Application of Titanium Carbide MXenes in Chemiresistive Gas Sensors

Abstract

The titanium carbide MXenes currently attract an extreme amount of interest from the material science community due to their promising functional properties arising from the two-dimensionality of these layered structures. In particular, the interaction between MXene and gaseous molecules, even at the physisorption level, yields a substantial shift in electrical parameters, which makes it possible to design gas sensors working at RT as a prerequisite to low-powered detection units. Herein, we consider to review such sensors, primarily based on Ti₃C₂T_x and Ti₂CT_x crystals as the most studied ones to date, delivering a chemiresistive type of signal. We analyze the ways reported in the literature to modify these 2D nanomaterials for (i) detecting various analyte gases, (ii) improving stability and sensitivity, (iii) reducing response/recovery times, and (iv) advancing a sensitivity to atmospheric humidity. The most powerful approach based on designing hetero-layers of MXenes with other crystals is discussed with regard to employing semiconductor metal oxides and chalcogenides, noble metal nanoparticles, carbon materials (graphene and nanotubes), and polymeric components. The current concepts on the detection mechanisms of MXenes and their hetero-composites are considered, and the background reasons for improving gas-sensing functionality in the hetero-composite when compared with pristine MXenes are classified. We formulate state-of-the-art advances and challenges in the field while proposing some possible solutions, in particular via employing a multisensor array paradigm.

1. Introduction

Since 2011, there has been a high level of scientific interest in 2D nanomaterials, called MXenes, with the general formula M_n+1X_nT_x, where M is an early transition metal such as Ti, V, Nb, Cr, or Mo; X is C or N; and T is a surface functional group, primarily –F, –OH, or –O, which have unique properties due to a layered structure and variations in elemental composition [1,2,3]. These materials are distinguished by: (i) a high surface-to-volume ratio, which allows for an efficient transfer of surface gas adsorption into electrical properties when combined with an extremely low electrical noise; (ii) a high electrical conductivity that extremely depends on the synthesis method and may vary from metallic to semiconducting nature; and (iii) a high variability of available compositions that could be tuned via the chemical nature of the metal and/or functional groups on the surface in order to adjust the bandgap and work function. These remarkable features of MXenes provide researchers with great opportunities for numerous applications in diversely targeted units. Among others, the high specific surface makes it possible to harden composite materials of different chemical natures [4,5,6,7,8]. The possibility to intercalate various ions or molecules between the 2D layers [9,10,11] and high electrical conductivity led to a focus of major interest on the MXene application in R&D of energy generation and storage devices, lithium/sodium-ion batteries and supercapacitors [12,13,14,15,16,17,18], and fuel cells [19,20]. They are also widely considered in photo- and electrocatalytic processes for CO₂ reduction, hydrogen production, industrial water purification, electrocatalytic nitrogen reduction to form ammonia, etc. [21,22,23] and to develop conducting coatings in optoelectronics [24,25].

However, one of the most prospective ways to employ such a material lies in gas sensors. As noted, the targeted surface functionalization of MXenes could advance a selectivity for gas adsorption that could be registered even at room temperature (RT), which is important for energy-efficient sensor devices. Therefore, MXene structures are widely studied as sensing materials [26,27,28,29,30,31,32]. The number of publications in 2021–2022 with keywords “Sensor* and MXene” (CAS, SciFindern, September 2022) cumulatively increases by more than 2.5 times compared with the ones reported in 2013–2020 (Figure 1a), which was 1374 against 492. So far, the major focus has been on investigating the receptor properties, such as carbide MXenes, of Ti₃C₂T_x and Ti₂CT_x formulas (Figure 1b).

In spite of the abundance of numerous review reports on the application of MXenes for liquid and gas sensors [33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51], there is some gap in the systematic research on employing titanium-containing phases of Ti₃C₂T_x and Ti₂CT_x as receptor materials for chemiresistive gas sensors, including (i) a theoretical analysis of sorption and charge transfer, (ii) generalization of experimental data on the influence of the nature of functional surface groups and different modifiers, (iii) systematization of information on the gas sensitivity of nanocoating, and (iv) the analysis of the influence of the nature of functional surface groups and different modifiers. These issues are thoroughly considered in the present review.

2. Ti₃C₂T_x and Ti₂CT_x MXenes Preparation Methods

First report on synthesis of 2D Ti₃C₂ carbide as a result of selective removal of Al atom layer from the structure of Ti₃AlC₂ MAX-phase [52] was published just 11 years ago by Naguib and coworkers [1]. Already in the primary works by Gogotsi and Barsoum group [1,2,3], the method of aluminum etching by HF acid according to the following reaction was considered:

M_n+1AlX_n + 3HF = M_n+1X_n + AlF₃ + 1.5H₂

(1)

M_n+1X_n + 2HF = M_n+1X_nF₂ + H₂

(2)

M_n+1X_n + 2H₂O = M_n+1X_n(OH)₂ + H₂

(3)

This method became possible due to the considerably different nature and strength of the M–X (X = C, N) chemical bonds in M_n+1X_n layers and M–Al chemical bonds, which are responsible for keeping these layers together in MAX-phases. The former bonds are strong covalent–metal interactions, while the latter M–Al bonds are relatively weak, mostly metallic bonds, which can be torn by reacting with certain compounds. In case of MAX-phases including more electronegative element of Si instead of Al as binding element, an oxidizing agent is needed in addition to HF, such as H₂O₂, FeCl₃, HNO₃, NH₄S₂O₈, KMnO₄ etc. [53].

The high toxicity of hydrofluoric acid used in the etching process and its excessively high reactivity result in not only more defective MXenes but also a dramatic heating of the reaction system at the initial stage of synthesis. In its turn, together with a large outgassing, it often leads to spewing from the vessel or to oxidation of the resulting MXene particles. Therefore, another milder approach has been developed to selectively etch Al under the influence of HF generated in situ via interacting hydrochloric acid with lithium, sodium, potassium, and iron fluorides [54,55,56,57,58]. With relatively slow release of HF molecules, the synthesis process of MXenes can be easier controlled, and the output layered products appear to be less defective and to contain relatively few M_n+1X_n layers. In addition, the etchant composition significantly affects the kind of surface functional groups that form, such as –F, –OH, and –O, –Cl, which frequently manage the electrochemical properties of MXenes. Application of the NH₄HF₂-HCl system also allows one a simultaneous delamination due to intercalation of NH⁴⁺ cation into the interlayer space [59].

Obtaining single or few-layered MXene plates using the “HCl–MF” etching system, where M is Li⁺, Na⁺, K⁺, NH⁴⁺ cations, etc., usually requires an additional stage of delamination, i.e., exposure to intense ultrasound or organic molecule intercalation followed by splitting. This often results not only in the M_n+1X_n layers’ separation but also in the reduction of their surface area and an increased defect rate. In order to get large and non-defective Ti₃C₂T_x particles, the technique of Ti₃AlC₂ MAX-phase interaction with HCl-LiF system was optimized to be called Minimally Intensive Layer Delamination (MILD) [60,61,62,63,64]. In this case, the higher n(LiF):n(M_n+1AlX_n) and n(HCl):n(LiF) ratios facilitate etching of aluminum layers and make it possible to separate accordion-like MXene aggregates into individual plates without using ultrasound; just shaking is sufficient. The morphology of MXenes synthesized by the MILD method, their arrangement in aggregates after drying, as well as their mechanical and electrophysical properties, essentially differ from those of products derived by the conventional etching method with lithium fluoride solution in hydrochloric acid.

Hydrothermal etching technique conducted at 270 °C with assistance of NaOH [65] is one of the most practically interesting for obtaining Ti₃C₂T_x MXenes with only OH– and O– species on the surface.

Some works [66,67] show the prospects of preparing fluorine-free MXenes as a result of electrochemical etching via taking, for example, HCl acid as a medium under moderate heating [67]:

Ti₃AlC₂ − 3ē + 3Cl⁻ = Ti₃C₂ + AlCl₃

(4)

Ti₃C₂ + 2OH⁻ − 2ē = Ti₃C₂(OH)₂

(5)

Ti₃C₂ + 2H₂O = Ti₃C₂(OH)₂ + H₂

(6)

The known problem of rather fast oxidation of MXenes in aqueous media as well as the fact that the nitride MXenes are formed with a very low yield in contrast to the carbide phases when using etching with HF or HCl-LiF solutions led to the development of synthesis methods in molten salts. In Refs. [55,68], a reaction route to fabricate Ti-containing MXenes in the melts of LiF-NaF-KF and NaF-KF fluoride systems was proposed as follows:

Ti₂AlC + Na⁺ + 2K⁺ + 6F⁻→Ti₂CT_x + K₂NaAlF₆

(7)

The functional parameters of MXenes, such as bandgap, electron mobility at RT, work function, etc., depend very strongly on the surface groups. However, for some applications, the fluoride substituents are not suitable. Therefore, the approach of thermal treatment in molten chlorides, such as ZnCl₂ [69,70] and other Lewis acids, such as CuCl₂ [71], proved to be quite effective. In the case of ZnCl₂ melt, either the substitution of Al atoms by Zn in the MAX phase, Reactions (8) and (10), or the synthesis of Cl-substituted Ti₃C₂Cl₂ MXene with release of zinc metal particles as a by-product, Reactions (9) and (11) occur in dependence on the proportions of the components [69,70]:

Ti₃AlC₂ + 1.5ZnCl₂ = Ti₃ZnC₂ + 0.5Zn + AlCl₃↑

(8)

Ti₃AlC₂ + 1.5ZnCl₂ = Ti₃C₂ + 1.5Zn + AlCl₃↑

(9)

Ti₃C₂ + Zn = Ti₃ZnC₂

(10)

Ti₃C₂ + ZnCl₂ = Ti₃C₂Cl₂ + Zn

(11)

The prepared chlorine- and bromine-substituted MXenes are convenient to further modify the surface groups to –S, –Se, and –Te ones in case of necessity to adjust the material properties [72].

In addition, alternative methods for MXene preparation, such as chemical vapor deposition (CVD), other gas-phase methods, and template synthesis, have been actively developed in recent years [73,74,75,76,77].

As a result of liquid-phase etching, packages of MXenes are formed with layers bound to each other by Van der Waals forces. Their morphology is very recognizable and has been almost universally named “accordion-like MXenes.” Depending on the method of synthesis, the degree of aggregation can vary greatly. For example, employing acid solutions of lithium/sodium/ammonium fluorides not only allows one to selectively remove layers of aluminum atoms from MAX-phase structures, but due to cation intercalation, the parallel process of delamination is also possible.

Typical and historically first method of delaminating the MXene layers was a separation under the influence of ultrasound [78], whose prolonged application results in a reduction of the MXene plate area. However, intercalation of some organic molecules greatly facilitates and accelerates the process. Among the most frequently used ones are tetramethylammonium and tetrabutylammonium hydroxides, dimethyl sulfoxide, isopropylamine, urea, and others.

3. Application of Individual Ti₃C₂T_x and Ti₂CT_x MXenes in Chemiresistive Gas Sensors

3.1. Sensors Based on Pristine Ti₃C₂T_x

One of the first theoretical works aimed at the analysis of the possibility to employ MXene Ti₃C₂T_x as a receptor material in chemiresistive gas sensors, particularly for ammonia detection, has been published in [79], which in addition to experimental studies simulated the interaction between MXene with completely substituted oxygen surface groups of the Ti₃C₂O₂ formula and eight gases: CH₄, H₂S, H₂O, C₂H₅OH, CH₃OH, CH₃C(O)CH₃, NH₃, and NO. First-principles calculations made it possible to determine the optimal adsorption geometry and evaluate the energy of this process [79]. The E_ads values indicate significantly stronger interactions of NH₃, ethanol, methanol, and acetone molecules with the Ti₃C₂O₂ surface compared with that of CH₄, H₂S, and H₂O. The expression of E_ads in normalized eV/atom units, not in eVs, strongly emphasizes the ammonia molecule, which delivers the characteristic charge transfer value of 0.153e.

In the work by Khakbaz et al. [80], the electrostatic surface potential (ESP), projected density of states (PDOS), and thermal analysis methods were utilized to study the adsorption of NH₃, NO, NO₂, N₂O, CO, CO₂, CH₄, and H₂S molecules on the surface of MXene phases of Ti₃C₂T_x with various ratios of surface species that are most commonly formed by the often applied techniques for their synthesis: Ti₃C(OH)_0.44F_0.88O_0.66 (relatively high F-group content), Ti₃C₂(OH)_0.66F_0.22O_1.11 (relatively low F-group content), and Ti₃C₂(OH)_0.44F_0.88O_0.66(H₂O)_0.44 (with adsorbed water molecules). The calculation results have shown that the composition with high fluoride surface groups of Ti₃C₂(OH)_0.44F_0.88O_0.66 is most sensitive to ammonia due to the maximum charge transfer (−0.098e) and minimum adsorption energy (−0.36 eV). It was found that sorption of the NO₂ molecule on the surface of this composition results in the formation of nitrous acid from HNO₂ as a result of proton detachment from the OH-group by the adsorbed molecule of the gas, the chemical bonding in the OH-group being weakened due to the high content of electronegative F-groups. It was shown that for MXene substrates, the adsorption energy goes down from −0.36 eV to −0.49 eV with a reduction in the number of F-groups and a corresponding enhancement in the number of oxygen-containing surface substituents, which leads to an increase in the sensor recovery time. At the same time, the charge transfer value for MXenes with different –F/(–OH and –O) ratios does not substantially change. If water molecules are present in the MXene receptor layer, a stabilization of the sorbed NH₃ molecule on the surface occurs due to the formation of hydrogen bondings. However, in this case, a decrease in the interaction between the titanium and nitrogen atoms and, accordingly, the charge transfer down to 0.012e, i.e., a reduction by almost 90%, were also recorded. Thus, it is shown [80] that the chemical composition of the surface functional groups of MXene Ti₃C₂T_x as well as the presence of surface-adsorbed water molecules have a significant influence on the chemosensor properties of this material.

In work [81], density functional theory (DFT) was applied to examine features of the adsorption of ammonia molecules on Ti₃C₂O₂ MXene, which is characterized by titanium defects. It is shown that the stronger physical interaction of the NH₃ molecule with the MXene monolayer is observed at the titanium deficit, with a higher negative corresponding adsorption energy of −7.44 eV when compared with −6.29 eV for the defect-free Ti₃C₂O₂. At the same time, the charge transfer has slightly decreased, from 0.226e to 0.115e. The calculations using an atomic-scale gas sensor model for studying electron transport indicate (Figure 2) that, in accordance with the I-V curves under ammonia molecule adsorption, there is a more significant current drop for titanium-deficient Ti₃C₂O₂ in contrast to the reference MXene. According to the authors of [81], this proves that the introduction of titanium vacancies can increase the sensitivity of Ti₃C₂O₂-based sensors for NH₃ detection.

In theoretical work [82] for MXenes with different surface substituents (Ti₃C₂O₂, Ti₃C₂F₂, and Ti₃C₂(OH)₂) and defects (point vacancies), the adsorption process of SF₆ decomposition products (SO₂, SOF₂, and H₂S) as markers for inevitable partial discharge in gas-insulated switchgear was investigated by DFT. It was shown that no stable adsorption configuration and low adsorption energy were found for the interaction of the defect-free Ti₃C₂(OH)₂ with SOF₂ and H₂S gases, whereas for pure Ti₃C₂O₂ and Ti₃C₂F₂, all the optimized adsorption configurations are stable, but the low energy and large adsorption distance imply a very weak adsorption of all the above-mentioned SF₆ decomposition products. Due to this, calculations with embedded point vacancies for oxygen and for fluorine were performed for the latter, which indicate a 2–4-fold increase in the adsorption energy for all selected test gases. In this case, physical adsorption is replaced by chemisorption with significant charge transfer. The most negative values of adsorption energy on Ti₃C₂O₂ and Ti₃C₂F₂ with the corresponding point vacancies were obtained for SO₂, which indicates that the highest sensitivity should be observed for its detection.

Another group of researchers, Zeng and colleagues [83], also studied the features of the adsorption of gaseous products of sulfur hexafluoride oxidation (H₂S, SO₂, SOF₂, SO₂F₂) on Ti₃C₂T_x MXene with different functional groups (O, F, OH). Here, DFT calculations allowed the authors to establish that all selected gases should adsorb spontaneously on Ti₃C₂O₂ and Ti₃C₂(OH)₂, and charge transfer reduces in the series Ti₃C₂(OH)₂ > Ti₃C₂O₂ > Ti₃C₂F₂. For the most promising composition, Ti₃C₂(OH)₂, the charge transfer upon adsorption of these four marker gases is in the range from 0.19e (H₂S) to 1.27e (SO₂F₂). The work published in [84] continues a series of theoretical studies by Zeng’s group aimed at studying the efficiency of detection of SF₆ decomposition products formed as a result of insulation overheating or partial discharge. However, in this case, it is taken into account that sulfur hexafluoride is often used in a mixture with nitrogen to reduce the negative effects of SF₆ release into the environment. As a result, in addition to sulfur-containing oxidation products, it is necessary to consider the possibility of formation of such molecules as NO, NO₂, N₂O and NF₃. The data of the first-principle calculations exhibited that all compositions of Ti₃C₂T_x (T = F, O, OH) are predicted to have good prospects for the detection of NO and NO₂, while Ti₃C₂(OH)₂ has a large negative adsorption energy for all four gases (from −1.7 eV to −9.1 eV, characterizing NO and NF₃, respectively) and significant charge transfer (−1.6e for NO to −1.4e for NF₃), which should facilitate the detection, but long recovery times are assumed for these parameters.

As emphasized in one of the first sensing studies by Lee and colleagues [85] involving Ti₃C₂T_x MXene, continuous monitoring of human physiological state by exhalation composition requires the integration of gas sensors into portable and wearable electronic hardware that is operable at RT and possesses flexible substrates that provide a natural fit to the human body. To study the possibility of using Ti₃C₂T_x MXene as a gas-sensitive material, derived by selective etching of aluminum from the corresponding Ti₃AlC₂ MAX-phase by LiF + HCl (9M) with subsequent water washing and ultrasonic treatment, the dispersion was applied to a flexible polyimide substrate equipped with platinum counter-pin electrodes. The coating was dried in a desiccator. The thickness of the formed multilayer film was 30 nm. An examination of the response to 100 ppm of ethanol, methanol, acetone, and ammonia at RT showed (Table 1) that in all cases there was a response characterized by an increase in resistance and a systematic enhancement of baseline resistance, especially when detecting ammonia. The authors believe [85] that the binding of gas molecules to MXene in the present study is due to interactions with both Ti₃C₂T_x surface defects and surface functional groups, primarily –O and –OH. According to the authors, the adsorbed water and oxygen molecules introduced under etching the MAX-phase play an important role. Possible reactions occurring during the sorption of the NH₃ molecule on the MXene surface with a charge transfer have been suggested, with the formation of an electron that then recombines with the hole present in the MXene, resulting in the resistance increase [85]:

2NH₃ + 3O⁻ → N₂ + 3H₂O + 3ē

(12)

NH₃ + OH⁻ → NH₂ + H₂O + ē

(13)

The responses were 0.210, 0.143, 0.115, and 0.075 for ammonia, methanol, ethanol, and acetone, respectively [85]. The data obtained are consistent with the available theoretical studies [86] and indicate an increased absorption energy for the Ti₂CO₂-NH₃ pair, which may account for the baseline drift and the high sensor recovery times observed in study [85]. In addition, the change of response to acetone in the concentration range of 25–200 ppm was studied in [85] to observe a linear dependence.

The scientific task set forth in [87] relates to the need for a correct determination of low concentrations of volatile organic compounds (VOCs) in the ppb range at RT, where a maximum signal-to-noise ratio should be observed. It was found that the deposited receptor layer of Ti₃C₂T_x MXene synthesized under etching by LiF + HCl according to a methodology similar to one reported in [85] showed a high response to 100 ppm ethanol, acetone, propanal, and ammonia. For ethanol, the response was almost twice as high as that for ammonia. DFT simulations highlighted this feature by the predominant presence of F and OH as surface functional groups, and the highest estimated sensitivity to both gases was predicted for the hydroxide-substituted Ti₃C₂(OH)₂ MXene. For acetone, ethanol, and ammonia, i.e., gases capable of participating in hydrogen bonds, the detection limit was defined as equal to 50–100 ppb [87]. The authors noted the record-low electrical noise of the sensor, especially under the detection of VOCs, which is due to the high electrical conductivity of 3250 S/m of the MXene.

Wu and colleagues [79] experimentally investigated the gas-sensitive properties of single-layer Ti₃C₂ MXene, synthesized by etching the corresponding MAX phase with sodium fluoride solution in hydrochloric acid, at 25 °C as part of a chemiresistive sensor in a cylindrical coating geometry versus to 500 ppm of a number of gases (CH₄, H₂S, H₂O, NH₃, NO, ethanol, methanol, and acetone) under a varying humidity background. A high selectivity towards ammonia was found; the second most intense response to ethanol is only 24% of that to NH₃. It seems to relate to the high adsorption energy of this analyte. The authors noted a significant increase in the response to ammonia when compared with the similar Ti₃C₂ sample obtained via etching with LiF + HCl [85]. This is explained by the cleaner impurities of metal cations on the surface of the MXene layer and, thus, an increased area for adsorption. In experiments performed to study the effect of humidity on ammonia detection [79], the response was found to go up with increasing humidity from 30 rel.% to 40–90 rel.%, and the response values at RH = 50–90% were very close. In addition, the phenomenon of some signal saturation at high NH₃ concentrations was observed, which seems to be associated with the difficulty of desorption of this gas at RT and leads to a gradual increase in the baseline resistance. As mentioned above, the simulation of Ti₃C₂O₂ interactions with eight gases (CH₄, H₂S, H₂O, C₂H₅OH, CH₃OH, CH₃C(O)CH₃, NH₃, and NO) based on first-principle calculations [79] has shown promising sorption of specifically ammonium molecules, which is characterized by a high adsorption energy and charge transfer. The significant discrepancy between the low experimental sensitivity to NO and the very high values of charge transfer and adsorption energy calculated for this molecule was attributed by the authors [79] to the low content of OH- and F-groups on the surface of MXene.

