other similar studies, as far as we are aware, only a few studies were conducted on earthen structures, such as adobe, rammed earth and renders. The bibliography for ethyl silicate, the most commonly and longest-used product, is slightly more extensive [12,13,15,16,17,21,40,46,47]. The results obtained by other authors are quite similar to those observed in this study. While ethyl silicate is effective in increasing surface mechanical strength, there may be slight to noticeable colour variations (ΔE > 5), depending on factors such as surface morphology, application technique, product concentration and the thermo-hygrometric conditions of the substrate [16,17]. Although ethyl silicate penetrates the earthen substrate well and reduces water ingress, it can also significantly decrease water vapour permeability [12,15,16,17,21,40,46,47,48]. This study used the DRMS test to directly observe that this issue is due to the formation of a hard, poorly porous crust.

The presence of a hard surface crust, not only in the case of ES, but also for the samples treated with MoES and NS, is probably due to the high concentration of the product, which, during the evaporation of the solvent, is drawn to the surface, leading to the formation of the crust. In specimens treated in mixtures with diNS and Wacker at concentrations of 3 and 5%, there is no crust formation. So, testing these products at lower concentrations could prevent crust formation. It is also possible that changing the application method, which involves repeated cycles rather than a single application, can help to avoid the crust formation or at least reduce its “hardness”. However, in the case of nanoproducts, the formation of the crust could also be due to poor penetration of the particles, resulting in their accumulation on the surface. This may have been the case with the MoES product. In any case, the formation of a surface crust is a potentially harmful occurrence that should be avoided. This is because it creates discontinuity that could lead to detachment over time.

The bibliography on silica nanoparticles is more limited, but the available studies report consistent results [16,17]. Although nano-silica treatment does not significantly affect water vapour permeability or colour changes compared to ethyl silicate, its consolidating power makes it suitable for slightly degraded surfaces, as it can penetrate a few hundred microns into the material [16]. These results are consistent with the findings of this study. It should be noted, however, that products applied to different materials (e.g., adobe, rammed earth or plaster/renders) with different compositions can be difficult to compare, and their effectiveness in the laboratory may not be confirmed in situ [16,17]. To the best of our knowledge, only one previous study has proposed using a mixture of nanoparticles and ethyl silicate to consolidate earthen structures [13]. Salazar-Hernández et al. [13] used a mixture of silica and TEOS nanoparticles on adobe samples, achieving positive results. Adding silica nanoparticles to TEOS improved its performance in terms of penetration depth and increased mechanical strength. As previously highlighted, these results align with our own. Additionally, Salazar-Hernández et al. [15] found that adding polydimethylsiloxane further improved the results. The authors of this article are not aware of any studies that have tested a mixture of calcium hydroxide and ethyl silicate nanoparticles on earthen materials, such as the SurfaPore FX WB product.

6. Conclusions

Six different earth consolidation treatments, five of which were nanoparticle-based, were tested on adobe brick samples taken from ancient rammed-earth walls in China. The results demonstrate that these products can be effective on such earthen structures. However, as previously noted, research on the application of nanoparticle-based consolidants to earthen materials remains limited.

Among the tested treatments, NSES appears to be the most promising. This treatment shows several positive effects, such as decreasing the water absorption coefficient and increasing cohesion without crust formation, while some side effects, such the slight increase in the evaporation resistance coefficient (5%) or colour variation close to the visible-by-naked-eye threshold (ΔE = 5), can be considered acceptable.

Concerning the MoES, ES and NS treatments, it should be noted that the formation of a hard surface crust is a critical event for conservation purposes. This is because it creates discontinuity that could lead to detachment over time.

Furthermore, this issue could be mitigated or avoided by adjusting either the concentration of the products or the method of application—particularly in the case of nanoparticles. These results suggest promising prospects for the consolidation of earthen surfaces, even in specific climatic contexts such as that of Ulanbay, China.

Nevertheless, a comprehensive evaluation of the compatibility and efficacy of these products requires further investigation. Additional experimental studies should be carried out to optimise the formulations for direct on-site application and to assess their durability and resistance to weathering, particularly with regard to their long-term performance under outdoor exposure conditions.

Furthermore, the provenance and composition of earthen materials should be carefully considered, since variations in mineralogical and granulometric characteristics can result in different responses to treatment. Consequently, experimental campaigns and preliminary laboratory investigations remain essential for identifying the most suitable consolidation product.

This study provides preliminary insights that could contribute to the development of more targeted conservation strategies for earthen architecture. At the same time, however, it highlights the need for further research to assess the effectiveness and applicability of these in situ consolidation treatments.