A significant influence of the composition of the surface groups and the ion intercalation into the interlayer space on the chemosensory behavior of MXenes has been concerned in a fairly wide range of studies [88,89,90]. For example, Yang and colleagues [88] studied the effect of additional treating the Ti₃C₂T_x MXenes with sodium hydroxide solution on RT sensitivity to ammonia and atmospheric moisture. These MXene structures were synthesized from Ti₃AlC₂ MAX-phase with etching in a 45% HF solution at RT. The authors found that such a MXene processing results in (1) the removal of some of the surface F-groups and their replacement by OH-species, increasing the ratio of n(O)/n(F) by 2.5–3 times, and (2) the intercalation of Na⁺ cations into the interlayer space of the material. For the NaOH-impregnated Ti₃C₂T_x MXene, an inversion of the response to moisture and ammonia influx was observed: there was an increase in conductivity rather than an increase in resistance, as opposed to the original Ti₃C₂T_x. Other sensory characteristics also changed. The experiments at 38 rel.% humidity showed that there was an almost two-fold increase in the response to ammonia and improved selectivity towards, in particular, NO₂. The authors explained [88] the increased sensitivity to humidity by the formation of aqua-complexes, of [Na(H₂O)_m]⁺, on the surface of MXene layers, which advances the adsorption capacity of the material. Increasing the number of surface OH-groups led to improved NH₃ binding. The simulation data presented in [87] show that the most negative adsorption energy is observed for Ti₃C₂(OH)₂ in comparison to Ti₃C₂F₂ and Ti₃C₂O₂. The change in the signal direction, however, as believed by the authors [88], may be due to a change in the carrier type in Ti₃C₂T_x after functionalization with oxygen.

Another study [89] also addressed the role of the process of gas intercalation into the interlayer space of the Ti₃C₂T_x MXene upon detecting gases of different chemical natures as well as the influence of the intercalation of sodium cations as a result of processing in alkaline solutions. It was found that ethanol vapors, to which Ti₃C₂T_x is sensitive, resulted in a significant shift of the (002) reflex position, which corresponds to an increase in the interlayer distance of the MXene film from 13.24 Å to 14.06 Å. At the same time, such a process does not occur upon CO₂ exposure, for which Ti₃C₂T_x lacks gas sensitivity. Preliminary intercalation of sodium cations as a result of impregnation of MXene layers with NaOH solutions of various concentrations leads to an increased response to both ethanol and CO₂, since the adsorption of ethanol in the interlayer space is controlled by hydrogen bonds with functional surface groups of O/OH and the water present. It was found that there is an optimal concentration of NaOH solution of 0.3×10⁻³ mol/L, which allows one to obtain the maximum selectivity when detecting C₂H₅OH vapors. As a result, the authors in [89] conclude that the cause of the universal positive response to gaseous analytes of different natures (oxidants and reducing agents), along with the influence of the metallic nature of MXene conductivity, arises from interlayer swelling due to difficulties in electron transfer outside the Ti₃C₂T_x layer plane, where an increase in electrical resistance occurs.

The effect of ionic intercalation upon addition of electrolyte solutions (HCl acid, neutral KCl salt, and KOH alkali) to the colloidal solution of delaminated Ti₃C₂T_x MXenes, synthesized upon exposing Ti₃AlC₂ powder to a mixture of lithium fluoride with hydrochloric acid, on the sensitivity of chemiresistive gas sensors has been considered in research reported in [90]. As a part of the study, the optimal concentrations of modifying solutions, which do not lead to MXene aggregates, were established as being equal to 0.01 M for KCl and KOH and 0.001 M for HCl. It was shown that the introduction of KOH, in contrast to KCl and HCl, makes it possible to increase the number of surface OH-groups and also leads to the formation of films with the largest interlayer distance in MXenes due to the intercalation of potassium cations. The authors attribute the efficiency of potassium ion intercalation into the interlayer space of MXene specifically for the Ti₃C₂T_x-KOH sample to the increased content of surface OH-groups capable of deprotonation, the formation of the enolate form (Ti-O-), and further intermediate compounds (Ti-O-K). The obtained Ti₃C₂T_x -KOH structure exhibited the highest response when detecting 100 ppm ammonia [90]. The response was 3.8%, which is higher than the 0.9% observed in the pristine Ti₃C₂T_x. The minimum response was observed for the KCl-treated sample, equal to 0.4%. Sensitivity analysis to ethanol vapor, 100 ppm, showed that the Ti₃C₂T_x-KOH sample had the highest response to this analyte, ~3.4%, which was significantly higher than that of the Ti₃C₂T_x (0.8%) and the other samples (<0.5%). The study of the response of the obtained sensors under a higher level of background humidity (RH = 30%) revealed that the response to ethanol for the Ti₃C₂T_x-KOH, though significantly decreased, was still approx. four times higher than the response of other samples. With further increasing the humidity up to 50 rel.%, responses significantly deteriorate and become close for all the samples, equal to 0.25–0.5%, which the authors attribute to the additional adsorption of water molecules on the surface of MXene, which blocks adsorption centers for the analyte gases.

In a study by Shuvo et al. [91], taking into account such drawbacks of MXenes as their tendency to agglomerate under coating formation and a rather small interlayer distance, which hinders gas diffusion and reduces the effective surface-to-volume ratio, the effect of Ti₃C₂T_x doping with sulfur heteroatoms on the chemisensor properties was studied. For this purpose, the fresh Ti₃C₂T_x nanomaterial synthesized by the exposure of Ti₃AlC₂ powder to LiF + HCl was ground in a mortar with thiourea and heat-treated at 500 °C for 3 h in an Ar atmosphere. It was found that the alloying allowed for increasing the interlayer distance, which was expressed in a shift of the (002) reflex relative to that for Ti₃C₂T_x as well as confirmed by HR-TEM inspection; a more than two-fold increase in the distance between the layers was observed, from 0.96 nm (Ti₃C₂T_x) to 1.91 nm (Ti₃C₂T_x-S). For the Ti₃C₂T_x and Ti₃C₂T_x-S receptor materials deposited on a flexible polyethylene terephthalate substrate equipped with 20 Ag counter-pin electrodes, responses to a number of VOCs were measured. An unexpected fact was established: among the studied series of analytes (hexyl acetate, toluene, hexane, and ethanol), the highest response for both sensors was recorded for toluene vapor, as shown in Figure 3. For the Ti₃C₂T_x sample, the conventional increase in resistance was observed upon a gas injection, while for the Ti₃C₂T_x-S sample, there was a decrease in resistance. This anomalous behavior of the sensor based on sulfur-doped Ti₃C₂T_x was attributed by the authors to the presence of sulfur atoms on the surface, which leads to the extraction of electrons from the conduction band of MXenes with the formation of ionic forms of S^γ−, e.g., S⁻, S²⁻, and an electron-depleted layer. These processes lead to increased receptor layer resistance. Following the suggestion of the authors, when a gaseous analyte appears, chemical processes with a reduction in sulfur compound concentration are possible, such as forming oxygen ionic species in MOS sensors, which can thereby increase the electron concentration and, consequently, advance the conductivity of the material.

The response value for sulfur-doped MXene exceeded the same for pristine MXene by 3–4 times. For example, the response to 50 ppm toluene for Ti₃C₂T_x-S was ca. −79.5%, while for Ti₃C₂T_x it was about +19.3%, which the authors related to increasing the interlayer distance in the MXene [91]. For the Ti₃C₂T_x-S sample, a detection limit of 500 ppb toluene was adjusted. Still, the higher sensitivity to toluene for both sensors is explained by the authors to come from the higher activity of the aromatic ring, which is enhanced by the methyl group due to its electron-donating effect. DFT calculations showed that the bond energy of the toluene molecule to the sulfur-doped functional surface groups has a much more negative value than for Ti₃C₂O₂, which is consistent with the experimentally observed increased response for the Ti₃C₂T_x-S sample.

The problem of the rather low oxidative stability of MXenes is discussed in the study by Chen et al. [92], who proposed an alternative way to stabilize their structure via a surface functionalization with perfluoroalkoxysilane for further application in the detection of VOCs. For this purpose, the Ti₃C₂T_x MXene powder obtained by etching the MAX phase of Ti₃AlC₂ with LiF/HCl was functionalized in (3-chloropropyl)trimethoxysilane and 1H,1H,2H,2H-perfluorooctyltriethoxysilane solutions, which interacted with the existing O- and OH-groups on the surface to yield Ti₃C₂T_x-Cl and Ti₃C₂T_x-F samples, respectively. In this case, superhydrophobicity was demonstrated for the latter, expressed in a sharp increase of the boundary wetting angle from 33 to 156. It was also noted that the microstructure of Ti₃C₂T_x-F is not characterized by the accordion-like aggregates, typical for the original Ti₃C₂T_x and Ti₃C₂T_x-Cl, but instead appears in a porous three-dimensional architecture consisting of crumpled nanosheets, which facilitates gas diffusion. It was noted that storing the samples for 2 weeks under 100 rel.% humidity led to only a slight reduction of the electrical conductivity of the modified Ti₃C₂T_x-F sample of ca. 14%, which was further restored to 93% of the original value under a dry environment. In contrast, the same humidity storage resulted in a catastrophic 97% decrease in the electrical conductivity of the Ti₃C₂T_x sample. For both modified receptor materials, an enhanced response to all the studied VOCs was recorded, primarily to ethanol, acetone, and 2-propanol, which the authors attributed to a significant enlarging of the interlayer distance (Figure 4) [92].

Thus, the responses to 120 ppm ethanol for the Ti₃C₂T_x-F sample exceeded those for the original Ti₃C₂T_x MXene by more than two times and those for the Ti₃C₂T_x-Cl sample by approximately 1.5 times. For the Ti₃C₂T_x-F sample, it was found that the response to 30 ppm ethanol systematically decreased from 5.7% to 1.5% when the relative humidity of the gas environment went up from 5% to 80%. This seems to be due to a partial filling of the receptor material’s active centers with water molecules. Altogether, it was found that modification of the surface of Ti₃C₂T_x nanosheets with a perfluoroalkyl substituent leads to a significant increase in coating stability under the influence of oxygen and moisture for a long time, as well as to a significant improvement in gas-sensitive properties to VOCs at room temperature.

The study by Li et al. [93] deals with developing a Ti₃C₂T_x-based sensor for VOCs in exhaled air, aimed at noninvasive detection of diseases. As numerous works highlight [31], the MXene structures are extremely sensitive to humidity, which could interfere with VOCs and eliminate the response to other gases. The peculiarity of the exhaled air, where the water vapors exceed by hundreds of times the concentration of the target analytes, can lead to the negation of the available signals of the latter. In view of this problem, the authors tried to increase the hydrophobicity of Ti₃C₂T_x receptor coatings synthesized from MAX phase exposure to LiF/HCl by grafting a branched hydrocarbon substituent onto the surface OH-groups as a result of interaction with trimethylacetic anhydride to obtain the sample called Ti₃C₂T_x-M1. The second sample of Ti₃C₂T_x-M2 was additionally exposed to NaHCO₃ solution to increase the surface hydroxo-groups required for grafting the (CH₃)₃COO substituent. It was found that there is a significant shift in the XRD reflex position of the (002) plane in Ti₃C₂T_x-M1 when compared with Ti₃C₂T_x, going from 8.8° to 6.8°, which indicates an increase in the interlayer distance due to the introduction of bulk surface substituents. For the sample of Ti₃C₂T_x-M2, however, in addition to an even greater shift of the same reflex position down to 6.3°, a significant change in morphology with the formation of curved porous flake aggregates was observed. The authors explained it by a random distribution of (CH₃)₃COO-groups on the surface, which weakens a polarity for local areas and leads to non-uniform interaction between Ti₃C₂T_x layers. Due to an almost four-fold increase in the specific surface area, from 10 m²/g characterizing pristine Ti₃C₂T_x to 39 m²/g for Ti₃C₂T_x-M2, and the facilitation of gas diffusion into the interlayer space, a sharp increase in responses to all gases under test, including non-polar gases such as hexane, octane, and aromatic hydrocarbons, has been observed, as pictured in Figure 5.

However, the largest response was obtained for 40 ppm ethanol [93]. At the same time, by introducing a bulk organic surface modifier, the response to water vapor decreased most significantly, from 2% for the original Ti₃C₂T_x to 0.6% for Ti₃C₂T_x-M2. When detecting ethanol in the range of 20–500 ppm, it was found that there was an increase in the response in a power-dependent manner for all three samples, with the modified Ti₃C₂T_x-M2 sensor having the highest response [93]. Experiments on the long-term stability of the sensor performance showed that the Ti₃C₂T_x-M2 sample is characterized by a mild increase in the baseline associated with a decrease in conductivity due to some oxidation of the carbide core. At the same time, the responses to hexane and ethanol were reduced after 15 days of ambient exposure by only 41.9% and 31.7%, while the unmodified Ti₃C₂T_x lost ca. 90% of its sensitivity during this time. It is worth noting that the most promising sample, Ti₃C₂T_x-M2, was integrated into a wearable tag for VOC detection as a part of a face mask and demonstrated the ability to distinguish between normal breath and breath after alcohol consumption.

In recent years, there has been increasing interest in the use of non-delaminated MXenes, which are characterized by an accordion-like microstructure to be preserved during coating with a large specific surface area [94,95]. The structures of Ti₃C₂T_x were successfully employed to produce an ultrafast humidity sensor [94]. To enrich the MXene surface with OH-substituents under an accordion-like architecture, after selective etching of Al from Ti₃AlC₂ by a concentrated HF acid, washing with distilled water, and drying in a vacuum, the layered aggregates of Ti₃C₂T_x were not subjected to delamination. It was found that the chemiresistive response is reversible and follows a linear function in the relative humidity range of 11–90%. The sensor response and recovery times were estimated to be ca. 60 s. FT-IR studies for Ti₃C₂T_x in dry air and at 95% relative humidity indicate the appearance of a layer of liquid water on the surface of the MXene in the latter case. The authors explain this by the abundance of OH-groups on the surface, which can form continuous and ordered films of H₂O in the hierarchical Ti₃C₂T_x nanostructures, which are quickly grown by exposure to water vapor. In this case, the water molecules are stacked on top of each other, making an electrostatic field directed perpendicular to the surface of the MXene, which prevents the carrier transfer and, consequently, leads to an increase in the resistance of the receptor material.

The further evaluation of the efficiency of employing the accordion-like Ti₃C₂T_x derived by selective etching of Al layers from the MAX phase of Ti₃AlC₂ with concentrated HF (without delamination) as a sensor to detect low concentrations of acetone at 23 °C was performed by Majhi et al. [95]. They found out that the screen-printed receptor coatings could allow one to measure acetone at concentrations down to 0.25 ppm when the observed response was 17.3%. The response to acetone was quite selective. At a concentration of 100 ppm, it was equal to 99%, while one to 100 ppm of H₂ was only 15%. Notably, the sensor performance was preserved versus at least acetone vapors while storing the unit for 3. Under RH = 50%, the acetone response reduced from 99% (at RH = 0%) to 80%, though the humidity at the 90 rel.% pressure forced the response to drop to ca. 5%.

Another study published in [96] is somewhat different from the already described studies. However, it is worth noting as an important example of a virtual sensor design capable of fairly accurate detection of ethanol content in the presence of water and methanol. The idea was to employ a broadband impedance spectrum from a single Ti₃C₂T_x-based sensor and to process it by principal component and linear discriminant analysis methods. For this purpose, the sensor was exposed to 100–800 ppm of different VOCs (methanol, ethanol, acetone, isopropanol, acetonitrile, dichloromethane, hexane, and toluene) to record impedance over a wide range of AC frequencies. The authors have extracted eight representative parameters for each analyte at each concentration. Thus, unique fingerprints were created to characterize the gaseous test analytes, which were further used for their identification, including their appearance in complex gas mixtures. By performing multivariate mathematical data processing as well as cross-checking, the authors ensured a high level of gas identification and prediction of its concentration for the detection of tested VOCs in the presence of known and unknown interferences.

3.2. Sensors Based on Pristine Ti₂CT_x

The number of works, both theoretical and experimental ones, dealing with the application of Ti₂CT_x MXene as a receptor material for chemiresistive gas sensors is significantly less when compared with Ti₃C₂T_x. Nevertheless, the adsorption of gases on the Ti₂CT_x surface has been considered theoretically already in 2015. In this study, Yu et al. [86] performed first-principles simulations to simulate the interaction between Ti₂C MXene with oxygen surface functional groups and ammonia molecules. It was found to have a rather high charge transfer value, 0.174e from the nitrogen atom to the MXene, and a negative adsorption energy of −0.37 eV, which should result in the chemisorption of a given molecule on the Ti₂CO₂ surface (which is also expressed in distortion of both NH₃ and solid phase geometry in the Ti₂CO₂ monolayer approach) and, alternatively, should allow for a relatively fast desorption at RT. At the same time, it is shown that application of 3% strain to the layer leads to an even greater increase in the adsorption energy of NH₃ on the Ti₂CO₂ surface, up to 0.51 eV.

Thomas and coworkers used [97] the DFT method for estimation of phosgene adsorption onto the Ti₂CT_x MXene with surface substituents by F-, OH-, and O- to evaluate the effect of both the nature of functional groups and atomic defects. Analysis of the adsorption energy of the COCl₂ molecule over Ti₂C, Ti₂CF₂, Ti₂CO₂, and Ti₂C(OH)₂ monolayers indicated that Ti₂C(OH)₂, including the defective one, is the most promising structure to efficiently detect this gas. Moreover, the maximum charge transfer, of 0.310–0.311e, is characteristic for pure Ti₂C(OH)₂ and carbon-deficient materials. Another theoretical study [98] shows that a sufficiently high amount of water molecules adsorbed on the surface of the Ti₂C MXene facilitates a sharp, almost by an order of magnitude, increase in the response when detecting ethanol when compared with that for pure Ti₂C.

For the thinnest carbide MXene of Ti₂CT_x, both chemistry and experiments concerning gas-sensing properties have been studied in much less detail compared with Ti₃C₂T_x, despite the fact that, due to the smaller single layer thickness, it seems more promising for use in areas where surface processes are critical. The influence of the synthesis method and hence the composition of the surface functional groups on the sensory properties of the Ti₂CT_x has been studied by Sun et al. [99]. It was found that in the case of selective removal of Al from the Ti₂AlC MAX-phase using concentrated HF, which leads to an increased content of F-groups and a reduced content of OH-groups, 2D nanomaterials with p-type semiconductor properties are formed. At the same time, the technique of exposure to a mixture of hydrochloric acid and lithium fluoride allows one to obtain samples with metallic conductivity and a relatively high content of surface OH-groups. The ammonia sensitivity of this MXene synthesized by hydrofluoric acid etching is approximately 35% higher when compared with one of the Ti₂CT_x MXenes obtained by etching with LiF/HCl.

It was also noted that the Ti₂CT_x (HF) MXene is also characterized by faster desorption and restoration of the baseline; purging the cell with N₂ was sufficient for this purpose, while the second MXene required additional UV irradiation [99]. This seems to be due to the more negative adsorption energy calculated in [86] for the Ti₂C(OH)₂ compound in comparison with Ti₂CO₂. Thus, etching of Ti₂AlC with hydrofluoric acid allows one getting more promising results from a chemical gas-sensing standpoint of Ti₂CT_x MXenes, both in terms of the response value and kinetic parameters, when compared with using a mixture of lithium fluoride and hydrochloric acid.

The application of multilayer Ti₂CT_x MXene, obtained by selective etching of Ti₂AlC layers with a mixture of LiF and hydrochloric acid without delamination, as a methane sensor operating at RT was demonstrated in a study by Wang and colleagues [100]. They discovered a photocatalytic activity of this compound in the oxidation reaction of CH₄ under the influence of visible light, which helped them achieve their difficult goal. The authors showed that the rather low response to 1% CH₄ in darkness, equal to ca. 18.2%, strongly enhances up to ca. 142% when the sensor is illuminated by a visible light from a Xe lamp with a cutoff filter (λ > 420 nm) with a power of 0.85 mW∙cm⁻² (Figure 6).

A respective reduction of response time from 136 s to 38 s has been noted, too. It is worth noting that here the humidity also plays an important interference role. When the relative humidity was increased to 60–80%, a decrease in the response to 1% CH₄ down to ca. 111% was observed. Storage in the air for 90 days led to a decrease in response, down from 130% to 142%. Exposing this MXEne-based sensor to other gases, such as H₂S and CO, which usually accompany methane in coal mines under their normal concentrations, showed [100] that the selectivity to CH₄ also enhances under irradiation with visible light. The authors substantiate their idea of the observed photocatalytic activity of Ti₂CT_x by visible light via the complete inactivity of this MXene with respect to methane oxidation in darkness, while CH₄ is fully oxidized to CO₂ under illumination. In situ IR spectroscopy measurements of the MXene prior and after viz-light illumination were indicated by the appearance of a characteristic absorption band of CO vibrations at ~2340–2360 cm⁻¹, and gas chromatography data that identified CO₂ at the sensor outlet have been supplied as a confirmation.

In another study [101], a chemiresistive sensor based on few-layer Ti₂CT_x synthesized by etching in NaF + HCl with a subsequent delamination with tetramethylammonium hydroxide and ultrasonic processing exhibited an increased sensitivity to 100 ppm of NO₂, (a ca. 6% response) at 30 °C. With increasing temperature, however, the greatest response was observed toward 100 ppm ammonia; the response was equal to 13%, with a clear baseline drift. An interesting fact about the reversible detection of oxygen at 1% and 5% concentrations at 30 °C with high sensitivity was found; the corresponding chemiresistive responses were 8.6 and >270 (Figure 7). This opens the prospects for the application of MXene materials, in particular those based on Ti₂CT_x, as low-temperature sensors for oxygen.

3.3. Features of Gas-Detection Mechanisms in Pristine MXenes

Summarizing the studies presented above, it is necessary to consider the concepts formulated by their authors on a possible mechanism of detection of gaseous analytes, including water vapors, by receptor layers of individual MXenes.

The family of titanium carbide MXenes, such as Ti₃C₂T_x and Ti₂CT_x, is known to have a metallic conductivity while their outer surface is ordinarily coated with polar functional groups. This combination makes this class of compounds attractive for high signal-to-noise ratio (SNR) gas sensors. High SNR, i.e., the ratio of high signal intensity upon gas detection to low electrical noise intensity, is achieved due to the functionalization of the MXenes surface by various groups that provide a strong chemical binding to the analyte gases, as well as the metallic nature of the MXenes’ conductivity, which results in low noise levels [87].

The detection mechanism of MXenes differs significantly from that of semiconductor metal oxide (MOS) materials, which are considered the basic conventional receptors for the whole class of chemiresistive sensors. For MOS-gas sensors, a universal model can be employed to describe the interaction of the analyte gas with ion-adsorbed oxygen and the formation of an electron-depleted layer (EDL) in the near-surface layer of the sensing material [102]. At present, the question of the mechanism of gas detection by MXenes is debatable. Nevertheless, theoretical and experimental models that allow us to describe the appearance of the chemiresistive signal in the atmosphere of various gases have already been found. In general, two possible reasons for the change in resistance, which is positive in the vast majority of cases, are singled out upon gas intake:

(1)

Transfer of charge from titanium atoms to adsorbed molecules, supporting a reduction in the transport of charge carriers along MXenes and sometimes their quantity;

(2)

Intercalation of gas species into the interlayer space, which causes an enlargement of the distance between the layers and difficulties in the transport of electrons outside the plane of the MXenes.

At the same time, there is a fairly wide range of factors that can influence the response value and kinetic characteristics of MXene materials.

Primarily, the sensitivity of MXenes to various gases is strongly influenced by the functional groups on their surfaces, such as oxygen (–O), hydroxyl (–OH), fluoride (–F), etc., which are defined by a synthesis method. Due to their predisposition to the formation of hydrogen bonds through polar functional groups, MXenes are extremely sensitive to moisture, as noted in Section 3.2 and Section 3.3. Using FT-IR spectroscopy and the study of the wetting angle on the surface of MXenes, Zhang C. et al. found [94] that H₂O molecules from the gas phase can chemically adsorb on the surface and in the interlayer space of Ti₃C₂T_x MXenes. The authors proposed a mechanism where the hydroxyl groups appearing on the surface create an electrostatic field and prevent a charge transfer, which explains the positive (p-type) response with increasing humidity. The nanoscale interlayer space of ca. 1.01 nm in Ti₃C₂T_x and the superhydrophilicity of the surface prevent capillary condensation of water and limit moving hydroxyl groups, which are characteristics for bulk nanomaterials. The 2D accordion-like structure that is characteristic for MXenes just after a selective etching of Al from the MAX-phase without a delamination, the metallic type of conductivity, and the fact that H₂O molecules do not form covalent bonds with the surface of MXenes make them extremely sensitive to humidity and allow for a reproducible and reversible response.

Structural water plays an important role in reducing the overall interaction between the MXene layers and creating additional space for the adsorption and diffusion of gases with affinity to functional groups. This feature is an advantage over other 2D-layered materials, such as, for example, graphene, where strong Van der Waals interactions between 2D sheets and small interlayer distances prevent the intercalation of gas molecules. Koh H-J. et al. demonstrated [89] in situ using XRD that the interlayer space of MXenes contains OH-fragments, such as hydroxyl groups or water molecules, which can be removed when the material is exposed to dry nitrogen. The authors showed that ethanol molecules can chemically bind with Ti₃C₂T_x MXene in its interlayer space. As a result of these interactions, the MXene swells and its interlayer distance enlarges due to the steric effect. The 2D MXene swollen as a result of ethanol adsorption gets a lower number of electrons involved in a charge transfer between the layers, which leads to the growth of electrical resistance with a positive (p-type) chemiresistive response.

MXenes show the best sensitivity when detecting gases capable of forming hydrogen bonds. For instance, Majhi et al. showed [95] the positive (p-type) response in the detection of acetone vapors by Ti₃C₂T_x MXene. They proposed a model of the interaction of acetone molecules with various functional groups on the surface or in the interlayer space of Mxenes. As a result of these interactions, hydrogen bonds are formed. In the course of the reactions on the surface of the MXenes, a significant transfer of the charge carriers takes place, and in the interlayer space—inhibition of conductivity in channels occurs due to the steric effect. In both cases, it is assumed that there is a decrease in the electrical conductivity of the material, which allows a positive p-type response to be recorded.

A study by Kim et al. [87] devoted to Ti₃C₂T_x MXene also proved that the receptor material has the greatest response to gases with a composition and structure capable of forming hydrogen bonds with surface functional groups, such as acetone, ethanol, propanol, and ammonia. At the same time, for acid-forming gases such as NO₂, SO₂, and CO₂, the response is much lower. The authors have demonstrated the ability to detect 50–1000 ppb of acetone, ethanol, and ammonia with a low SNR value. According to DFT calculations, the greatest contribution to the binding of acetone with functional groups on the surface of Ti₃C₂T_x MXene comes from hydroxyl groups (–OH), while oxygen (–O) and fluoride (–F) substituents are less active. For the –OH groups, the minimum binding energy is significantly lower compared with –O and –F, at −0.774 eV versus −0.317 eV and −0.311 eV, respectively. Thus, the hydroxyl terminal groups on the surface of MXenes likely play a key role in the detection of various gases capable of forming a hydrogen bond.

In an earlier work by Lee et al. [85], another mechanism of interaction of NH₃ with the functional groups (–O) and (–OH) on the surface of Ti₃C₂T_x MXene was considered. To explain the positive (p-type) response, the interactions of ammonia molecules with the available oxygen-containing functional groups were proposed. As a result of these reactions, the released electrons recombine with holes, which results in increasing the resistance. Unfortunately, the authors do not experimentally confirm in their work the formation of the products of Reactions (12) and (13), which should be desorbed from the surfaces of MXene, N₂, and NH₂. Therefore, the described mechanism is not obvious at this time.

Wu et al. [79] studied the responses of Ti₃C₂T_x MXene to a wide group of gases. The largest positive (p-type) response was also recorded toward ammonia. The authors performed DFT modeling of the adsorption of various gas molecules on the surface of the MXenes. It was found that the gas molecules do not dissociate upon adsorption. Therefore, the sorption of gases is purely of a physical nature, which contradicts the model proposed by Lee et al. in [85]. It has been shown that the adsorption energy of acetone and ammonia molecules is approximately the same, to be 0.326 and −0.311 eV. However, taking into account the steric factor, because the volume of the ammonia molecule is much smaller than that of the acetone, additional calculations were performed using Bader charge analysis. It was found that the amount of charge transferred to the ammonia molecule was much greater than for all other test molecules. This can explain the pronounced selective sensitivity of Ti₃C₂T_x MXene with respect to NH₃.

The sensory characteristics of Ti₃C₂T_x and Ti₂CT_x in relation to the technique of their synthesis are summarized in Table 1. It can be stated that experimental and theoretical works have shown that modification of surface functional groups can most significantly affect not only the selectivity but also the value of the response upon the adsorption of gaseous analytes. In some studies, increased response to various gases in the case of non-delaminated, accordion-like MXenes due to their high specific surface area has been noted.

Table 1. Sensing characteristics of coatings based on individual Ti₃C₂T_x and Ti₂CT_x MXenes.

Analyte Gas	Concentration	Operating Conditions	Response (ΔR·100%/R0), %	Detection Limit	Synthesis Features	Ref.
Ti3C2Tx
Ammonia	500 ppm	RT	+6.13	10 ppm	NaF + HCl (18–19 wt.%), 60 °C, 48 h, delamination: DMSO, RT, ultrasound for 6 h in nitrogen flow	[79]
Ethanol			+1.5	-
Methane			+0.5	-
NO			+0.4	-
H2O			+0.37	-
Acetone			+0.3	-
Methanol			+0.2	-
H2S			+0.16	-
Ammonia	100 ppm	RT	+0.24	25 ppm	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound for 30 min	[85]
Methanol			+0.143	-
Ethanol			+0.115	-
Acetone			+0.075	-
Ethanol	100 ppm	RT	+1.7	100–1000 ppb	NaF + HCl (18–19 wt.%), 60 °C, 48 h, delamination: DMSO, RT, ultrasound for 6 h in nitrogen flow	[87]
Acetone			+0.97	50–1000 ppb
Propanal			+0.88	-
Ammonia			+0.8	100–1000 ppb
NO2			+0.2	-
SO2			+0.15	-
CO2			+0.12	-
Ammonia	100 ppm	RT	+17	10 ppm	HF (45 wt.%), 60 °C, 24 h	[88]
NO2			+8	-
Ammonia			+28	-	+NaOH (5M), rinsing
NO2			+11	-
Ethanol	0.1%	RT	+0.8	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[89]
Ethanol	0.1%		+10	-	+NaOH (0.3 mM)
Ammonia	100 ppm	RT	+0.9	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[90]
Ethanol			+0.8	-
Ammonia			+3.8	10 ppm	+KOH (0.01M)
Ethanol			+3.4	-
Ammonia			+0.4	-	+KCl (0.01M)
Ethanol			+<0.5	-
Toluene	50 ppm	RT	+19.32	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[91]
Toluene			−79.5	500 ppb	Ti3C2Tx + thiourea, 500 °C, Ar, 3 h
Ethanol	120 ppm	RT	+6.6	-	LiF + HCl(9M), 35 °C, 24 h	[92]
Ethanol			+10.1	-	+(3-chloropropyl)-trimethoxysilane (5%), 24 h
Ethanol			+14	5 ppm	+1H,1H,2H,2H-perfluorooctyltriethoxysilane (5%), 24 h
Ethanol	40 ppm	RT	+5	-	LiF + HCl(9M), 40 °C, 24 h, delamination: ultrasound 30 min	[93]
Octane			+0.3	-
Toluene			+0.1	-
H2O			+2	-
Ethanol			+6.2	20 ppm	+DMSO, trimethylacetic anhydride, 40 °C, 36 h
Octane			+0.9	-
Toluene			+0.8	-
H2O			+1.3	-
Ethanol			+8	-	+1) NaHCO3 (1M), +2) DMSO, trimethylacetic anhydride, 40 °C, 36 h
Octane			+2.5	-
Toluene			+1.4	-
H2O			+0.6	-
H2O	RH = 33%	RT	+7	-	HF (40 wt.%), 40 °C, 36 h	[94]
H2O	RH = 67%		+10	-
H2O	RH = 95%		+15	-
Acetone	100 ppm	23 °C	+99	-	HF (48 wt.%), 60 °C, 15 h	[95]
	250 ppb		+17.3	-
Ti2CTx
Ammonia	10 ppm	RT	+1.1	-	HF (20%), RT, 12 h, delamination: DMFA, ultrasound 12 h	[99]
Ammonia	10 ppm		+0.7	-	LiF + HCl(12 M), 40 °C, 24 h, rinsed with HF (10%)
Methane	200 ppm	RT, visible light irradiation	+16	-	LiF + HCl(9M), 35 °C, 36 h	[100]
Methane	1000 ppm		+68	-
Methane	10,000 ppm		+142	-
NO2	100 ppm	30 °C	+6	-	NaF + HCl(6M), 40 °C, 24 h, delamination: tetramethylammonium hydroxide, ultrasound 30 min	[101]
Ammonia	100 ppm	50 °C	+13	-
Oxygen	1%	30 °C	+8.6 *	-
Oxygen	5%		>276 *	-

* SO₂ = R/R₀ [101].

Table 1. Sensing characteristics of coatings based on individual Ti₃C₂T_x and Ti₂CT_x MXenes.

Analyte Gas	Concentration	Operating Conditions	Response (ΔR·100%/R0), %	Detection Limit	Synthesis Features	Ref.
Ti3C2Tx
Ammonia	500 ppm	RT	+6.13	10 ppm	NaF + HCl (18–19 wt.%), 60 °C, 48 h, delamination: DMSO, RT, ultrasound for 6 h in nitrogen flow	[79]
Ethanol			+1.5	-
Methane			+0.5	-
NO			+0.4	-
H2O			+0.37	-
Acetone			+0.3	-
Methanol			+0.2	-
H2S			+0.16	-
Ammonia	100 ppm	RT	+0.24	25 ppm	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound for 30 min	[85]
Methanol			+0.143	-
Ethanol			+0.115	-
Acetone			+0.075	-
Ethanol	100 ppm	RT	+1.7	100–1000 ppb	NaF + HCl (18–19 wt.%), 60 °C, 48 h, delamination: DMSO, RT, ultrasound for 6 h in nitrogen flow	[87]
Acetone			+0.97	50–1000 ppb
Propanal			+0.88	-
Ammonia			+0.8	100–1000 ppb
NO2			+0.2	-
SO2			+0.15	-
CO2			+0.12	-
Ammonia	100 ppm	RT	+17	10 ppm	HF (45 wt.%), 60 °C, 24 h	[88]
NO2			+8	-
Ammonia			+28	-	+NaOH (5M), rinsing
NO2			+11	-
Ethanol	0.1%	RT	+0.8	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[89]
Ethanol	0.1%		+10	-	+NaOH (0.3 mM)
Ammonia	100 ppm	RT	+0.9	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[90]
Ethanol			+0.8	-
Ammonia			+3.8	10 ppm	+KOH (0.01M)
Ethanol			+3.4	-
Ammonia			+0.4	-	+KCl (0.01M)
Ethanol			+<0.5	-
Toluene	50 ppm	RT	+19.32	-	LiF + HCl(9M), 35 °C, 24 h, delamination: ultrasound 1 h	[91]
Toluene			−79.5	500 ppb	Ti3C2Tx + thiourea, 500 °C, Ar, 3 h
Ethanol	120 ppm	RT	+6.6	-	LiF + HCl(9M), 35 °C, 24 h	[92]
Ethanol			+10.1	-	+(3-chloropropyl)-trimethoxysilane (5%), 24 h
Ethanol			+14	5 ppm	+1H,1H,2H,2H-perfluorooctyltriethoxysilane (5%), 24 h
Ethanol	40 ppm	RT	+5	-	LiF + HCl(9M), 40 °C, 24 h, delamination: ultrasound 30 min	[93]
Octane			+0.3	-
Toluene			+0.1	-
H2O			+2	-
Ethanol			+6.2	20 ppm	+DMSO, trimethylacetic anhydride, 40 °C, 36 h
Octane			+0.9	-
Toluene			+0.8	-
H2O			+1.3	-
Ethanol			+8	-	+1) NaHCO3 (1M), +2) DMSO, trimethylacetic anhydride, 40 °C, 36 h
Octane			+2.5	-
Toluene			+1.4	-
H2O			+0.6	-
H2O	RH = 33%	RT	+7	-	HF (40 wt.%), 40 °C, 36 h	[94]
H2O	RH = 67%		+10	-
H2O	RH = 95%		+15	-
Acetone	100 ppm	23 °C	+99	-	HF (48 wt.%), 60 °C, 15 h	[95]
	250 ppb		+17.3	-
Ti2CTx
Ammonia	10 ppm	RT	+1.1	-	HF (20%), RT, 12 h, delamination: DMFA, ultrasound 12 h	[99]
Ammonia	10 ppm		+0.7	-	LiF + HCl(12 M), 40 °C, 24 h, rinsed with HF (10%)
Methane	200 ppm	RT, visible light irradiation	+16	-	LiF + HCl(9M), 35 °C, 36 h	[100]
Methane	1000 ppm		+68	-
Methane	10,000 ppm		+142	-
NO2	100 ppm	30 °C	+6	-	NaF + HCl(6M), 40 °C, 24 h, delamination: tetramethylammonium hydroxide, ultrasound 30 min	[101]
Ammonia	100 ppm	50 °C	+13	-
Oxygen	1%	30 °C	+8.6 *	-
Oxygen	5%		>276 *	-

* SO₂ = R/R₀ [101].

4. Nanocomposites Based on Ti₃C₂T_x and Ti₂CT_x MXenes

While there are undeniable advantages of MXenes as receptor materials for gas sensors, primarily the possibility to operate at RT and to consume a lowered power, there are also serious negative factors that greatly complicate the application of Ti₃C₂T_x and Ti₂CT_x layers. For instance, when applying coatings, it is inherent to MXenes to agglomerate, to self-assemble into fairly dense stacks of monolayers, which leads to a sharp reduction in the specific surface area and, consequently, to the deterioration of sensitivity. Furthermore, due to the large interaction energy of the Ti₃C₂T_x/Ti₂CT_x surface with molecules of gaseous analytes, as well as the need for diffusing gases into the interlayer space of MXenes, they have overly long sensor response and recovery times, up to tens of minutes, which often exceed these parameters for commercial MOS sensors.

High reactivity, in particular, when interacting with oxygen in a humid environment, typical for nano-dispersed carbides and metal nitrides, leads to constant and fairly fast degradation of MXenes. Therefore, there is a problem for sensors based on Ti₃C₂T_x and Ti₂CT_x to perform steadily within even just a few days. Another experimental problem may be the non-reproducibility of the properties of the receptor materials from batch to batch due to contamination by various synthesis by-products and different storage times of the starting substance prior to applying the receptor layers. Furthermore, the high surface hydrophilicity of MXenes, on which electronegative substituents –F, –OH, and –O are localized, causes their high sorption capacity toward H₂O molecules, which is associated with prospects to employ these layers as humidity sensors, as discussed in Section 3. However, this is a negative factor for the detection of other gaseous analytes, including VOCs, ammonia, and other inorganic gases, because a change in relative humidity in the vast majority of cases causes a significant modulation of the response value.

To overcome these drawbacks, it is proposed to apply approaches known from experience with MOS sensors. In particular, the most powerful one is designing nanocomposites involving semiconducting metal oxides and chalcogenides, nanoparticles of noble metals, carbon materials such as graphene and nanotubes, and polymeric materials. The modification of the surface of MXene sheets by such components creates steric difficulties for the formation of dense aggregates; at least, it dramatically enhances the interlayer distance. In some cases, it leads to the appearance of hierarchically organized mesoporous 3D structures with a much larger specific surface area compared with individual MXenes. In addition, more adsorption centers are induced. The heterojunctions yielded between semiconductor and/or metallic nanoparticles and MXene layers frequently stimulate and advance the chemiresistive response. The combination of the positive characteristics of components of different natures often causes a synergistic effect and a significant change in the gas-detection mechanism of MXenes. Still, we leave out here the finer approaches related to the replacement of various atoms in the MXene crystals by foreign ones [27].

Interestingly, only one publication involving the Ti₂CT_x phase was found in the body of work devoted to the use of nanocomposites with carbide MXenes in chemiresistive gas sensors.

Table 2 provides a short summary of the sensory properties of nanocomposite materials based on Ti₃C₂T_x and Ti₂CT_x in relation to the synthesis method.

4.1. Modification of Ti₃C₂T_x and Ti₂CT_x by Metal Oxide Semiconductors

4.1.1. Modification with n-Type Metal Oxide Semiconductors

Non-stoichiometric metal oxides of n-type conductivity are the most known ones to be employed in current commercial chemiresistors. Primarily, these are TiO₂, SnO₂, and ZnO to consider employing as additives to MXene structures. In this subsection, we look sequentially at how these metal oxide/MXene structures were elaborated.

TiO₂. Modification of Ti₃C₂T_x-based receptor materials with titanium dioxide is most common, primarily because the inevitable oxidation of these MXenes yields nano-dispersed, often amorphous, TiO₂. Already in this case, the improved sensory properties of the resulting composite material are noted. Thus, a sample of Ti₂CT_x/TiO₂ composite formed by incubation of MXenes obtained under the influence of LiF/HCl in an ethanol-water mixture at 4 °C for 16 h [99] was characterized by weak n-type semiconductor properties, close to rutile. The authors showed that the initial receptor material of Ti₂CT_x possessed high sensitivity to NH₃, while the sensor based on the Ti₂CT_x/TiO₂ composite allowed one to detect even 100 ppb of NH₃, as given in Figure 8. Moreover, the response value at 100 ppm ammonia for this sensor exceeds that of individual Ti₂CT_x MXenes obtained by etching with HF acid and LiF/HCl by almost three and two times, respectively [99]. While studying Ti₂CT_x/TiO₂ against a number of other VOCs, including hydrogen sulfide and nitrogen dioxide (10 ppm), it was found that the ratio of responses to ammonia and other gases was at least higher than ca. 3.8, which is characteristic for the second most sensitive gas, formaldehyde. It indicates rather high selectivity to NH₃. Moreover, with enhancing the degree of Ti₂CT_x oxidation by longer time soaking in an ethanol-water mixture, up to 12 h, and increasing the TiO₂ nanoparticle content, the sample response to ammonia has been raised, but after 16 h of oxidation the response dropped due to a destruction of the original MXene sheet into porous and smaller fragments. The authors suggested that the appearance of TiO₂ nanoparticles on Ti₂CT_x increases the number of adsorption centers and adjusts the film resistance in the range of 10⁴–10⁷ Ohm, optimal for reading by electronic circuits.

Choi et al. studied [103] the gas-sensitive properties of the Schottky barrier that appeared at the interface between the metal-like MXene Ti₃C₂T_x and the semiconductor titanium oxide that was generated by holding the MXene in an aqueous dispersion at 80 °C for 4–48 h. This method of Ti₃C₂T_x oxidation was investigated to obtain the most uniform distribution of the oxide because the oxidation degree for the surface layers is higher than for the inner ones when oxidizing the films that have already been deposited. The authors have elucidated the maximum time for such an oxidation to be around 8 h; the longer times resulted in the oxide phase predominating over the MXene one. At this optimum oxidation time of 8 h, TiO₂ nanoclusters have been grown on the edges of Ti₃C₂T_x sheets. Such a structure exhibited a 260-fold growth of the electrical resistance when compared with the pristine MXene, but also an advanced chemiresistive response of the p-type toward a few VOC analytes, 5 ppm of toluene, ethanol, propanol, acetone, and toxic gases, 5 ppm of NO₂ and NH₃, at RT. Such a hybrid Ti₃C₂T_x/TiO₂ structure sample yielded the enhanced chemiresistive response by 13.7 times to NO₂, by 4.7 times to toluene, and ca. three times for the other gases. At the same time, a significant drift of the baseline was observed, which was most likely related to the long recovery time of the sensor. The authors suggested a possible detection mechanism for this heterostructure, taking into account the determined work functions and Fermi levels of their components and constructing the band structure of the nanocomposite (Figure 9), which affects mainly the NO₂ detection mechanism. It was revealed that the response value to the analyte reducing gases does not almost change with varying the time and, accordingly, the degree of oxidation of Ti₃C₂T_x, which has been explained by the formation of the interlayer structure of the Ti₃C₂T_x/TiO₂ composite [103].

Yang and colleagues [104] raised objections to the detection mechanism proposed in [103] based on Schottky barrier formation because there should be observed an ohmic transition at the Ti₃C₂T_x/TiO₂ heterojunction upon a certain work function for the n-type TiO₂ semiconductor that exceeded that for Ti₃C₂T_x. An additional argument by Yang et al. against the signal upon the NO₂ detection amplification mechanism proposed by the authors of [103] is that the response is preserved with increasing resistance, regardless of whether the oxidizing gas or the reducing gas are adsorbed. It is likely that the resolution of this question directly relates to the processes involved in the formation of TiO₂ nanoparticles. For sensor experiments, crumpled spheres containing atomic Ti defects and Ti₃C₂T_x/TiO₂ heterojunctions and samples of similar morphology containing only Ti₃C₂T_x/TiO₂ heterojunctions [104] were fabricated via ultrasonic sputtering pyrolysis technology. The temperature of heat treatment in air, in the 100–200 °C range, regulated the number of atomic defects of Ti and TiO₂. To create a real Schottky heterojunction, graphene oxide was also introduced into the material composition, and its reduction was performed by in situ oxidation of Ti₃C₂T_x, giving the appearance of a more complex composition and structure in the Ti₃C₂T_x/TiO₂/rGO composite. The performed gas sensitivity measurements enabled us to show that there was no increase in the NO₂ response when only TiO₂ appeared on the MXene surface. However, if titanium atoms from Ti₃C₂T_x layers have been involved in building titanium dioxide and, accordingly, atomic deficiencies have been formed, the response to NO₂ goes up significantly, especially at the oxidation temperature of 150 °C, from ca. 11.8% to ca. 26.7%. Realization of the real Schottky heterojunction in the synthesis of Ti₃C₂T_x/TiO₂/rGO nanocomposite resulted in a negative response to NO₂, with positive and relatively low responses to other gases, which is especially remarkable for the sample obtained with a 2:1 mass ratio of Ti₃C₂T_x to GO (Figure 10). The authors also formulated ideas about the detection mechanism of Ti₃C₂T_x/TiO₂/rGO composites involving the Schottky barrier at the rGO/TiO₂ interface.

One of the most frequently used methods of partial oxidation of Ti₃C₂T_x in order to increase the response value due to the appearance of highly dispersed TiO₂ particles is thermal oxidation, primarily in an oxygen-containing atmosphere [29,105,106]. It was shown [105] that a partial oxidation in air of the Ti₃C₂T_x MXene layer at 350 °C, to be conducted in situ on the sensor chip unit, resulted in high responses of ca. 40–180%, with fast response and recovery rates for a number of VOCs down to 2 ppm concentration at a fairly high temperature of 350 °C, which was significantly different from the behavior of the initial Ti₃C₂T_x. The unoxidized material was characterized by responses to VOC vapor intake with an increase in resistance at an operating temperature of 20 °C, while the Ti₃C₂T_x/TiO₂ composite, which contained titanium dioxide in two modifications, anatase and rutile, exhibited a decrease in resistance, which is more typical for n-type MOS-based sensor behavior. The authors showed that the oxidation temperature significantly affects the sensor properties: at temperatures of 100–200 °C, a calcination of Ti₃C₂T_x film has not yielded any response even to 250 ppm of ethanol, while at temperatures of 300–350 °C, the coating resistance has shown a remarkable reduction in the presence of ethanol vapors (Figure 11). At the same time, for the Ti₃C₂T_x/TiO₂ nanocomposite at RT in the presence of alcohols, responses with resistance increases were still observed, which the authors attributed to reactions in internal regions of the multilayer coating with the preserved MXene structure.

Discussing the detection mechanism given in [105], and based on the measured yields for the original Ti₃C₂T_x and the Ti₃C₂T_x/TiO₂ nanocomposite, it is assumed that a potential barrier of ~1 eV is formed between the components. Electrons from TiO₂ nanoparticles enter the metal channel of MXene, while the reverse diffusion is prevented by the potential barrier. Given the particle size and the Debye length of non-stoichiometric TiO₂, the resulting clusters are expected to be completely depleted of electrons. Thus, the reduced conductivity of Ti₃C₂T_x/TiO₂ films is explained by the fact that electron-depleted TiO₂ clusters form potential barriers between the electrically conductive MXene sheets. When analyte molecules are adsorbed, they supply electrons to TiO₂ nanoparticles, which causes a decrease in the potential barrier and a corresponding increase in the conductivity of the material as a whole.

Because the response to all of the gaseous analytes under study occurred with decreasing resistance and was close in magnitude, the device was integrated in a multielectrode matrix equipped with 38 pairs of independent electrodes to improve selectivity. Processing the obtained vector signals from all the sensors on the chip using linear discriminant analysis (LDA) made it possible to reliably recognize even analytes that were close in chemical nature, including low-molecular-weight alcohols (methanol, ethanol, and isopropanol) [105].

The results of the surface oxidation of Ti₃C₂T_x by oxygen plasma at temperatures of 350–550 °C are presented in another work [107]. In order to obtain modified TiO₂ samples, an accordion-like MXene coating was synthesized under the influence of a lithium fluoride solution in concentrated hydrochloric acid and placed into the MPCVD reactor, where plasma was generated in an O₂ atmosphere at a residual pressure of 2 kPa using microwave radiation. At the same time, the substrate temperature was varied from 350 °C to 550 °C.

It was found that the Ti₃C₂T_x phase is preserved in all samples, in addition to the formation of anatase and defects, and that with increasing the activation temperature of the MXene, the number of surface F-groups reduces. It was also noted that a systematic growth of the film resistance from the kOhm to MOhm range is observed as the exposure temperature goes up. By varying the operating temperature in the range of 250–400 °C, the authors found that the sensitivity to ethanol is highest at 325 °C; the sensor being plasma-activated at 500 °C exhibited the greatest response, which was 34.7% higher than that of original Ti₃C₂T_x. This result was explained by increasing the number of –O functional groups and defects in the MXene layers under exposure to O₂ plasma, which, in turn, provided more oxygen adsorption centers. Moreover, the room-temperature response to ethanol for the oxidized sample was found to be almost twice as high as for the original MXene, at ca. 22.5% compared with ca. 13.3%. Still, the response was of the n-type and related to resistance reduction. The chemiresistive response was rather stable within 10 days of continuous testing. When explaining the mechanism of VOC detection in the material, the authors rely on noting the interaction of analyte molecules with ion-adsorbed oxygen, which yields an electron-depleted layer on the surface of the particles of the receptor material, an n-type Ti₃C₂T_x semiconductor. The treatment of MXene with oxygen plasma in an oxygen-rich gas atmosphere causes reactions between hydroxyl groups with the production of H₂O and the removal of fluoride groups from the surface. These processes lead to an increase in the number of functional O-groups and promote the adsorption/desorption of gas molecules to improve sensitivity. In addition, the higher number of O-groups on the surface of MXene layers provides an opportunity to raise the work function.

In order to improve the response of Ti₃C₂T_x MXene to ethanol vapors at RT, the receptor material was also treated in another study [108] with oxygen plasma, which led to the appearance of titanium dioxide on the surface. The authors attributed the observed hundred-fold growth of response and improvement of the sensor kinetic characteristics to the lattice distortion, the formation of a mesoporous structure, and the presence of a Schottky barrier at the Ti₃C₂T_x-TiO₂ interface.

To further improve the moisture sensitivity on the surface of Ti₃C₂T_x layers, TiO₂ nanowires were grown [109]. For this purpose, the Ti₃C₂T_x MXene powder obtained by treating the MAX-phase with concentrated HF after delamination in N-methylpyrrolidone was introduced into a KOH solution and heated at 50 °C. It was found that this resulted in the growth of Ti₃C₂T_x nanowires of 10–100 nm in diameter and 500–1000 µm in length, depending on KOH concentration and duration of exposure, by diminishing fluoride functional groups on the surface of Ti₃C₂T_x nanosheets. The whole structure was shown to have the specific surface area to increase from 10.8 m²/g to 52.6 m²/g. As a result, the sensitivity of the prepared Ti₃C₂T_x/TiO₂ nanocomposite to humidity was high, which was explained via complex impedance spectroscopy and Schottky transition theory.

The application of Ti₃C₂T_x MXene as a source of titanium oxide in the hydrothermal synthesis of Ti₃C₂T_x/TiO₂ nanocomposites as a receptor material for the detection of VOCs has been considered by Kuang et al. [110]. For this purpose, Ti₃C₂T_x MXene powder synthesized by etching the Ti₃AlC₂ with concentrated hydrofluoric acid was dispersed in a 50% ethanol solution and then subjected to hydrothermal treatment at 200 °C for 25 h. An XRD study revealed that such processing resulted in (i) forming an additional anatase phase and (ii) enlarging the inter-planar distance as observed as a shift in the 002 reflex from 9.19° to 9.02° due to the appearance of oxide nanoparticles. The measurement of the specific surface area for the Ti₃C₂T_x/TiO₂ nanocomposite gave a value of 81.9 m²/g, which is an order of magnitude higher than that for the original Ti₃C₂T_x, of 8.6 m²/g. At the same time, the MXene layers were uniformly covered by globular nanoparticles with a diameter of approx. 30–50 nm. These composite samples were found sensitive to VOCs, particularly hexanal, in the 10–40 ppm concentration range, with the responses being 5–6 times higher than those of the original Ti₃C₂T_x (Figure 12). Moreover, both samples of the receptor materials exhibited p-type responses, indicating a decrease in the number of major carriers or blockage of carrier transport under the adsorption of gaseous analytes. It is noted that some baseline drift and high response and recovery times, 293 s and 461 s, respectively, were observed for Ti₃C₂T_x/TiO₂ in signal reproducibility experiments versus 10 ppm of hexanale. The promising application of the synthesized Ti₃C₂T_x/TiO₂ composite for disease detection by exhalation analysis has been demonstrated, since the very low sensitivity of the sensor to CO₂ can reduce its negative effect on the detection of VOC markers. The increase in humidity in the background air from 0 to 80 rel.% forces a reduction in the response to hexanal, 10 ppm as an example, linearly from ca. 2.8% to ca. 0.7%, in accordance with observations about competition between H₂O molecules and hexanal ones at the sensor surface for adsorption centers. The kinetic characteristics of Ti₃C₂T_x/TiO₂ are significantly improved at RH = 20%. The authors cited [110] a combination of three key factors as reasons for the advanced sensitivity of the Ti₃C₂T_x/TiO₂ nanocomposite: (1) the preservation of electrical properties of the multilayer 2D metal framework of MXene provides a fast charge carrier transfer at RT, which underlies the observed high SNR; (2) the advanced specific surface area significantly improves the adsorption of VOC molecules that, in turn, facilitates a greater response value; and (3) the strong electronic interaction between the surfaces of Ti₃C₂T_x and TiO₂ due to heterojunctions can modulate the charge transfer upon test gas exposures.

Thus, the formation of TiO₂ nanoparticles will produce crystal defects in the anion positions, for each of which the two Ti⁴⁺ cations in the crystal lattice nodes are reduced to Ti³⁺ to maintain electrical neutrality, forming an oxygen vacancy with two positive charges [110]. When a hexanal molecule is adsorbed, these vacancies should tend to capture its oxygen atom and the corresponding electrons, i.e., TiO₂ defects act as the major adsorption centers. The migration of electrons from the conduction zone of TiO₂ to the Ti₃C₂T_x MXene leads to a depleted layer at their interface and a Schottky barrier, which is associated with the higher resistivity of the Ti₃C₂T_x/TiO₂ composite; see Figure 13. The absorption of hexanal molecules on the TiO₂ surface containing oxygen vacancies and the transfer of electrons to it will lead to a change in the Fermi levels and a decrease in the thickness of the free carrier-depleted layer. These processes result in a reduction of the concentration of hole carriers in the basic planes of the MXene and force the material resistance to rise. In general, the authors confirm that the continuous heterogeneous Ti₃C₂T_x/TiO₂ interface and reduced charge carrier density are responsible for the high SNR and improved gas response [110].

In work by Liu et al. [111], it was found that the temperature of the hydrothermal synthesis of Ti₃C₂T_x/TiO₂ nanocomposite is a key factor in improving the sensory characteristics. The materials were synthesized by oxidation of the initial Ti₃C₂T_x MXene obtained using concentrated hydrofluoric acid under hydrothermal conditions at temperatures of 140–220 °C for 12 h. It is shown that the number of TiO₂ nanoparticles growing in situ on the surface of multilayer MXene sheets increases as the processing temperature goes up, while their size enlarges until the temperature rises to 180 °C and is further stabilized. At the same time, the response to NO₂, 100 ppm, has gradually increased from ca. 2.2 to ca. 4.5 with the advancement of the hydrothermal oxidation temperature from 140 °C to 180 °C; for comparison, the response of unoxidized Ti₃C₂T_x was equal to ca. 0.3. Still, no significant change in the response value and kinetic parameters occurred with further increases in the hydrothermal treatment temperature [111]. In this regard, the authors chose a temperature of 180 °C as the optimum one for the hydrothermal synthesis of Ti₃C₂T_x/TiO₂ nanocomposite. This material exhibited a response vs. NO₂, 100 ppm, that was ca. 2.1 times greater than that for 100 ppm of SO₂, 6.3 times greater than that for 100 ppm of NH₃, and 6.7 times greater than that for 100 ppm of CO, which indicates quite good selectivity with respect to NO₂.

The higher operating temperatures allowed the sensor to advance the response to 100 ppm of NO₂; at the optimal operating temperature of 175 °C, it was ca. 19.8 when compared with about 4.5 at RT. In addition, the expected decrease in response and recovery times was observed with higher detection temperatures. It is worth noting that in describing the mechanism of NO₂ detection by the Ti₃C₂T_x/TiO₂ nanocomposite, the authors rely on data from the literature [103,112]. In contact with air, the adsorbed oxygen molecules on the surface of the MXenes can change into the anionic form (O₂⁻), which leads to the growth of the Schottky barrier and the formation of an electron-depleted layer at the surface of TiO₂ nanoparticles, as shown in Figure 14. When NO₂ molecules are collected on the surface of Ti₃C₂T_x/TiO₂ composite, they accept electrons according to two possible Reactions (14) and (15) given below, which are accompanied by a deformation of the band structure and an increase in the resistance.

NO_2(g) + ē → NO₂⁻ _(ads)

(14)

2NO_2(g) + O₂⁻ + ē → 2NO₃⁻ _(ads)

(15)

Overall, the research published in [111] shows that the choice of the optimal temperature not only for the hydrothermal oxidation of Ti₃C₂T_x but also for the sensor operating temperature makes it possible to improve the sensor characteristics of the MXene.

In a study by Zhang and coworkers [113], a (001)-oriented TiO₂ was grown on 2D Ti₃C₂T_x carbide using a hydrothermal method to develop a highly sensitive and selective sensor for NH₃ as a food spoilage degree marker. To form predominantly the (001) plane, Ti₃C₂T_x multilayer MXene powder obtained by HF in a 50% solution was mixed with an aqueous NaBF₄ solution and hydrothermally treated at 160 °C for 8–32 h. With increasing the time of heat treatment, the number of TiO₂ nanosheets raised to be located laterally to the MXene planes. At the same time, the 32 h treatment results in the destruction of the layered accordion-like structure of MXene and the formation of disordered aggregates consisting of TiO₂ particles. Exposure to ammonia in a concentration range of 0.05–30 ppm forced the nanocomposites to reduce a resistance that is rather unusual for the pristine Ti₃C₂T_x MXene, similar to the response of n-type semiconductors.

The sample of Ti₃C₂T_x/TiO₂ derived after 16 h of hydrothermal treatment demonstrated the best sensitivity to ammonia; its sensitivity to 10 ppm of NH₃ was seven times higher than that of Ti₃C₂T_x. Its selectivity to ammonia compared with similar amounts of typical gaseous reducing analytes (formaldehyde, trimethylamine, methanol, ethanol, benzene, CO, H₂S) and oxidants (NO₂, SO₂, CO₂) was noted.

An additional UV activation of the sensors has further increased the sensitivity of this Ti₃C₂T_x/TiO₂ nanocomposite to NH₃ by 2–5 times and reduced its detection limit to 5 ppb. The sensor based on the Ti₃C₂T_x/TiO₂ nanocomposite obtained after 12 h of hydrothermal treatment exhibited a stable response at background humidity varying from 25 rel.% to 67 rel.%, which is the most frequently occurring in practice. Implemented into the alarm system designed by the authors, the material was found promising to indicate the beginning of decomposition of pork, fish, and shrimp [113].

The detection of trace amounts of NH₃ at RT was also the goal of work by Zhou et al. [114]. For this purpose, the oxidation of multilayer Ti₃C₂T_x MXene produced using LiF solution in hydrochloric acid was performed simultaneously with its functionalization with nitrogen via a solvothermal treatment at 180 °C of its suspension in a mixture of ethanol-ethylene glycol solvent with dissolved urea. As the duration of treatment increased, anatase nanoparticles were observed to grow as a result of reactions of MXene-derived surface oxygen-containing groups with urea oxygen, which gradually yielded rather large particles of rutile. At the same time, there is a simultaneous alloying of MXene with nitrogen due to urea decomposition reactions with ammonia release (16) and (17), which was recorded during the prolonged treatment for 12 h and 18 h.

$$\mathrm{CO}({\mathrm{NH}}_2{)}_2+({\mathrm{CH}}_2\mathrm{OH}{)}_2\stackrel{heating}{\to }{\mathrm{C}}_3{\mathrm{H}}_4{\mathrm{O}}_3+2{\mathrm{NH}}_3$$

(16)

$$\mathrm{CO}({\mathrm{NH}}_2{)}_2\stackrel{H_{2}O}{\to }\mathrm{HNCO}+{\mathrm{NH}}_3$$

(17)

For the sample prepared at the maximum duration of solvothermal synthesis, for 18 h, an increased sensitivity to low concentrations of ammonia was observed; at 20 °C, a response to 200 ppb of NH₃ was equal to ca. 7.3% [114]. Since the material was highly sensitive to humidity, the authors proposed combining the sensitive element with a waterproof polytetrafluoroethylene membrane.

In order to manufacture sensitive humidity sensors capable of working in a wide concentration range, a method of growing potassium titanate nanofibers on the surface of MXene as a result of the interaction of accordion-like or delaminated Ti₃C₂T_x with a concentrated KOH solution was developed by Wu et al. [115]. Here, the adsorption of H₂O molecules is promoted both by increasing the interlayer distance and by the formation of hydrophilic layered K₂Ti₄O₉ nanowires, 10–50 nm thick, for which the ability to attract water molecules is assumed. It was shown that delamination of the original accordion-like MXene with its subsequent modification by K₂Ti₄O₉ allows for an 8-fold increase in sensitivity when compared with the original MXene. The authors attribute this advancement in sensory properties to three aspects [115]:

(1)

Enlarging the interlayer distance in Ti₃C₂T_x MXene caused by the intercalation of K⁺ cations and contributing to the intercalation of H₂O molecules into it;

(2)

The formation of a layered hydrophilic K₂Ti₄O₉, which is thought to promote the adsorption and intercalation of H₂O molecules;

(3)

The filamentous and porous nanostructure of the material has an increased specific surface area, which provides more efficient contacts between the sensitive material and H₂O molecules.

Thus, partial oxidation of both low-layer and accordion-like Ti₃C₂T_x MXenes allows us a significant increase in sensitivity, mainly to ammonia, moisture, VOCs, and NO₂. The detection mechanism likely depends significantly on the degree of degradation of the original MXene and its transformation into highly dispersed and porous TiO₂.

SnO₂. Another oxide of interest, SnO₂, is currently the most common receptor material used to design MOS sensors [116]. Therefore, quite a number of publications are devoted to the study of the possibilities of combining its positive properties with the advantages of 2D MXene nanomaterials.

For example, in a study published in [117], an electrostatic assembly of the Ti₃C₂T_x/SnO₂ nanocomposite was performed by mixing dispersions of SnO₂ nanoparticles synthesized by the hydrothermal method with a few-layer Ti₃C₂T_x MXene obtained by the MILD process (Table 2). Analysis of the technique as well as XRD and SEM data indicate that SnO₂ is predominant in the composition of the composite. The composite layers demonstrated a greater sensitivity compared with the individual components when detecting 0.5–100 ppm of NH₃, as well as a significant decrease in response time.

Wang and colleagues [118] also studied the effect of modifying SnO₂ microspheres with some amount of multilayer Ti₃C₂T_x MXene synthesized using concentrated hydrofluoric acid on the sensory properties. To obtain SnO₂/Ti₃C₂T_x nanocomposites, 10–50 wt.% of accordion-like MXene particles were introduced into the reaction mixture for hydrothermal production of porous SnO₂ microspheres. It was found that in the normal temperature range of the highest SnO₂ sensitivity, 160–260 °C, the response to 10 ppm of ethanol for the composites was significantly higher than that for pure SnO₂ and even more when compared with that of Ti₃C₂T_x. The material containing 20 wt.% of Ti₃C₂T_x exhibited the highest response at the optimum temperature of 230 °C. It was found that the obtained SnO₂ microspheres grafted onto accordion-like Ti₃C₂T_x particles were almost insensitive to inorganic analytes, NH₃, NO₂, but showed a relatively low response to 10 ppm H₂S and high responses to VOCs, especially to ethanol, to equal 5% upon exposure to 10 ppm of this analyte. Increasing the background humidity from 40 rel.% to 90 rel.% reduces the response to 3.7%. The authors conclude that the main role in the adsorption of gaseous analytes over the nanocomposites is played by microspheres of semiconducting tin dioxide, while the introduced MXene plays the role of a sensitizer. In this work, the authors could clarify the background reasons for the positive effect of Ti₃C₂T_x addition on the gas-sensitive properties of SnO₂ microspheres:

(1)

The presence of a large number of functional groups (–O, –OH, and –F) on the Ti₃C₂T_x surface, which contribute to the nucleation and growth of SnO₂ with a large number of active adsorption centers;

(2)

Schottky barrier formation between the metal surface of MXene and semiconductor SnO₂, which can lead to distortion of its band structure and reduce the thickness of the electron-depleted surface layer in contact with adsorbed molecules of ethanol;

(3)

Acceleration of the mobility of charge carriers due to conductive Ti₃C₂T_x between SnO₂ particles leads to improvements in kinetic characteristics such as response and recovery rates of the sensor.

The hydrothermal method to synthesize SnO₂/Ti₃C₂T_x (SNTM) nanocomposites was also employed in another work [119]. For this purpose, an aqueous solution of urea and cetriltrimethylammonium bromide was mixed with a dispersion containing tin tetrachloride monohydrate and ultrasound-delaminated Ti₃C₂T_x MXene, to be followed by hydrothermal treatment at 140 °C. Both the ratio of the initial reagents and the yellowish color of the output product indicate that the major phase in the synthesized nanocomposite is layered SnO₂, which is oriented perpendicular to the plane of MXene. For the obtained hierarchical structure, the optimum operating temperature versus the detection of 50 ppm of triethylamine was found to be equal to 140 °C, which is significantly lower than that characteristic for pure SnO₂ obtained by a similar technique, which is equal to 260 °C. Moreover, the response value, R₀/R, for the SnO₂/Ti₃C₂T_x nanocomposite was an order of magnitude higher than that for individual tin dioxide, being 33.9 when compared with 3.4 (Figure 15).

It was found that the influence of humidity at RH < 60% for this material is practically absent. The authors justified the detection mechanism here by using the traditional MOS sensor notions of redox surface reactions between adsorbed gas molecules and ion-adsorbed oxygen species on the surface of SnO₂ particles. The reason for the positive effect of the introduction of Ti₃C₂T_x MXene is believed to be the orienting role of its surface functional groups, –F, –O, and –OH, which set the growth direction of SnO₂ nanosheets perpendicular to the Ti₃C₂T_x layers, which prevents their aggregation. A corresponding growth of the specific surface area leads to an increase in the number of active adsorption centers. The removal of F-groups from the surface of MXene under the hydrothermal process promotes the content of other functional groups, –O and –OH, due to the formation of hydrogen bonds that enhance the adsorption energy of the triethylamine molecule and the corresponding number of charge carriers. As the third cause of sensitization, the authors suggest Schottky barrier formation at the interface between SnO₂ and MXene.

In order to design a highly sensitive NO₂ sensor operable at RT, a hydrothermal synthesis method was used in a study published in [120] to synthesize SnO₂/Ti₃C₂T_x nanocomposites. For this purpose, a water-alcohol solution containing multilayer Ti₃C₂T_x MXene, SnCl₄-5H₂O, polyvinylpyrrolidone, and hydrochloric acid was hydrothermally treated. The content of hydrochloric acid was varied from 4% to 11% in order to control the {221} and {110} SnO₂ planes’ growth. During the treatment, the hydrolysis of SnCl₄-5H₂O and the appearance of SnO₂ particles on the surface of accordion-like particles of MXene take place, where particle microstructure depends on the concentration of HCl. Thus, spindle-shaped particles with a predominance of {110} faces are grown at a minimum amount of acid.

Then, there occurs an octahedral formation of dodecahedral crystals with four {110} faces at 7% of HCl and eight {221} faces at 9% of HCl. For the samples with a combination of faces {110} and {221}, the sensing properties were observed at an operating temperature of 25 °C, followed by a pulse heating up to 100 °C for desorption. Under these conditions, the maximum response was recorded toward 10 ppm of NO₂, equal to ca. 1.6%, which is an order of magnitude greater than the one observed in the case of just Ti₃C₂T_x, equal to ca. 0.1%, as shown in Figure 16. At the same time, other test gases—10 ppm of CO, NH₃, SO₂, and H₂S—yielded a lower sensor signal when compared with nitrogen dioxide.

When analyzing the detection mechanism, the authors stated that for {221} faces, all the Sn cations are unsaturated and are characterized by a greater number of broken chemical bonds as compared with {110} faces. This may be the reason for the higher sensitivity of samples synthesized at higher acidity (9% HCl). In addition, a heterojunction can form on the surface of SnO₂ particles (Figure 17) due to the difference in the band structure of these facets, contributing to the efficient electron transfer between the SnO₂/Ti₃C₂T_x composite and the adsorbed NO₂ molecules. The difference in Fermi levels leads to an electron overflow from the Ti₃C₂T_x and {110} SnO₂ facets to the {221} SnO₂ facets, resulting in curved energy bands and electron accumulation at the surface. These electrons can be localized at the surface while being captured by ion-adsorbed species of oxygen. When NO₂ adsorbs, it can localize free electrons in accordance with Reactions (14) and (15), which additionally cause a distortion of energy zones by enhancing both the Schottky barrier and resistance of the SnO₂/Ti₃C₂T_x composite. That is, the formation of Schottky heterojunctions and surface heterojunctions promotes accelerating a charge transfer, thereby improving the sensitivity of the SnO₂/Ti₃C₂T_x composite to NO₂ at low concentrations [120].

Yao et al. [121] found that the hydrothermal treatment of multilayer Ti₃C₂T_x MXene in an aqueous solution containing SnCl₂-2H₂O, urea, and hydrochloric acid resulted in the formation of a Ti₃C₂T_x/SnO nanocomposite. XRD analysis of the material indicates a partial oxidation of SnO to SnO₂, but SEM and TEM images did not allow one to find the particles whose interplanar spacing corresponded to the SnO₂ phase among the disordered array of SnO nanosheets grown on undelaminated MXene. Sn 3D XPS spectra verify that there is a fairly large component from Sn(II), in addition to the Sn(IV) state, which was absent for individual SnO_x synthesized under the given conditions. This allows the authors to conclude that the presence of Ti₃C₂T_x inhibits the oxidation of SnO to SnO₂, which occurs very easily in air. The study of exposing the material at RT to 200 ppm of various gases demonstrated its high p-type sensitivity to ammonia with an R₀/R response equal to ca. 7.8. For the other studied VOCs, except for ethanol, the resistance has decreased under the n-type response but is still rather low, not exceeding 1.6. It indicates a decent selectivity toward NH₃ vapors. Nevertheless, an aging of the material for more than 7 days was observed, which the authors attributed to SnO oxidation as well as a high sensitivity to background humidity.

The authors of [121] attributed the superior gas sensitivity of the Ti₃C₂T_x/SnO composite material over that of the individual components to a number of factors. First, the –O, –OH, and –F functional groups existing on the surface of MXene layers promote the growth of ultrathin SnO sheets, which is promising for accelerating electron diffusion. The higher specific surface area contributes to more available adsorption centers for oxygen and analyte molecules from the surrounding environment. Accounting for the observed chemiresistive behavior of Ti₃C₂T_x/SnO composites, the authors concluded that here SnO exhibits n-type semiconductor properties while Ti₃C₂T_x with metallic conductivity plays the role of a conductive channel, which reduces the resistance compared with pure SnO. In addition, MXene enforces the electron transfer rate and consequently improves the kinetic characteristics of the sensor. The p-n junctions that appeared between Ti₃C₂T_x and SnO enabled higher concentrations of charge carriers and the suppression of their recombination.

The synthesis of the nanocomposite, similar to that reported in [121], was performed by Wang et al. [112]. However, these authors emphasized observing two oxide phases, SnO and SnO₂, as a result of Sn²⁺ oxidation by residual oxygen in the autoclave. The resulting SnO-SnO₂/Ti₃C₂T_x nanocomposite displayed high sensitivity to acetone in a wide range of concentrations, 10–100 ppm. The R/R₀ response to 100 ppm of acetone was ca. 12.1 at RT, which is almost eleven and four times higher than one of the original Ti₃C₂T_x MXene and SnO-SnO₂ oxide composites, respectively. At the same time, the resistance gradually enhanced under the gas exposures for the SnO-SnO₂ and SnO-SnO₂/Ti₃C₂T_x composites, while an unusual drop in resistance was observed in the original multilayer MXene obtained by etching the MAX-phase with concentrated hydrofluoric acid. The synthesized composite material also showed good responses to toluene, ethanol, and methanol, the value of which, however, is significantly lower than that to acetone and is in the range of 2.9–4.3. The long-term stability of the SnO-SnO₂/Ti₃C₂T_x-based sensor was evidenced by recording the response value to acetone for 35 days. The authors considered the acetone detection by the obtained nanocomposite via the existence of p-n-heterojunctions in the SnO-SnO₂ oxide particles formed on the MXene surface, which caused the increased concentration of charge carriers and the suppression of charge carrier recombination. This system seems to have a layered structure: SnO primarily appears on the surface of Ti₃C₂T_x from the solution, which is converted to SnO₂ at the outer surface of the SnO/Ti₃C₂T_x composite via reacting with oxygen to be a n-type semiconductor. Therefore, a charge transfer occurs from SnO₂ and Ti₃C₂T_x with metallic conductivity to the p-semiconducting SnO, which leads to a distortion of the band structure (Figure 18).

When oxygen molecules are adsorbed from air to induce primarily O₂⁻ ions on the surface, a depleted layer is generated at the p-n junction boundary while a layer of hole accumulation (HALs) is formed at the SnO/Ti₃C₂T_x interface. This increases the mobility of charge carriers and reduces the Schottky barrier height, i.e., the resistance at RT has a relatively low value. Acetone reacts with ion-adsorbed oxygen following Reaction (18), releasing electrons that recombine with holes and cause a higher resistance.

CH₃COCH₃ + 4O₂⁻ → 3CO₂ + 3H₂O + 4ē

(18)

The authors of the study [112] highlighted the main contribution of Ti₃C₂T_x MXene to the yielding of the composite morphology in the form of a layered framework with a high specific surface area and, consequently, with a large number of active adsorption centers. In addition, due to the high electrical conductivity of Ti₃C₂T_x, an electron transfer under the gas detection is accelerated, which leads to a reduction in the sensor’s response and recovery time.

To summarize, it should be emphasized that the modification of Ti₃C₂T_x MXene with n-type SnO₂ semiconductors has not been considered in this section. Rather, the talk is about SnO₂ sensitization by 2D MXene with metallic conductivity. In this regard, it was possible to reduce in some cases the operating temperature of the SnO₂–based sensor down to RTRT. In addition, the increased sensitivity of the resulting nanocomposites is noted, primarily due to the appearance of Schottky junctions and the significantly increased specific surface area because the MXene plays a role as a framework template for the growth of oriented SnO₂ nanoparticles. Nevertheless, a number of questions arise following the reported data. For instance, the synthesis of a number of SnO₂/Ti₃C₂T_x nanocomposites is carried out in the framework of a hydrothermal approach, which promotes a significant degradation of MXene structures with growing TiO₂ nanoclusters, whose appearance is frequently documented by XRD via corresponding reflections. Therefore, the reasoning about the detection mechanisms, at least for a number of experimental results, should most likely be complicated by adding the TiO₂ phase, an n-type semiconductor, into consideration.

ZnO. Zinc oxide is the second most common semiconductor oxide used for the detection of gaseous analytes in chemiresistive sensors. Recently, the number of publications devoted to employing ZnO in gas sensors has almost equaled that on SnO₂ [116]. Therefore, many scientific groups attempted to combine MXenes and ZnO [122,123,124,125,126].

In order to create a flexible NO₂ sensor allowing a chemiresistive response at RT, the Ti₃C₂T_x/ZnO nanocomposite was considered in the study by Yang et al. [122]. In addition to the polyvinylpyrrolidone, zinc acetate was also introduced into an initial solution for ultrasonic pyrolysis in the process of obtaining the crumpled MXene spheres. Thermal treatment of the aerogel in a N₂ flow at 800 °C resulted in the formation of a nanocomposite with the morphology of a 3D crumpled MXene sphere, whose surface, especially on the protruding ribs, was covered with ZnO nanoparticles of ca. 10 nm diameter that were evenly distributed. Among a large number of analyte gases tested on the Ti₃C₂T_x/ZnO sample, there was a sharp increase in the response value versus NO₂ compared with pristine Ti₃C₂T_x; the response to 100 ppm of NO₂ at RT and humid air was ca. 41.9% in the case of the Ti₃C₂T_x/ZnO sample compared with ca. 27.3% in the case of Ti₃C₂T_x, respectively. A significant improvement in the kinetic characteristics of the obtained nanocomposite was also noted, including a relatively fast recovery of resistance to the baseline, as shown in Figure 19. The authors showed that the sensor signal is stable under repeated bending, including an angle of up to 120°, which is important for portable electronics devices on flexible substrates. It was shown that the signal magnitude decreased by 31% even after performing 1000 bends at 90°, together with some loss of film conductivity. There was a tendency for the response value toward 100 ppm of NO₂ to advance from ca. 28.9% to ca. 51.4% when the humidity in the background air increased from 20 rel.% to 90 rel.%.

An unusual reduction of resistance was observed [122] in Ti₃C₂T_x/ZnO upon NO₂ exposure (in contrast to ammonia, which led to a resistance enhancement). This behavior differed from that of crumpled, pristine Ti₃C₂T_x spheres, which exhibited an increase in resistance versus both analytes. The authors considered that the relative content of oxidized titanium, TiO₂, as verified by XPS, is higher in the Ti₃C₂T_x/ZnO nanocomposite. As a result, the metal conductivity of the MXene is converted to a p-type semiconducting one. Thus, these composite structures contain heterogeneous p-n junctions with ZnO nanoparticles synthesized on the Ti₃C₂T_x surface, yielding an electron-depleted layer at this interface. The adsorption of NO₂ molecules leads to the expansion of this layer, which improves charge transfer efficiency. The adsorbed water molecules, as stated by the authors, also lead to the effect of n-alloying and additionally promote the sensor response to NO₂, with which they are able to form hydrogen bonds.

The work by Fan et al. [123] is also devoted to obtaining highly sensitive nitrogen dioxide sensors using the Ti₃C₂T_x/ZnO composite. To obtain the receptor material, a Ti₃C₂T_x dispersion synthesized by etching aluminum from the MAX-phase structure was mixed with hydrothermally synthesized porous layered ZnO nano-powder calcined at 450 °C. After ultrasonic treatment of the combined suspension, a receptor layer was applied onto an electrode structure to serve as a sensor. The hybrid material obtained in this manner showed high sensitivity toward NO₂, a response to 20 ppm of ca. 367.6%, and some selectivity (Figure 20). The authors believe that the porous morphology of the layers and the oxygen vacancies in ZnO contribute to the favorable adsorption of gaseous NO₂, as does the presence of surface functional groups (–F, –OH, –O) on the surface of the Ti₃C₂T_x MXene layers. Furthermore, the authors found the UV irradiation useful to substantially reduce the sensor recovery time, down to 22 s.

This effect is attributed to the photogeneration of charge carriers in ZnO under UV (Figure 21), which release chemically adsorbed NO₂ molecules according to Reactions (19)–(21).

NO_2(gas) + ē → NO₂⁻_(ads)

(19)

Hν → h⁺(hν) + ē(hν)

(20)

NO₂⁻_(ads) + h⁺(hν) → NO_2(gas)

(21)

DFT calculations indicate that the main adsorption centers belong to ZnO nanosheets, while Ti₃C₂T_x acts as a conductive channel to accelerate charge transfer [123].

A more complex multistep synthesis was employed to obtain a mesoporous Ti₃C₂T_x/ZnO composite as part of an efficient NO₂ optoelectronic sensor [124], where ZnO nanorods were grown on the surface of MXene from seed ZnO nanoparticles. The oxide seed nanoparticles were deposited on the surface of delaminated Ti₃C₂T_x from a methanol–zinc acetate solution by interaction with NaOH. The specific surface area of the material was higher than 145 m²/g. The resulting composite powder was further dispersed in an aqueous NaOH solution together with the previously synthesized ε-Zn(OH)₂ and held at 80 °C for the aging and growth of ZnO nanorods. The EDS analysis data revealed that the ZnO content was ca. 63.5 wt.%. The comparison of responses of the composite to 50 ppb of NO₂ in the dark and under UV irradiation showed that UV forces the sensor signal to grow from ca. 6.5% up to ca. 81% with low response, 17 s, and recovery, 24 s, times. Such a high sensitivity gives an option to measure ultra-low NO₂ concentrations down to a detection limit of 0.2 ppm (Figure 22). The response to NO₂ was quite selective when compared with other gaseous pollutants under test such as SO₂, NH₃, H₂S, and CO. In particular, the response to 50 ppb of NH₃, second in value to that for NO₂, was almost four times lower. At a higher relative humidity of RH = 80%, the response decreased to approx. 50%, which the authors attribute to the competing adsorption of water molecules.

The effect of the component ratios in the ZnO/Ti₃C₂T_x nanocomposite on the gas-sensing properties was investigated by Zhu et al. [125]. To produce the receptor materials in two types, containing multilayer and few-layer Ti₃C₂T_x, the dispersion of the corresponding MXene in zinc acetate solution and CTAB (cetyltrimethylammonium bromide) was hydrothermally treated at a relatively low temperature of 120 °C. The content of Ti₃C₂T_x in the different samples varied from 1 wt.% to 3 wt.%.

It was found that the composites containing delaminated Ti₃C₂T_x MXene exhibit greater sensitivity to acetone, with an R₀/R response to a 100 ppm concentration equal to 14.1 at 320 °C of operating temperature for the 2 wt.% Ti₃C₂T_x when compared with samples based on multilayer Ti₃C₂T_x.

The study by Liu et al. [126] was also dedicated to the effect of doping zinc oxide with small amounts of delaminated Ti₃C₂T_x MXene on its sensitivity to nitrogen dioxide. To obtain the ZnO/Ti₃C₂T_x composite in the form of spherical porous particles, zinc chloride, sodium acetate, and sodium citrate were added to the dispersion of MXene in ethylene glycol, followed by solvothermal synthesis at 200 °C. For the composite material containing 2% of Ti₃C₂T_x with the maximum specific surface area, the dependence of the response value toward 10 ppm of NO₂ on the operating temperature in the range of 150–180 °C was studied. It was found that the response at 160 °C is 1.4–2.9 times higher as compared with that at other temperatures. Measurement of response to various gases (CO, H₂S, NO₂, C₂H₅OH, NH₃, CH₂OH) at 10 ppm concentration showed that the ZnO/Ti₃C₂T_x material obtained by the method is more selective to NO₂. The authors explain the advancement of the sensory characteristics of ZnO in terms of reducing the operating temperature from 220 °C to 160 °C and increasing the response when it is modified with some amount of few-layer Ti₃C₂T_x due to the higher specific surface area and appearance of mesoporous structure, which facilitates more active sorption centers as compared with pristine ZnO. Another factor is the high electrical conductivity of metallic Ti₃C₂T_x layers, which enables fast transport of charge carriers. Because the porous ZnO spheres are anchored on both sides of the Ti₃C₂T_x layers (Figure 23), they can act as channels for free carriers and effectively reduce the potential barriers at the grain boundary between the ZnO spheres.

In order to modify nanoporous cubic tin-zinc oxide (ZnSnO₃) aggregates with delaminated Ti₃C₂T_x sheets, Sima and colleagues [127] performed a hydrothermal treatment of a pre-synthesized suspension of MXene and ZnSnO₃, with its surface modified with the cationic surfactant CTAB to electrostatically assemble the individual components. The measured CVC curves demonstrated that the ZnSnO₃/Ti₃C₂T_x hybrid material has a higher electrical conductivity compared with the initial ZnSnO₃. Furthermore, the high sensitivity of the obtained composite to 100 ppm of formaldehyde with a response equal to 194.7% at RT and short response and recovery times of 6 s and 5 s, respectively, was found out. Measuring the responses to 100 ppm of other toxic and practically significant VOCs (ammonia, alcohol, benzene, acetone, and trimethylamine) revealed that the sensitivity to formaldehyde was at least 2.1 times higher than that to each of the above test analytes (Figure 24). The authors discussed this as advancing to the higher enthalpy of formaldehyde adsorption. The detecting mechanism of ZnSnO₃-based semiconductor materials is based on the well-known redox interaction between formaldehyde molecules and chemisorbed oxygen (Reactions (22) and (23)), resulting in a free electron release and thinning of the electron-depleted semiconductor layer [127].

HCHO_(gas) → HCHO_(ads)

(22)

HCHO_(ads) + O₂⁻_(ads) → CO₂+H₂O+ē

(23)

The authors see the contribution of impurities in delaminated MXene to improving the sensory properties of the composite material via:

(1)

Preventing the aggregation of ZnSnO₃ cubes with Ti₃C₂T_x sheets that advance the specific surface area of the material;

(2)

Promoting the adsorption of organic molecules due to the presence of polar functional groups (–F, –OH, –O) on the surface of Ti₃C₂T_x;

(3)

Increasing the free carrier mobility due to the enhanced electrical conductivity of Ti₃C₂T_x;

(4)

Appearing as a heterojunction at the interface between ZnSnO₃ and Ti₃C₂T_x.

WO₃. Another oxide, tungsten(VI), draws significant attention as an n-type semiconductor characterized by high sensitivity to ammonia. In order to achieve a response to NH₃ at RT, Guo et al. [128] studied the potential of Ti₃C₂T_x-WO₃ composites by varying the WO₃ content from 17 wt.% to 67 wt.%. The authors mixed ultrasonically treated Ti₃C₂T_x dispersions, obtained with HF, and hydrothermally synthesized WO₃ to obtain the composite samples. Using the Ti₃C₂T_x-50 wt.% WO₃ sample as an example, XPS spectroscopy data revealed that this method of nanocomposite fabrication does enhance the Ti–X bond due to the oxygen atoms transferring from the WO₃ nanoparticles to the MXene surface. The research on the Ti₃C₂T_x/WO₃ nanocomposites exposed to 1 ppm NH₃ at RT indicated that an optimal phase ratio exists, which refers to the Ti₃C₂T_x-50 wt.%WO₃ composition. In this case, the observed response was ca. 22.3%, compared with the ca. 1.5% characteristic for pristine Ti₃C₂T_x [128]. At the same time, all samples except pure WO₃ demonstrated a p-type response related to enhancing resistance upon analyte introduction, indicating that the chemiresistive response of the nanocomposites is determined by the Ti₃C₂T_x MXene. The reproducibility study on the example of four detection cycles for samples of Ti₃C₂T_x-50 wt.% WO₃ and Ti₃C₂T_x confirmed that the composite-based chemiresistor has better repeatability while the SNR matches that of a pure MXene with its metallic conductivity. The kinetic properties of the nanocomposite, alternatively, improve insignificantly compared with Ti₃C₂T_x. The responses of the nanocomposite to other analytes (H₂S, NO, CH₃CH₂OH, CH₃CHO, and CH₃COCH₃) were ca. 2.4–17.3 times lower than those to NH₃. The humidity suppressed the signal to NH₃ and deteriorated the sensor baseline, which the authors attributed to the antagonism between H₂O and NH₃ molecules for adsorption centers [128]. The suggested mechanism of the p-type response to ammonia primarily relates to reducing carrier transport due to the adsorption of the gas molecules as surface defects. The increased sensitivity for Ti₃C₂T_x-WO₃ nanocomposites, similar to that of other n-type metal semiconducting oxides, is explained by the higher porosity and number of active centers provided by WO₃ as well as by the heterojunction existing at the boundary between the metal Ti₃C₂T_x and the n-type semiconductor WO₃.

Still, it is worth noting that the negative effect of humidity is observed in the study of [128] in all the sensors based on the pristine metal oxide, MXene, and Ti₃C₂T_x/WO₃ nanocomposites. In contrast, the work by Gasso et al. [129] dealing with the design of a NO₂ sensor showed preserving the response value over a wide range of humidity in the background air (0–99 rel.%). In this work, the hydrothermal synthesis of WO₃ nanorods in the presence of dispersion of Ti₃C₂T_x, to be synthesized with concentrated HF, was performed to obtain a Ti₃C₂T_x/WO₃ nanocomposite. It was demonstrated that MXene plays in this case the role of an anchoring platform for WO₃ nanorods, which limits the aggregation of the latter and increases the specific surface area [129]. The chemosensory properties of the nanocomposite tested at RT in a dry atmosphere toward a number of gases of oxidizing nature (NO₂, I₂, Br₂, Cl₂) and reducing nature (ammonia, acetone, ethanol, isopropanol) showed that it has the highest sensitivity to NO₂ at a 30–1000 ppb concentration range. However, it is noted that the response to this analyte, at 0.2 ppm, goes from ca. 197% to ca. 145% while the humidity increases from 0 rel.% to 99 rel.%. Aiming to improve the stability over time and reduce the sensitivity of the material to humidity, the MXene dispersion was pre-treated with sodium L-ascorbate under the Ti₃C₂T_x/WO₃ nanocomposite synthesis [129], because there is evidence in the literature [126,130] that this approach is effective due to the screening of titanium atoms in Ti₃C₂T_x sheets from oxygen and water attack by C₆H₇O₆ anions. These modified Ti₃C₂T_x/WO₃ composites exhibited a response to 200 ppb of NO₂, equal to ca. 184% of the total range of RH = 0–99 (Figure 25).

Furthermore, Gasso’s group [131] designed a sensor for even lower NO₂ concentrations, down to 15 ppb of concentration, considering the WO₃/Ti₃C₂T_x nanocomposite synthesized by a stepwise hydrothermal method. In particular, the mixture of thiourea and nitric acid with Na₂WO₄ solution containing 10–40 wt.% of Ti₃C₂T_x, up to pH = 3, was stirred for 1.5 h prior to hydrothermal treating. The gas-sensing properties of the sensor placed over a flexible polyamide substrate were studied only for the WO₃-20 wt.% Ti₃C₂T_x sample. This composition exhibited high sensitivity and selectivity versus NO₂, as it was documented in the previous work on WO₃/Ti₃C₂T_x nanocomposites [129] obtained by another technique. The response to 200 ppb of NO₂ at RT was found to be 78%, while for the same concentration of ammonia, it was 20%, SO₂—13%, ethanol—11%, and for other gases under test, it did not exceed 10% [131].

At the same time, the 20 wt% Ti₃C₂T_x-decorated tungsten oxide showed a reduced sensor response and recovery time compared with both pristine Ti₃C₂T_x MXene and pure WO₃. The WO₃/Ti₃C₂T_x material maintained sensor properties for 30 days. However, when the humidity in the background air varied from 0 rel.% to 99 rel.%, the response diminished from 78% to 56%, accompanying a rise in the sample resistance both in humid air and upon analyte injection [131].

A solvothermal growth of non-stoichiometric tungsten oxide (W₁₈O₄₉) nanorods on the surface of Ti₃C₂T_x MXene sheets was reported by Sun and colleagues [132] to fabricate a sensor for low concentrations of acetone. The choice of W₁₈O₄₉ as the major phase of the nanocomposite was due to the presence of a large number of oxygen vacancies, which can serve as additional active adsorption centers for the target gases. To synthesize the W₁₈O₄₉/(1–2.5 wt.%)Ti₃C₂T_x composite material in an acetylacetone-modified tungsten ethoxide solution, multilayer MXene powder was dispersed by ultrasound prior to placing it in an autoclave for solvothermal processing. The gas-sensing properties of the material were analyzed using VOCs and ammonia, at 20 ppm concentration as an example, at 200–400 °C temperatures. The measurements ensured an improved selectivity for acetone detection with W₁₈O₄₉/Ti₃C₂T_x structures characterized by a resistance reduction, or n-type response, which is characteristic for n-semiconducting oxides, including WO_x≤3. It was shown that the operating temperature, optimal in terms of the response value and the response/recovery rates, is 300–350 °C, again similar to those observed in oxides. Moreover, the material decorated with 2 wt.% Ti₃C₂T_x stands out with improved chemiresistive properties: the response, R₀/R, to 200 ppm of acetone at 300 °C was ca. 12.1, while in the case of the pristine W₁₈O₄₉ it was ca. 3.8. The reason for this phenomenon, according to the authors, is (i) advancing the specific surface area when W₁₈O₄₉ is placed on the Ti₃C₂T_x sheets surface, (ii) removing the fluoride surface groups in the MXene under the solvothermal synthesis and their replacement with OH– and O–groups, as well as (iii) an inter-phase interaction between the components having semiconductor and metallic conductivity characteristics, or the appearance of heterojunctions sensitive to environment.

In₂O_3. Another appropriate oxide that is frequently employed to develop chemiresistors is non-stoichiometric In₂O₃. Therefore, Liu et al. [133] tried to develop In₂O₃/Ti₃C₂T_x composite materials fabricated from the interaction of hydrothermally pre-synthesized In₂O₃ nanocubes and Ti₃C₂T_x MXene via an additional hydrothermal treatment. In order to anchor and evenly distribute these porous cubes on the MXene sheet surface for producing multiple heterostructures, the authors modified the indium oxide surface with a cationic surfactant, (3-aminopropyl)triethoxysilane (Figure 26).

While exposing such a sensor to VOCs (acetone, xylene, triethylamine, methanol, trimethylamine, and toluene) of 5 ppm concentration at RT, it was found that the nanocomposite exhibits an enhanced sensitivity to methanol with the response of the n-type, which can be explained by the minor Ti₃C₂T_x addition. Still, the kinetic properties were found promising for practice; the response/recovery times were 6.5 s and 3.5 s, respectively, although some baseline drift has been observed too [133].

A self-assembly of metal organic frameworks (MOFs) was employed in study [134] to prepare an In₂O₃/Ti₃C₂T_x composite with indium oxide having a highly developed cage-like porous structure. For this purpose, an aqueous dispersion of Ti₃C₂T_x MXenes was introduced into a solution containing indium nitrate and terephthalic acid in dimethylformamide, followed by heat treatment at 120 °C (Figure 27). As a result, MOFs of the MIL-68 (In) type were formed, whose crystals have a hexagonal microtube shape with a small amount of MXene sheets distributed in their volume. To remove the organic moieties, the precursors were calcined in air at 500 °C to design the In₂O₃/Ti₃C₂T_x composite with a giant specific surface area of 447 m²/g. It was established [134] that the detection of NH₃ at a 5 ppm concentration for such a sensor follows the n-type response characterized by a resistance reduction. Increasing the operating temperature from 25 °C to 100 °C facilitates some raising of the response from ca. 60.6% to ca. 67.5% but has a negative effect on the response/recovery times. Under heating up to 200 °C, there is a sharp deterioration in kinetics accompanying the reduction of the response to ca. 41.2%. The In₂O₃/Ti₃C₂T_x porous sample also yielded a good response to methanol (equal to ca. 20.2%) and ethanol (equal to ca. 12.8%) at 5 ppm concentration; the response to other VOCs, CO and NO₂, at the same concentration, did not exceed ca. 9%. It is interesting that the sensor response has not changed significantly in a moderately humid atmosphere, RH ≤ 60%, although the higher humidity level resulted in decreasing both the resistance and the response. Obviously, such an effect relates to filling adsorption centers with H₂O molecules, as discussed in other works.

In order to understand the effect of varying ratios of components in the Ti₃C₂T_x/In₂O₃ nanocomposite on its gas-sensing performance at RT, Zhou and colleagues [135] obtained a series of samples containing In₂O₃ in a range of 15 wt.% to 80 wt.%. For this purpose, In₂O₃ nanopowder synthesized by the solvothermal method in glycerol and isopropyl alcohol to be calcined at 500 °C was introduced into the dispersion of multilayer Ti₃C₂T_x MXene synthesized under the NaF–HCl system.

This reaction system was subjected to ultrasonic processing for 5 h, which additionally led to some delamination of Ti₃C₂T_x, as confirmed by the obvious shift of the (002) reflection to the low-angle region when compared with the original multilayer MXene. Testing the Ti₃C₂T_x/In₂O₃ samples by introducing 30 ppm of NH₃ allowed authors to select the composition containing 50 wt.% In₂O₃ [135] because its response value was the highest, equal to ca. 63.8% at RH = 40%, and was almost two times greater than that for the second gas-sensitive sample of Ti₃C₂T_x–30 wt.% In₂O₃. The introduction of even minor amounts of In₂O₃, such as 15 wt.%, promoted the sensitivity towards ammonia as compared with pristine In₂O₃ and MXene by 2.7 and 12.5 times, respectively. At the same time, the authors note that the humidity level caused an almost linear reduction of the response value down to ca. 36.5% observed at RH = 85% in the case of the Ti₃C₂T_x–50 wt.% In₂O₃ sample. The measured responses to 30 ppm of other gases (methanol, nitrogen monoxide, hydrogen sulfide, ethanol, and acetone) were in the range of 3.2–8.9%, which indicates an elevated selectivity of this sensor specifically versus ammonia, at least in the 2–100 ppm concentration range. When discussing the mechanism of ammonia detection by the Ti₃C₂T_x/In₂O₃ nanocomposite, the authors considered that the NH₃ injection resulted in higher resistance, while the individual In₂O₃, as a typical n-type semiconductor, should exhibit in such a case an opposite change in resistance. Therefore, it is concluded that the sensitivity of Ti₃C₂T_x/In₂O₃-based sensors is mainly dependent on Ti₃C₂T_x. Summarizing, the authors attributed the improved sensing performance of the composite material when compared with the individual components to two aspects:

• Owing to the deposition of nanoparticles on the surface and in the interlayers of multilayer MXenes, a larger specific surface area appears to be 39.06 m²/g compared with the 6.95 m²/g characterizing one for pristine Ti₃C₂T_x. This promotes the adsorption and diffusion of gas molecules into the volume of the receptor material.
• When mixing semiconducting In₂O₃ and Ti₃C₂T_x, which is considered a metallic-like crystal because of its high conductivity and charge carrier mobility, a Schottky barrier and an electron-depleted layer are formed at the interface. When exposed to ammonia, its molecules are predominantly adsorbed on the In₂O₃ surface, where ion-adsorbed forms of oxygen from air (O₂⁻) are present due to their higher adsorption energy. These species interact according to the Reaction Scheme (24):

4NH₃ + O₂⁻ → 4NO + 6H₂O + 5ē

(24)

As a result, a large number of free electrons are released to be recombined with holes on the Ti₃C₂T_x surface, which increases the integral resistivity of the Ti₃C₂T_x–50 wt.% In₂O₃ material.

The next metal oxide worth testing in combination with MXene to develop a sensor is Fe₂O₃. The method employed earlier to compose In₂O₃/Ti₃C₂T_x [133] was applied in the study described in Ref. [136] to compose composites consisting of α-Fe₂O₃ nanocubes anchored to the surface of the multilayer Ti₃C₂T_x MXene as a result of electrostatic self-assembly (Figure 28). When examining the gas-sensing properties of such α-Fe₂O₃/Ti₃C₂T_x composite at RT and a relative humidity of approx. 22.3%, taking 5 ppm of some analytes (toluene, methanol, formaldehyde, acetone, xylene, ethanol, SO₂, and NO₂), the authors revealed its higher sensitivity to acetone with a ca. 16.6% response, while the second largest response was to xylene, at ca. 3.5%. The reducing character of the resistance change when detecting acetone vapors indicates that α-Fe₂O₃ as an n-type semiconductor plays the most significant role in this process. High rates of response/recovery were also observed.

Fe₂O₃. Liu et al. [137] reported on the hydrothermal synthesis of α-Fe₂O₃/Ti₃C₂T_x composites, where the rose-like α-Fe₂O₃ particles were grown on the surface of MXene sheets. In this case, isopropanol and a dispersion of delaminated Ti₃C₂T_x were added to a solution of iron(III) chloride and urea in ethylene glycol, followed by a heat treatment at 120 °C. Because such mild conditions do not facilitate the appearance of iron oxide but rather α-FeOOH metahydroxide, the produced and purified product was calcined in an Ar atmosphere at 500 °C, which excluded a significant oxidation of the 2D nanomaterial. Measuring the response of the material to 5 ppm of several VOCs (acetone, methanol, trimethylamine, ethanol, toluene, NH₃, SO₂, and NO₂) at RT and 22 rel.% of humidity revealed an increased sensitivity to ammonia, with the response equal to ca. 18.3%. Fast response and recovery, as well as the fact that the resistance decreased rather than increased upon the analyte intake, also indicate the predominance of α-Fe₂O₃ in the detection mechanism. The role of Ti₃C₂T_x is limited to forcing an enlarged nanocomposite specific surface area due to spatially-restricted aggregation of FeOOH nanoparticles, as well as enhanced charge carrier mobility and the formation of n-p-heterojunction at the interphase boundaries [137].

A similar approach to the synthesis of porous semiconductor oxide on Ti₃C₂T_x layer surfaces via intermediate MOF self-assembly, which was described for the In₂O₃/Ti₃C₂T_x structure in [134], was also employed in another research [138] to synthesize the Fe₂O₃/Ti₃C₂T_x composite, where the iron oxide was present in two phases, α- and γ-Fe₂O₃. The MXene content in the output compositions ranged from 2 to 6 wt.%. The gas-sensing properties of the material were measured by applying the receptor layer to the ceramic tube with a heater confined in the inner space. For the composite material containing 4 wt.% of Ti₃C₂T_x, the increase in the specific surface area fixed by the BET method according to XPS data was accompanied by a significant growth in the relative percentage of oxygen vacancies, equal to ca. 19.8%, as compared with pristine α-/γ-Fe₂O₃, of ca. 12.6%. Testing these materials at 100 ppm of acetone in the temperature range of 160–310 °C showed [138] that the optimal temperature for all samples, except for pristine Ti₃C₂T_x, is 255 °C. However, for the sample containing 4 wt.% of the MXene, the signal value exceeds that for α-/γ-Fe₂O₃ by eight times at close response and recovery rates. This sample also showed maximum selectivity with respect to acetone detection; its response exceeded the results for other VOCs, including ethanol, methanol, formaldehyde, and xylene, by 3–6 times, which the authors attributed to the differences in bond dissociation energies of these molecules.

Discussing the detection mechanism and, therefore, the reasons for the improved sensory properties of the Fe₂O₃-4 wt.% Ti₃C₂T_x composite, the authors summarize [138] that there is a large number of oxygen vacancies at the γ-Fe₂O₃ surface, such as titanium defects in MXene, which can serve as adsorption centers. MXene surface groups (–OH, –O, and –F) can interact with acetone molecules by forming hydrogen bonds, while the Ti₃C₂T_x core has a high conductivity, which contributes to the transport of charge carriers. Additionally, the Schottky barrier formed at the interface contributes to the transition of oxygen molecules adsorbed on the surface into the ionic state. The authors also noted that a partial Fe³⁺ reduction of the γ-Fe₂O₃ phase may occur in a reducing atmosphere, leading to the formation of Fe₃O₄ clusters with lower resistance, which are oxidized back to Fe₂O₃ in the air (Reaction Scheme (25)).

The presence of iron ions with different degrees of oxidation facilitates electron transport between defects.

$$\mathsf{\gamma }-F{e}_2^{3+}{O}_3^{2–}\stackrel{Reduction}{\to }\underset{Oxidation}{\leftarrow }F{e}^{2+}{O}^{2–}+F{e}_2^{3+}{O}_3^{2–}$$

(25)

Combination of Two n-Type Semiconductor Oxides. Some publications draw quite reasonable attention not only to the role of an added semiconducting metal oxide but also to the possibility of Ti₃C₂T_x partial oxidation resulting in TiO₂ nanoparticle formation at some stages of the composite material production process, e.g., hydrothermal/solvothermal synthesis of MO_x in the presence of MXene or thermal treating samples in air, which is not mentioned in most papers on the preparation and study of gas-sensitive properties of nanocomposites. For instance, Yao et al. highlight [139] the formation of the TiO₂ anatase phase during a hydrothermal treatment of the “(NH₄)₆Mo₇O₂₄·4H₂O–H₂O–HCl–Ti₃C₂T” reaction system at 180 °C. This is due to the fact that the MXene oxidation leads not only to sensory property changes caused by appearing heterojunctions at the interface but often results in a significant increase in the resistance of the Ti₃C₂T_x/TiO₂/MoO₃ composite, complicating or blocking the possibility of chemiresistive response measurement. In this work, the effect of MoO₃ content was also examined by varying the initial reagent content ((NH₄)₆Mo₇O₂₄: 0–50 wt.%) with respect to the Ti₃C₂T_x MXene. It was found that the introduction of MoO₃ with a large number of vacancies reduces the Ti₃C₂T_x/TiO₂ electrical conductivity and enhances the responses to various VOCs (acetone, isopropanol, toluene, methanol, ethanol, and ammonia) at RT by 20–40 times. The higher resistance of the receptor materials upon interaction with these gases indicates that Ti₃C₂T_x dominates in their detection. At the same time, there is an enhanced sensitivity to isopropanol, especially for the material with an initial ammonium molybdate content of 30 wt.%; the response to 50 ppm of C₃H₉OH at RH = 50% is 245%. The authors propose that this feature belongs to the optimal microstructure of the composite, with robust electronic bonding between MXene sheets and surface dopants, while MoO₃ excess wraps the MXene, causing a loss of its layered structure and weakening the interaction between the gas molecules and the sensor. It is noted [139] that since the resistance increases upon reducing gas adsorption, the charge carriers are holes, which are frequently characteristic for MXenes, whose number goes down as additional electrons appear due to the interaction of the material with gas molecules. The presence of defective MoO₃ on the nanocrystal surface simplifies the access of analytes to Ti₃C₂T_x layers, while the large number of oxygen-deficient sites, centers of interaction with atmospheric oxygen, also improves the sensitivity. Relying on the experimental data, the authors believe [139] that the Ti₃C₂T_x-MoO₃ heterojunction (Figure 29) plays a determining role in improving the sensory properties, while the influence of the Ti₃C₂T_x-TiO₂ heterojunction is rather limited. The authors attribute the latter effect to the low reactivity of TiO₂ and the fact that the crystallite size of this phase exceeds the Debye length of titanium dioxide. Due to the difference between the values of the work function of MoO₃ and Ti₃C₂T_x, 5.1 eV and 3.4 eV, respectively, a near-surface layer accumulating electrons in the case of MoO₃ and holes in the case of Ti₃C₂T_x is formed as a result of electron transfer from MXene to MoO₃. A negative charge growth at the MoO₃ surface due to the ion sorption of oxygen molecules leads to an evolution of the difference between its work function and that of Ti₃C₂T_x, as well as the corresponding energy band bending and broadening of the charge carrier accumulation layers. When the nanocomposite comes in contact with the gas, the negative charge of the surface decreases and the hole accumulation layer becomes narrower, i.e., the resistance of the material grows.

Wu and colleagues [140] synthesized the composite material of Ti₃C₂T_x/TiO₂/SnO₂ by mixing aqueous suspensions of previously synthesized Ti₃C₂T_x and SnO₂ nanofibers, followed by a long exposure of the resulting complex dispersion at 80 °C, which aimed at partial oxidation of MXene. The oxidation was confirmed experimentally by XRD, HRTEM, and XPS data. The mass ratio of Ti₃C₂T_x to SnO₂ was 2:1. It was found that the introduction of electrospun SnO₂ nanofibers into the composite resulted in an enhanced response to all the gases under test (NH₃, acetone, ethanol, and H₂, NO₂), not only compared with the individual components but also with respect to the Ti₃C₂T_x/TiO₂ composite, which the authors explain by the suitable energy band structure of the three phases. A higher sensitivity towards nitrogen dioxide was revealed: the response to 10 ppm of NO₂ was ca. 32.4%. It was observed that the response pattern to oxidizing and reducing gases of various chemical structures with increasing resistance indicates the detection mechanism is mainly related to the surface charge transfer model [140]. In other words, the surface of the Ti₃C₂T_x/TiO₂/SnO₂ heterostructure retains the functional groups characteristic of the Ti₃C₂T_x MXene, which provide multiple active adsorption centers and fast electron transport channels at RT.

Additionally, several types of heterojunctions, SnO₂–TiO₂ and Ti₃C₂T_x–TiO₂, are generated and can regulate the energy band structure of Ti₃C₂T_x.

Among n-type semiconductor metal oxides, indium and iron (III) oxides can also be highlighted, as they have been used to produce nanocomposites that are promising to employ in chemiresistive gas sensors.

Summarizing the data analyzed on the sensory properties of nanocomposites based on Ti₃C₂T_x and n-type semiconductor metal oxides (MO_x), we might distinguish two cases:

(1)

Mxene doped with nanosized metal oxide particles;

(2)

Semiconductor metal oxide modified with MXene.

There is an improvement in sensor properties in both cases due to an enhancement of the specific surface area because MO_x nanoparticles in the MXene interlayer space prevent their spontaneous agglomeration during coating application while MXene sheets limit the formation of large MO_x aggregates during their synthesis. Owing to the presence of polar functional groups on the Ti₃C₂T_x surface, new phase nucleation occurs on them, whose coalescence is further prevented by the MXene planes. This leads to a higher number of adsorption centers and, consequently, more sensitivity.

The formation of a heterojunction between MO_x semiconductor particles and Ti₃C₂T_x having metallic conductivity makes it possible to implement a charge separation at the interface and, therefore, to promote the response upon gas analyte intake.

In the case of the prevailing sensing mechanism of n-type semiconducting metal oxide, the introduction of MXene, which has high conductivity, promotes charge carrier transport, resulting in decreased sensor response and recovery times as well as a lower detection temperature. Moreover, the presence of –O, –OH, and –F functional groups on the surface of MXene sheets, as well as defects, improves analyte gas sorption, especially for those capable of forming hydrogen bonds.

4.1.2. Modification with p-Type Metal Oxide Semiconductors

CuO and Cu₂O. In contrast to the huge amount of scientific research on MXene composites containing n-type semiconducting metal oxides, the number of papers where Ti₃C₂T_x or Ti₂CT_x MXenes are modified with p-type semiconductor oxide is drastically smaller. For this review, we managed to collect articles describing the sensor properties of composite materials based on Ti₃C₂T_x and CuO, Co₃O₄, lithium, sodium, and potassium tungstate, as well as defective WO₃:Cr.

To improve primarily the sensitivity and kinetic characteristics of CuO–based sensors, which were found to be highly sensitive to H₂S and VOCs, including BTEX, Hermawan and colleagues [141] fabricated CuO/Ti₃C₂T_x composites with a MXene content ranging from 10 wt.% to 40 wt.% by electrostatic assembly under ultrasound treatment of the joint dispersion of components. Studies of gas-sensing properties toward toluene (10–50 ppm) showed that the CuO—30 wt.% Ti₃C₂T_x composition has the best performance due to the highest specific surface area. It was also highlighted that advancing the operating temperature from 100 °C to 250 °C, facilitates the response value, R/R₀, growing from ca. 1.2 to ca. 11.4 in the case of a 50 ppm concentration of toluene. This value exceeded the one characteristic for pristine CuO by almost five times. The described CuO/Ti₃C₂T_x composite material showed slightly lower sensitivity to other VOCs; the responses to ethanol, acetone, methanol, and hydrogen at 50 ppm concentration were 7.3, 7.2, 5.4, and 2.8, respectively. The authors have demonstrated experimentally the advantage of CuO hybridization with accordion-like Ti₃C₂T_x compared with other 2D materials, in particular MoS₂ and rGO [141]. The humidity, however, still dramatically deteriorates the CuO—30 wt.% Ti₃C₂T_x-based sensor performance. When the RH value increased from 0% to 70%, the response to 50 ppm toluene got lower, from ca. 11.4 to ca. 2.1. This is again explained by the competition between adsorbing toluene and water molecules for the active centers of the receptor material.

The improvement of CuO/Ti₃C₂T_x sensor performance is explained by the authors [141] via the appearance of a Schottky barrier at the interface due to the difference in work functions, which are 3.9 eV and 4.7 eV for Ti₃C₂T_x and CuO, respectively. Such a barrier facilitates a hole trapping region (HTR) on the MXene side (Figure 30), characterized by a higher amount of ion-sorbed oxygen species (O⁻, specific for the chosen temperature). The interaction of toluene molecules with the O⁻ ions on the CuO surface leads to a reduction in their surface concentration and, consequently, a narrowing the depletion region (hole accumulation layers) and a diminishing of sensor resistance. The authors also highlight the supporting role of MXene as a conductive layer for the faster charge carrier mobility through a network of CuO nanoparticles, which contributes to a shorter response/recovery time [141].

In addition, the research [142] showed the promising prospects for developing a self-powered ammonia sensor for pork quality control based on CuO obtained as a result of the thermal destruction of Cu-containing MOF modified with a minor amount of accordion-like Ti₃C₂T_x.

Zhou and colleagues [143] described the modification of the p-type semiconducting Cu₂O, with an energy band gap of 2.17 eV, by Ti₃C₂T_x MXene, 10–50 wt.%, in order to increase its sensitivity to triethylamine. For this purpose, hollow Cu₂O nanospheres produced by reducing Cu²⁺ with NH₂OH·HCl in the presence of sodium dodecyl sulfate (SDS) while adding sodium hydroxide were dispersed under ultrasound exposure along with accordion-like Ti₃C₂T_x. Detecting 10 ppm of trimethylamine (TEA) at RT and 50 rel.% humidity showed that the composition containing 20 wt.% MXene exhibited the highest sensitivity to TEA, with a response equal to ca. 181.4%, which was almost 3.5 times higher than that of pristine Cu₂O. At the same time, it was noted that doping allowed one to reduce the sensor response time by three times. It was found that a change in humidity from 30 rel.% to 90 rel.% decreased the response from 220% to 109%. Measuring the response of Cu₂O-20 wt.% Ti₃C₂T_x material to 10 ppm of other gas analytes (ammonia, methanol, ethanol, acetone, and hydrogen sulfide) exhibited that even the largest response to H₂S was 13.4%, while the response to other gases did not exceed 9%, which indicates a high selectivity toward triethylamine (Figure 31).

However, when testing the material for 30 days, the response decreased by 15%, displaying a continued degradation of the functional properties of the receptor material. Still, the authors explain advancing the Cu₂O sensing properties in the presence of Ti₃C₂T_x by the growth of a specific surface area, as well as the acceleration of carrier transport because of their high mobility in MXene and the formation of a Schottky barrier at the composite interfaces.

Co₃O₄. Doping with MXene, but at a much lower content of Ti₃C₂T_x, was also performed by Bu et al. [144] to synthesize highly dispersed Co₃O₄/Ti₃C₂T_x (1, 2, or 5%) composites. Similar to other works [134,138,142], the approach to obtain highly porous metal oxide semiconductors via MOFs’ decomposition/oxidation was applied. In the first step, rhombohedral dodecahedrons of organometallic frameworks (ZIF-67) are grown on the surface of delaminated MXene sheets during the interaction of cobalt(III) nitrate and 2-methylimidazole in an ethanol medium in the presence of Ti₃C₂T_x (Figure 32). After separation and drying, the product was calcined in air at 350 °C for 1 h to form a Co₃O₄/Ti₃C₂T_x nanocomposite. Gas-sensitivity screening of the materials on the example of 50 ppm ethanol in the temperature range of 150–275 °C allowed (1) to establish the optimal operating temperature of 200 °C and (2) to select Co₃O₄–2 wt.% Ti₃C₂T_x as the most sensitive composite; the response, R/R₀, was equal to 190 at RH = 30%. It was also found that there was an 18-fold improvement in response and recovery times compared with pristine Co₃O₄. A long-term stability test of the composite material over 60 days showed a gradual decrease in sensitivity over this time by approx. 10%. The authors point out that Co₃O₄–2 wt.% Ti₃C₂T_x also has a higher sensitivity than Co₃O₄ with respect to other gases, such as methanol, isopropanol, and acetone; however, the response value to a 50 ppm concentration of these analytes does not exceed 71. At the same time, when the humidity increases from 30 rel.% to 90 rel.%, the response to 50 ppm of ethanol decreases from ca. 190 to ca. 160. The background of the gas detection mechanism seems to follow the concept of the interaction of adsorbed analyte molecules with ion-adsorbed species of oxygen on Co₃O₄ particle surfaces as a typical p-type semiconductor. The authors attribute improving the sensor performance of cobalt oxide with 2 wt.% Ti₃C₂T_x to (1) the formation of a heterojunction between Co₃O₄, a p-type semiconductor, and MXene, which has metallic conductivity, and (2) the role of Ti₃C₂T_x as a platform for uniform nucleation and formation of nanoporous Co₃O₄, which promotes an increased specific surface area of nanocomposites and a higher number of adsorption centers.

Sun et al. [145] used a hydrothermal method to develop a composite material composed of layered Ti₃C₂T_x and cobalt-aluminum layered double hydroxide (n(Co³⁺):n(Al³⁺) = 2:1) in a 3D hierarchical structure, which was further calcined in a N₂ to grow Co₃O₄/Al₂O₃@Ti₃C₂T_x compositions. The XRD pattern of the latter displayed only Co₃O₄ reflections; the Raman spectra also yielded exclusively A_1g, E_g, and F_2g modes of the Co₃O₄ phase, while HRTEM images supplied Al₂O₃ interplanar distances as well, and the XPS spectrum of Ti2p was found to be similar to that of Ti₃C₂T_x. While measuring 100 ppm of NO_x at RT and RH = 26%, the sensitivity was noted not only for the Co₃O₄/Al₂O₃@Ti₃C₂T_x composite but also for the precursor containing cobalt-aluminum layered hydroxide. However, the response value of the composite is found to be significantly higher. Raising the humidity from 26 rel.% to 75 rel.% leads to a chemiresistive signal, R₀/R, lowering from ca. 40.3 to ca. 10.7. The improved sensing properties of Co₃O₄/Al₂O₃@Ti₃C₂T_x are explained by the synergy of three factors:

−

Interaction of Co₃O₄ and Ti₃C₂T_x phases, promoting electron transfer, and increasing the diffusion rate of charge carriers;

−

Three-dimensional structure of the receptor material, allowing channels for gas inflow that facilitates reducing the response/recovery time;

−

High specific surface area and porous structure, leading to the formation of adsorption centers; the Al₂O₃ addition also promotes the appearance of defects, thus improving the adsorption capacity of the composite.

Modification with n-Type Metal Oxide Semiconductors. Ama et al. [146] studied the features of gas-sensing properties of composite materials in the K₂W₇O₂₂–Ti₃C₂T_x system based on their previous findings on employing potassium tungstate (K₂W₇O₂₂) toward RT acetone detection. Thus, to obtain the receptor material, they mixed dispersions of hydrothermally pre-synthesized potassium tungstate nanorods with MXene obtained by etching Ti₃AlC₂ MAX-phase with 5% HF under hydrothermal conditions. After dwell time while stirring the reaction system, which supported an electrostatic assembly, the composites, K₂W₇O₂₂:Ti₃C₂T_x, with different component ratios ranging from 1:5 to 9:1, were separated from the mother liquor, washed, and dried. Measurement of responses to 2.86 ppm of ethanol at RT under RH = 20% showed that the 2K₂W₇O₂₂–1Ti₃C₂T_x composite exhibited the greatest sensitivity; the response was nearly 10 times higher than for pristine potassium tungstate. The authors suggested that this effect be facilitated by Ti₃C₂T_x’s contribution to boosting electrical conductivity and charge transfer. In addition, the interaction between the components raises the contribution of the (002) face, which provides enhanced sensitivity of the p-type semiconductor K₂W₇O₂₂ to ammonia at RT, as previously established elsewhere [147]. For the 2K₂W₇O₂₂–1Ti₃C₂T_x composition, a high signal stability in a wide range of humidity, 10–86 rel.%, was also found.

The M₂W₇O₂₂ nanowires of lithium, sodium, potassium, and chromium tungstate, as well as MXene-containing 2M₂W₇O₂₂–1Ti₃C₂T_x nanocomposites, were synthesized by a similar method [148]. Testing of the receptor materials at 2.86 ppm of acetone allowed authors to identify hybrid MXene/Cr-doped tungsten oxide as the most promising material. The favorable properties of these structures are explained by the large number of active centers available for acetone molecule adsorption as well as by the stronger ferroelectric characteristics of the synthesized p-type oxide semiconductor. Moreover, the response of this receptor material is ca. 4.7 times higher than that of the 2K₂W₇O₂₂–1Ti₃C₂T_x material described earlier [146]. A higher selectivity to acetone was noticed when compared with some other VOC vapors such as ethanol, methanol, toluene, and even water, whose response is almost two orders of magnitude lower.

The authors explain the improvement of sensory characteristics by a combination of several factors, including (i) the Schottky barrier formation at the interface between the semiconductor and MXene, (ii) the separation of charge carriers, (iii) chemical effects in terms of lower activation energy, catalytic activity, and surface reactions, and (iv) an increase in the specific surface area of the nanocomposite. Additionally, the characteristics of the underlying semiconductor material dominate as well.

Table 2. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and semiconducting metal oxides.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
Doping component is an n-type semiconductor metal oxide
	10 ppm	RT	+1.9	100 ppb	Ti2CTx/TiO2	LiF + HCl (12 M), 40 °C, 24 h, washing with HF (10%), +TiO2: Ti2CTx incubation in ethanol-water mixture, 4 °C, 16 h	[99]
NO2	5 ppm	RT	+16.05	–	Ti3C2Tx/TiO2	LiF + HCl (9 M), 35 °C, 24 h, +TiO2: Ti3C2Tx dispersion was maintained at 80 °C for 8 h	[103]
NO2	5 ppm	RT	−19.854	10 ppb	Ti3C2Tx/TiO2/rGO	Mixing of Ti3C2Tx and GO dispersions (mass ratio 2:1).	[104]
Acetone	2 ppm	350 °C	−180	20 ppb	Ti3C2Tx/TiO2	LiF + HCl (6 M), 35 °C, 24 h, +TiO2: Heat treatment of the coating in air at 350 °C, 24 h	[105]
Methanol			−110	35 ppb
Isopropanol			−100	40 ppb
Ethanol			−40	150 ppb
Ethanol	100 ppm	RT	−22.47	10 ppm	Ti3C2Tx/TiO2	LiF + HCl (38 wt.%), 60 °C, 24 h, US 20 min +TiO2: exposure to oxygen plasma in the MPCVD reactor at 500 °C, 1 h	[107]
Hexanal	100 ppm	RT	+8.8	217 ppb	Ti3C2Tx/TiO2	HF (40 wt.%), 40 °C, 36 h, +TiO2: hydrothermal treatment of Ti3C2Tx water-ethanol suspension at 200 °C, 24 h	[110]
Acetone			+3.0	-
Methanol			+3.5	-
Benzene			+0.9	-
NO2	100 ppm	RT	+4.455	-	Ti3C2Tx/TiO2	HF (40 wt.%), RT, 24 h, +TiO2: hydrothermal treatment of Ti3C2Tx water-ethanol suspension at 180 °C, 20 h	[111]
SO2			+2.111	-
NH3			+0.703	-
CO			+0.667	-
NO2		175 °C	+19.76	1 ppm
Ammonia	30 ppm	RT, UV activation (365 nm)	~−41 *	5 ppb	Ti3C2Tx/TiO2	HF (50 wt.%), RT, 2 h,+TiO2: hydrothermal treatment of Ti3C2Tx di spersion in 0.1M NaBF4 solution at 160 °C, 12 h	[113]
Ammonia	200 ppb	20 °C	+7.3	-	Ti3C2Tx (N-doped)/TiO2	LiF + HCl (6 M), 35 °C, 24 h, +TiO2: solvothermal treatment of Ti3C2Tx dispersion in urea-containing ethanol-ethylene glycol mixture at 200 °C, 18 h	[114]
Ethanol	10 ppm	230 °C	−5	0.5 ppm	SnO2/Ti3C2Tx (20 wt.%)	HF (40 wt.%), 60 °C, 48 h, +SnO2: hydrothermal treatment of K2SnO3·3H2O solution in the presence of urea and polyvinylpyrrolidone with the addition of Ti3C2Tx at 200 °C, 18 h	[118]
Trimethylamine	50 ppm	140 °C	−33.9 **	5 ppm	SnO2/Ti3C2Tx	HF (40 wt.%), 40 °C, 24 h, +SnO2: hydrothermal treatment of a mixture of SnCl4·H2O solution with Ti3C2Tx and urea- and CTAB-containing solution at 140 °C, 24 h	[119]
NO2	10 ppm	25 °C, desorption—pulse heating up to 100 °C	+1.57	0.5 ppm	SnO2/Ti3C2Tx	HF (40 wt.%), 35 °C, 24 h, +SnO2: hydrothermal treatment of Ti3C2Tx dispersion in water-ethanol medium containing SnCl4·H2O and 9% HCl, at 180 °C, 24 h	[120]
Ammonia	200 ppm	RT	+7.8 **	1 ppm	SnO/Ti3C2Tx	HFconc, RT, 24 h, +SnO: hydrothermal treatment of US-activated Ti3C2Tx dispersion in SnCl2·2H2O aqueous solution containing urea and hydrochloric acid at 120 °C, 24 h	[121]
Acetone	100 ppm	RT	+12.1 *	10 ppm	SnO-SnO2/Ti3C2Tx	HFconc, RT, 24 h, +SnO: hydrothermal treatment of US-activated Ti3C2Tx dispersion in SnCl2·2H2O aqueous solution containing urea and hydrochloric acid at 120 °C, 8 h	[112]
NO2	100 ppm	RT, RH = 70%	−41.6	-	Ti3C2Tx (crumpled spheres)/ZnO	Ultrasonic pyrolysis: ultrasonic spraying/drying of few-layered Ti3C2Tx dispersion in which polyvinylpyrrolidone and zinc acetate are dissolved, heat treatment in N2 at 800 °C	[122]
Ammonia			+4.2	-
NO2	20 ppm	RT, desorbtion—UV	+367.63	0.5 ppm	Ti3C2Tx/ZnO	HF, RT, 24 h, +ZnO: mixing and ultrasonic treatment of Ti3C2Tx dispersion with porous, layered ZnO nanopowder	[123]
NO2	20 ppb	RT, UV	+81	200 ppt	Ti3C2Tx/ZnO	LiF + HCl (9 M), 40 °C, 24 h, +ZnO: (1) dispersion of Ti3C2Tx in Zn(CH3COO)2 methanol solution, addition of NaOH methanol solution, (2) the powder obtained in step 1 was dispersed and maintained in an aqueous NaOH solution along with ε-Zn(OH)2, 80 °C, 0.5 h	[124]
Acetone	100 ppm	320 °C	−14.1 *	20 ppm	ZnO/Ti3C2Tx	HF, 30 °C, 24 h, delamination: TMAOH +ZnO: hydrothermal synthesis of ZnO from zinc acetate and CTAB with 2 wt% Ti3C2Tx MXene added	[125]
NO2	8 ppm	160 °C	+3.4 *	-	ZnO/Ti3C2Tx	LiF + HCl (12 M), 38 °C, 48 h, +ZnO: solvothermal treatment of a system containing ZnCl2, sodium acetate, sodium citrate, and dispersed Ti3C2Tx, in ethylene glycol	[126]
Formaldehyde	100 ppm	RT	+194.7	5 ppm	ZnSnO3/Ti3C2Tx	HF (10 wt.%), 50 °C, 8 h, Delamination: DMSO, RT, 24 h, + ZnSnO3: hydrothermal treatment of a suspension containing pre-synthesized CTAB-treated ZnSnO3 with the addition of Ti3C2Tx, 150 °C, 24 h	[127]
Ammonia	1 ppm	RT	+22.3	-	Ti3C2Tx/WO3 (50 wt.%)	HF (40%), RT, 24 h, +WO3: Ultrasonic treatment of suspension obtained by mixing Ti3C2Tx and WO3 dispersions	[128]
NO2	200 ppb	RT, RH = 0–99%	+184	30 ppb	Ti3C2Tx/WO3	HF (38%), RT, 24 h, sodium L-ascorbate treatment, +WO3: hydrothermal synthesis of WO3 from Na2WO4·2H2O, thioacetamide, and HCl in the presence of Ti3C2Tx, 200 °C, 20 h	[129]
NO2	200 ppb	RT	+78%	15 ppb	WO3/Ti3C2Tx (20 wt.%)	HF (38%), RT, 24 h, delamination: DMSO, US, +WO3: hydrothermal synthesis of WO3 from Na2WO4·2H2O, thiourea, and HNO3 in the presence of Ti3C2Tx, 180 °C, 24 h	[131]
Acetone	20 ppm	300 °C	−12.1 **	0.17 ppm	W18O49/Ti3C2Tx (2 wt.%)	HF, RT, 18 h, delamination: DMSO, US, +W18O49: solvothermal synthesis by heat treatment of WCl6 solution in ethanol with the addition of acetylacetone in the presence of Ti3C2Tx, 150 °C, 24 h	[132]
Methanol	5 ppm	RT	−29.6	-	In2O3/Ti3C2Tx	HF (40 wt.%), 35 °C, 6 h, delamination: DMSO, RT, 24 h, + In2O3: hydrothermal synthesis of In2O3 nanocubes from a solution of InCl3, sodium citrate, and urea, 140 °C, 24 h, In2O3 surface modification with surfactant (APTES), subsequent hydrothermal treatment after mixing with Ti3C2Tx, 120 °C, 14 h	[133]
Ammonia	5 ppm	RT	−60.6	-	In2O3/Ti3C2Tx	HF, 6 h, 35 °C, delamination: US,1 h, + In2O3: MOFs self-assembly upon thermal treatment of In(NO3)3 and terephthalic acid solution in DMF, 120 °C, 30 min, followed by annealing in air at 500 °C, 3 h	[134]
Ammonia	30 ppm	RT, RH = 40%	+63.8%	2 ppm	Ti3C2Tx/In2O3 (50 wt.%)	NaF + HCl, 60 °C, 24 h, +In2O3: ultrasonic treatment (RT, 5 h) of a mixed aqueous suspension of Ti3C2Tx and In2O3 obtained by the solvothermal method from In(NO3)3 in glycerol-isopropanol solution	[135]
Acetone	5 ppm	RT, RH = 22.3%	−16.6	-	α-Fe2O3/Ti3C2Tx	HF, 60 °C, 6 h, delamination: US, 80 min, +Fe2O3: Fe2O3 nanocubes synthesis by FeCl3 and NaOH mixing followed by precipitate aging, 100 °C, 4 days, modification of Fe2O3 surface with surfactant (APTES), subsequent hydrothermal treatment after mixing with Ti3C2Tx, 120 °C, 14 h	[136]
Ammonia	5 ppm	RT, RH = 22%	−18.3	-	α-Fe2O3/Ti3C2Tx	HF, 60 °C, 8 h, delamination: US, 1 h, +Fe2O3: solvothermal synthesis using a solution of FeCl3 and urea in ethylene glycol and isopropanol with Ti3C2Tx dispersion addition, 120 °C, 14 h, subsequent heat treatment at 500 °C, 2 h, Ar	[137]
Acetone	100 ppm	255 °C	−215.2 **	-	α-/γ-Fe2O3/Ti3C2Tx (4 wt.%)	LiF+HCl (6M), 35 °C, 24 h, DMF, US, 3 h, +α-/γ-Fe2O3: solvothermal treatment of Ti3C2Tx dispersion containing Fe(NO3)3 and terephthalic acid, 150 °C, 24 h, subsequent calcination, 450 °C, 4 h	[138]
Isopropanol	50 ppm	RT, RH = 50%	+245%	5 ppm	Ti3C2Tx/TiO2/MoO3	NaF + HCl (9M), 35 °C, 24 h, US, Ar, 1 h, +TiO2/MoO3: hydrothermal treatment of (NH4)6Mo7O24·4H2O–H2O–HCl–Ti3C2Tx” system, 180 °C, 24 h; the initial (NH4)6Mo7O24·4H2O content is 30% of Ti3C2Tx	[139]
NO2	10 ppm	RT	+32.4%	0.1 ppm	Ti3C2Tx/TiO2/SnO2	LiF + HCl, 70 °C, 36 h, US, 30 min, +TiO2/SnO2: aging of SnO2 nanofibers and Ti3C2Tx mixed dispersion, 80 °C, 10 h; mass ratio is 1SnO2-2Ti3C2Tx	[140]
Doping component is an p-type semiconductor metal oxide
Toluene	50 ppm	250 °C	+11.4 *	10 ppm	CuO/Ti3C2Tx (30 wt.%)	HF (40%), RT, 24 h, +CuO: ultrasonic treatment of CuO ethanol dispersion obtained by solvothermal method from copper acetate in anhydrous ethanol medium (150 °C, 12 h), with the addition of freshly prepared Ti3C2Tx powder, RT, 20 min	[141]
Triethylamine	10 pmm	RT, RH = 50%	+181.4	5 ppm	Cu2O/Ti3C2Tx (20 wt.%)	NaF+HCl, 60 °C, 48 h, +Cu2O: ultrasonic treatment of ethanol dispersion containing accordion-like Ti3C2Tx and Cu2O nanospheres derived from CuCl2 in the presence of SDS, NH2OH·HCl and NaOH, RT, 3 h	[143]
Ethanol	50 ppm	200 °C, RH = 30%	+190 *	1 ppm	Co3O4/Ti3C2Tx (2 wt.%)	HF (40%), 35 °C, 24 h, delamination: DMSO, 24 h, US, 1 h, +Co3O4: MOFs self-assembly on the Ti3C2Tx sheet surface out of Co(NO3)3 and 2-methylimidazole methanol solution, ultrasound, RT, 30 min, 24 h, subsequent calcination in air at 350 °C, 1 h	[144]
NOx	100 ppm	RT, RH = 26%	−40 **	0.01 ppm	Co3O4-Al2O3/Ti3C2Tx	LiF, HCl(9M), 35 °C, 48 h, delamination: US, 30 min, Ar, +Co3O4-Al2O3: hydrothermal synthesis of cobalt-aluminum layered double hydroxide on Ti3C2Tx surface from dispersion containing Co(NO3)3, Al(NO3)3, urea, and polyvinylpyrrolidone, 90 °C, 6 h, subsequent heat treatment at 450 °C, 3 h, N2	[145]
Acetone	2.86 ppm	RT, RH = 20%	+250	-	2K2W7O22–1Ti3C2Tx	HF (5 wt.%), hydrothermal treatment, 150 °C, 5 h, US, 1 h +K2W7O22: hydrothermal synthesis of K2W7O22 from Na2WO4·2H2O, oxalic acid, K2SO4, and HCl solution, 225 °C, 24 h, stirring of K2W7O22 + Ti3C2Tx dispersions, 12 h	[146]

* S = R/R₀ [112,113,125,126,141,144]; ** S = R₀/R [119,121,132,138].

Table 2. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and semiconducting metal oxides.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
Doping component is an n-type semiconductor metal oxide
	10 ppm	RT	+1.9	100 ppb	Ti2CTx/TiO2	LiF + HCl (12 M), 40 °C, 24 h, washing with HF (10%), +TiO2: Ti2CTx incubation in ethanol-water mixture, 4 °C, 16 h	[99]
NO2	5 ppm	RT	+16.05	–	Ti3C2Tx/TiO2	LiF + HCl (9 M), 35 °C, 24 h, +TiO2: Ti3C2Tx dispersion was maintained at 80 °C for 8 h	[103]
NO2	5 ppm	RT	−19.854	10 ppb	Ti3C2Tx/TiO2/rGO	Mixing of Ti3C2Tx and GO dispersions (mass ratio 2:1).	[104]
Acetone	2 ppm	350 °C	−180	20 ppb	Ti3C2Tx/TiO2	LiF + HCl (6 M), 35 °C, 24 h, +TiO2: Heat treatment of the coating in air at 350 °C, 24 h	[105]
Methanol			−110	35 ppb
Isopropanol			−100	40 ppb
Ethanol			−40	150 ppb
Ethanol	100 ppm	RT	−22.47	10 ppm	Ti3C2Tx/TiO2	LiF + HCl (38 wt.%), 60 °C, 24 h, US 20 min +TiO2: exposure to oxygen plasma in the MPCVD reactor at 500 °C, 1 h	[107]
Hexanal	100 ppm	RT	+8.8	217 ppb	Ti3C2Tx/TiO2	HF (40 wt.%), 40 °C, 36 h, +TiO2: hydrothermal treatment of Ti3C2Tx water-ethanol suspension at 200 °C, 24 h	[110]
Acetone			+3.0	-
Methanol			+3.5	-
Benzene			+0.9	-
NO2	100 ppm	RT	+4.455	-	Ti3C2Tx/TiO2	HF (40 wt.%), RT, 24 h, +TiO2: hydrothermal treatment of Ti3C2Tx water-ethanol suspension at 180 °C, 20 h	[111]
SO2			+2.111	-
NH3			+0.703	-
CO			+0.667	-
NO2		175 °C	+19.76	1 ppm
Ammonia	30 ppm	RT, UV activation (365 nm)	~−41 *	5 ppb	Ti3C2Tx/TiO2	HF (50 wt.%), RT, 2 h,+TiO2: hydrothermal treatment of Ti3C2Tx di spersion in 0.1M NaBF4 solution at 160 °C, 12 h	[113]
Ammonia	200 ppb	20 °C	+7.3	-	Ti3C2Tx (N-doped)/TiO2	LiF + HCl (6 M), 35 °C, 24 h, +TiO2: solvothermal treatment of Ti3C2Tx dispersion in urea-containing ethanol-ethylene glycol mixture at 200 °C, 18 h	[114]
Ethanol	10 ppm	230 °C	−5	0.5 ppm	SnO2/Ti3C2Tx (20 wt.%)	HF (40 wt.%), 60 °C, 48 h, +SnO2: hydrothermal treatment of K2SnO3·3H2O solution in the presence of urea and polyvinylpyrrolidone with the addition of Ti3C2Tx at 200 °C, 18 h	[118]
Trimethylamine	50 ppm	140 °C	−33.9 **	5 ppm	SnO2/Ti3C2Tx	HF (40 wt.%), 40 °C, 24 h, +SnO2: hydrothermal treatment of a mixture of SnCl4·H2O solution with Ti3C2Tx and urea- and CTAB-containing solution at 140 °C, 24 h	[119]
NO2	10 ppm	25 °C, desorption—pulse heating up to 100 °C	+1.57	0.5 ppm	SnO2/Ti3C2Tx	HF (40 wt.%), 35 °C, 24 h, +SnO2: hydrothermal treatment of Ti3C2Tx dispersion in water-ethanol medium containing SnCl4·H2O and 9% HCl, at 180 °C, 24 h	[120]
Ammonia	200 ppm	RT	+7.8 **	1 ppm	SnO/Ti3C2Tx	HFconc, RT, 24 h, +SnO: hydrothermal treatment of US-activated Ti3C2Tx dispersion in SnCl2·2H2O aqueous solution containing urea and hydrochloric acid at 120 °C, 24 h	[121]
Acetone	100 ppm	RT	+12.1 *	10 ppm	SnO-SnO2/Ti3C2Tx	HFconc, RT, 24 h, +SnO: hydrothermal treatment of US-activated Ti3C2Tx dispersion in SnCl2·2H2O aqueous solution containing urea and hydrochloric acid at 120 °C, 8 h	[112]
NO2	100 ppm	RT, RH = 70%	−41.6	-	Ti3C2Tx (crumpled spheres)/ZnO	Ultrasonic pyrolysis: ultrasonic spraying/drying of few-layered Ti3C2Tx dispersion in which polyvinylpyrrolidone and zinc acetate are dissolved, heat treatment in N2 at 800 °C	[122]
Ammonia			+4.2	-
NO2	20 ppm	RT, desorbtion—UV	+367.63	0.5 ppm	Ti3C2Tx/ZnO	HF, RT, 24 h, +ZnO: mixing and ultrasonic treatment of Ti3C2Tx dispersion with porous, layered ZnO nanopowder	[123]
NO2	20 ppb	RT, UV	+81	200 ppt	Ti3C2Tx/ZnO	LiF + HCl (9 M), 40 °C, 24 h, +ZnO: (1) dispersion of Ti3C2Tx in Zn(CH3COO)2 methanol solution, addition of NaOH methanol solution, (2) the powder obtained in step 1 was dispersed and maintained in an aqueous NaOH solution along with ε-Zn(OH)2, 80 °C, 0.5 h	[124]
Acetone	100 ppm	320 °C	−14.1 *	20 ppm	ZnO/Ti3C2Tx	HF, 30 °C, 24 h, delamination: TMAOH +ZnO: hydrothermal synthesis of ZnO from zinc acetate and CTAB with 2 wt% Ti3C2Tx MXene added	[125]
NO2	8 ppm	160 °C	+3.4 *	-	ZnO/Ti3C2Tx	LiF + HCl (12 M), 38 °C, 48 h, +ZnO: solvothermal treatment of a system containing ZnCl2, sodium acetate, sodium citrate, and dispersed Ti3C2Tx, in ethylene glycol	[126]
Formaldehyde	100 ppm	RT	+194.7	5 ppm	ZnSnO3/Ti3C2Tx	HF (10 wt.%), 50 °C, 8 h, Delamination: DMSO, RT, 24 h, + ZnSnO3: hydrothermal treatment of a suspension containing pre-synthesized CTAB-treated ZnSnO3 with the addition of Ti3C2Tx, 150 °C, 24 h	[127]
Ammonia	1 ppm	RT	+22.3	-	Ti3C2Tx/WO3 (50 wt.%)	HF (40%), RT, 24 h, +WO3: Ultrasonic treatment of suspension obtained by mixing Ti3C2Tx and WO3 dispersions	[128]
NO2	200 ppb	RT, RH = 0–99%	+184	30 ppb	Ti3C2Tx/WO3	HF (38%), RT, 24 h, sodium L-ascorbate treatment, +WO3: hydrothermal synthesis of WO3 from Na2WO4·2H2O, thioacetamide, and HCl in the presence of Ti3C2Tx, 200 °C, 20 h	[129]
NO2	200 ppb	RT	+78%	15 ppb	WO3/Ti3C2Tx (20 wt.%)	HF (38%), RT, 24 h, delamination: DMSO, US, +WO3: hydrothermal synthesis of WO3 from Na2WO4·2H2O, thiourea, and HNO3 in the presence of Ti3C2Tx, 180 °C, 24 h	[131]
Acetone	20 ppm	300 °C	−12.1 **	0.17 ppm	W18O49/Ti3C2Tx (2 wt.%)	HF, RT, 18 h, delamination: DMSO, US, +W18O49: solvothermal synthesis by heat treatment of WCl6 solution in ethanol with the addition of acetylacetone in the presence of Ti3C2Tx, 150 °C, 24 h	[132]
Methanol	5 ppm	RT	−29.6	-	In2O3/Ti3C2Tx	HF (40 wt.%), 35 °C, 6 h, delamination: DMSO, RT, 24 h, + In2O3: hydrothermal synthesis of In2O3 nanocubes from a solution of InCl3, sodium citrate, and urea, 140 °C, 24 h, In2O3 surface modification with surfactant (APTES), subsequent hydrothermal treatment after mixing with Ti3C2Tx, 120 °C, 14 h	[133]
Ammonia	5 ppm	RT	−60.6	-	In2O3/Ti3C2Tx	HF, 6 h, 35 °C, delamination: US,1 h, + In2O3: MOFs self-assembly upon thermal treatment of In(NO3)3 and terephthalic acid solution in DMF, 120 °C, 30 min, followed by annealing in air at 500 °C, 3 h	[134]
Ammonia	30 ppm	RT, RH = 40%	+63.8%	2 ppm	Ti3C2Tx/In2O3 (50 wt.%)	NaF + HCl, 60 °C, 24 h, +In2O3: ultrasonic treatment (RT, 5 h) of a mixed aqueous suspension of Ti3C2Tx and In2O3 obtained by the solvothermal method from In(NO3)3 in glycerol-isopropanol solution	[135]
Acetone	5 ppm	RT, RH = 22.3%	−16.6	-	α-Fe2O3/Ti3C2Tx	HF, 60 °C, 6 h, delamination: US, 80 min, +Fe2O3: Fe2O3 nanocubes synthesis by FeCl3 and NaOH mixing followed by precipitate aging, 100 °C, 4 days, modification of Fe2O3 surface with surfactant (APTES), subsequent hydrothermal treatment after mixing with Ti3C2Tx, 120 °C, 14 h	[136]
Ammonia	5 ppm	RT, RH = 22%	−18.3	-	α-Fe2O3/Ti3C2Tx	HF, 60 °C, 8 h, delamination: US, 1 h, +Fe2O3: solvothermal synthesis using a solution of FeCl3 and urea in ethylene glycol and isopropanol with Ti3C2Tx dispersion addition, 120 °C, 14 h, subsequent heat treatment at 500 °C, 2 h, Ar	[137]
Acetone	100 ppm	255 °C	−215.2 **	-	α-/γ-Fe2O3/Ti3C2Tx (4 wt.%)	LiF+HCl (6M), 35 °C, 24 h, DMF, US, 3 h, +α-/γ-Fe2O3: solvothermal treatment of Ti3C2Tx dispersion containing Fe(NO3)3 and terephthalic acid, 150 °C, 24 h, subsequent calcination, 450 °C, 4 h	[138]
Isopropanol	50 ppm	RT, RH = 50%	+245%	5 ppm	Ti3C2Tx/TiO2/MoO3	NaF + HCl (9M), 35 °C, 24 h, US, Ar, 1 h, +TiO2/MoO3: hydrothermal treatment of (NH4)6Mo7O24·4H2O–H2O–HCl–Ti3C2Tx” system, 180 °C, 24 h; the initial (NH4)6Mo7O24·4H2O content is 30% of Ti3C2Tx	[139]
NO2	10 ppm	RT	+32.4%	0.1 ppm	Ti3C2Tx/TiO2/SnO2	LiF + HCl, 70 °C, 36 h, US, 30 min, +TiO2/SnO2: aging of SnO2 nanofibers and Ti3C2Tx mixed dispersion, 80 °C, 10 h; mass ratio is 1SnO2-2Ti3C2Tx	[140]
Doping component is an p-type semiconductor metal oxide
Toluene	50 ppm	250 °C	+11.4 *	10 ppm	CuO/Ti3C2Tx (30 wt.%)	HF (40%), RT, 24 h, +CuO: ultrasonic treatment of CuO ethanol dispersion obtained by solvothermal method from copper acetate in anhydrous ethanol medium (150 °C, 12 h), with the addition of freshly prepared Ti3C2Tx powder, RT, 20 min	[141]
Triethylamine	10 pmm	RT, RH = 50%	+181.4	5 ppm	Cu2O/Ti3C2Tx (20 wt.%)	NaF+HCl, 60 °C, 48 h, +Cu2O: ultrasonic treatment of ethanol dispersion containing accordion-like Ti3C2Tx and Cu2O nanospheres derived from CuCl2 in the presence of SDS, NH2OH·HCl and NaOH, RT, 3 h	[143]
Ethanol	50 ppm	200 °C, RH = 30%	+190 *	1 ppm	Co3O4/Ti3C2Tx (2 wt.%)	HF (40%), 35 °C, 24 h, delamination: DMSO, 24 h, US, 1 h, +Co3O4: MOFs self-assembly on the Ti3C2Tx sheet surface out of Co(NO3)3 and 2-methylimidazole methanol solution, ultrasound, RT, 30 min, 24 h, subsequent calcination in air at 350 °C, 1 h	[144]
NOx	100 ppm	RT, RH = 26%	−40 **	0.01 ppm	Co3O4-Al2O3/Ti3C2Tx	LiF, HCl(9M), 35 °C, 48 h, delamination: US, 30 min, Ar, +Co3O4-Al2O3: hydrothermal synthesis of cobalt-aluminum layered double hydroxide on Ti3C2Tx surface from dispersion containing Co(NO3)3, Al(NO3)3, urea, and polyvinylpyrrolidone, 90 °C, 6 h, subsequent heat treatment at 450 °C, 3 h, N2	[145]
Acetone	2.86 ppm	RT, RH = 20%	+250	-	2K2W7O22–1Ti3C2Tx	HF (5 wt.%), hydrothermal treatment, 150 °C, 5 h, US, 1 h +K2W7O22: hydrothermal synthesis of K2W7O22 from Na2WO4·2H2O, oxalic acid, K2SO4, and HCl solution, 225 °C, 24 h, stirring of K2W7O22 + Ti3C2Tx dispersions, 12 h	[146]

* S = R/R₀ [112,113,125,126,141,144]; ** S = R₀/R [119,121,132,138].

4.2. Modification of Ti₃C₂T_x and Ti₂CT_x by Semiconducting Metal Chalcogenides

A Ti₃C₂T_x/MoS₂–based sensor for detecting low NO₂ concentrations at RT was developed by the authors in their study [149]. To obtain this composite, a dispersion of multilayer Ti₃C₂T_x MXene and MoS₂ nanopowder synthesized from MoO₃, KSCN, and HCl was hydrothermally treated [150]; the MoS₂ content varied from 10 to 30 wt.%. SEM and TEM visualizations indicated that the MoS₂ particles were distributed both on the Ti₃C₂T_x surface and in its interlayer space, which was confirmed by the broadening and shift of the (002) reflection towards smaller diffraction angles, from 9.7° to 8.6°. The increase in the interlayer distance of the MXene should have a positive effect on the generated electron transfer during gas detection. The gas-sensing properties of the obtained products were studied for NO₂ in the concentration range of 10–100 ppm at RT (Table 3). It was established [149] that the optimum amount of dopant that advances the composite sensor performance is 20 wt.% MoS₂. The response of Ti₃C₂T_x–20 wt.% MoS₂-structure to 10–100 ppm of NO₂ is observed in the range of 35.8–72.5%, which is still affected by a significant baseline drift. The selectivity of the obtained coatings was justified with respect to methanol, ammonia, and NO₂, at 50 ppm concentration. It was found that the response to NO₂ was 60%, while for the other gases it did not exceed 10%. The authors do not explain the reason for such a high selectivity of the Ti₃C₂T_x/MoS₂-based composite to NO₂, rather giving a conventional mechanism of oxidizing gas detection through the chemical adsorption of oxygen molecules with their further dissociation on the surface as well as physical adsorption of the analyte gas. Due to such processes, dissociated oxygen ions migrate easily from MXene to MoS₂ and act as electron donors for NO₂ binding.

The one-step in situ hydrothermal synthesis of 2D Ti₃C₂T_x@TiO₂@MoS₂ composite was proposed in [151] to be employed for nitrogen dioxide detection in a wide concentration range, 0.023–100 ppm. For this purpose, a hydrothermal synthesis of MoS₂ on the surface of delaminated Ti₃C₂T_x was carried out in the Na₂MoO₄·2H₂O–CN₂H₄S reaction system. XRD data confirmed the formation of a complex Ti₃C₂T_x/TiO₂/MoS₂ composite. TEM micrographs indicated the vertical growth of MoS₂ particles on the surface of partially oxidized MXene, and the EDX analysis demonstrated the uniform distribution of molybdenum sulfide particles without obvious aggregation, which, according to the authors, is prevented by the formed TiO₂-anatase particles. The integral area of the Ti-O peaks in the XPS spectra revealed the degree of Ti₃C₂T_x MXene oxidation to be around 86.7%. The impedance spectroscopy established that among the rows of Ti₃C₂T_x/MoS₂, MoS₂, and Ti₃C₂T_x/TiO₂/MoS₂, the highest conductivity is observed in the latter, which is associated with a high conductivity of MXene and with the presence of a large number of heterojunctions that contribute to the charge transfer. The negative slope of the Mott-Schottky plot determined the p-type conductivity of this nanocomposite. When examining the gas-sensing properties of the obtained coatings, the significant response to 50 ppm of NO₂ (55%), the short response time (1.8 s), and the average recovery time (70 s) were observed, which the authors reasoned were due to (i) the high conductivity of the MXene component of the nanocomposite; (ii) the large number of heterojunctions at the grain boundaries, which accelerate electron transfer throughout the composite; and (iii) the vertical morphology of molybdenum sulfide growth and the introduction of defective structures, achieved by excess urea during synthesis. The experiments on finding the selectivity of the receptor material to NO₂ versus H₂, CO, CH₄, NH₃, H₂S, and NO showed that the Ti₃C₂T_x/TiO₂/MoS₂ nanocomposite exhibited, at high concentrations of 5–50 ppm, a considerably greater response to NO₂; the closest signal corresponds to NO of 7.5% (Figure 33). The material was found to have good reproducibility and long-term stability over 8 weeks in combination with a linear response to a wide range of NO₂ concentrations (R > 0.977).

The authors proposed an electron-hole mechanism of gaseous analyte detection for the composite [151]: when the sensor is exposed to air, O₂ molecules capture electrons from the TiO₂ electron accumulation layer to form chemically adsorbed O_x⁻ oxygen forms. The adsorbed NO₂ electrophilic molecule captures electrons at the nanocomposite surface and reacts with adsorbed oxygen species to form NO₃⁻ particles, whose existence has been proven by IR spectroscopy. In this case, the electric field balance at the heterojunction is disturbed, leading to hole overflow into MoS₂ to reach Fermi level equilibrium and decreasing the width of the heterojunction barrier.

To detect ammonia in a wide concentration range, 10–800 ppm, Tian et al. synthesized the Ti₃C₂T_x/TiO₂/MoS₂ composite material by the hydrothermal method described in [151], with the extra addition of citric acid hydrate to the reaction mixture as a surfactant [152]. Using a set of physicochemical analysis methods, the qualitative composition and morphology of the Ti₃C₂T_x/TiO₂/MoS₂ nanocomposite were studied. MoS₂ nanosheets and a small amount of TiO₂ particles were grown on the surface of the Ti₃C₂T_x MXene. The authors observed that the TiO₂ diffraction peak is significantly weaker in the MoS₂-containing nanocomposite than in the Ti₃C₂T_x/TiO₂. They explain it by the MoS₂ growth on the MXene surface, which inhibits TiO₂ appearance under the hydrothermal synthesis. The Ti₃C₂T_x/TiO₂/MoS₂ nanocomposite was found to have a higher response to NH₃ at RT compared with the pristine components; a detection limit of 500 ppb was defined (Figure 34).

It was noted [152] that the Ti₃C₂T_x MXene has p-type conductivity, while MoS₂ is an n-type semiconductor. The resistance of the Ti₃C₂T_x/TiO₂/MoS₂-based sensor is significantly higher compared with the initial components due to the newly formed p-n-heterojunctions. The high reproducibility within five cycles of the target gas and the long-term stability within 60 days supported the material’s ability to deliver a highly stable sensing signal. The selectivity tests carried out with ammonia, triethylamine, trimethylamine, n-butanol, acetone, and formaldehyde, all of 100 ppm concentration, and nitrogen dioxide, of 800 ppm concentration, supported the Ti₃C₂T_x/TiO₂/MoS₂ nanocomposite’s ability to exhibit an exceptional selectivity to ammonia with a response equal to 173.3%, while that to other reducing gases did not exceed 15%. The response to the NO₂ was 84%, but in another direction, n-type, maturing from the oxidizing nature of this analyte. Experiments on the humidity effect, where it was varied in the range of 43–75%, revealed a typical problem for MXene-based sendors as a drop in response values with increasing RH. The authors explained this by H₂O molecules’ adsorption on the active centers and suggested that the interaction between adsorbed water molecules and oxygen with H⁺ and OH⁻ formation releases electrons into the conduction band of the sensing material, which suppresses the resistance. The ammonia detection for the Ti₃C₂T_x/TiO₂/MoS₂ composite material is described via an electron-hole model by the creation of a surface depletion layer due to atmospheric oxygen adsorption [152]. It was found that MoS₂ acts as the major phase material in the nanocomposite, as it exhibits the same resistance change trend as the pristine molybdenum chalcogenide. The energy band diagram of Ti₃C₂T_x/TiO₂/MoS₂ nanocomposite shows (Figure 35) the energy band changes between Ti₃C₂T_x MXene and MoS₂ when forming heterojunction accompanied by Fermi level alignment, while TiO₂ mainly increases the distance between MXenes’ layers and provides more adsorption centers. The DFT method demonstrated that the ammonia molecules provide more charge transfer at the heterojunction of the nanocomposite than in the case of individual nanocomposite components, which leads to a greater change in resistance.

Chen et al. [153] synthesized a Ti₃C₂T_x/WSe₂ (2 and 4 wt.%) nanohybrid combining the properties of a semiconductor (WSe₂) and MXene with a metallic conductivity. They were electrostatically assembled by mixing the dispersions of delaminated Ti₃C₂T_x and WSe₂ nanoflakes, whose surfaces were modified with a cetyltrimethylammonium cation at elevated temperatures. This Ti₃C₂T_x/WSe₂ composite was studied upon exposure to 1–40 ppm of ethanol at RT. It was found that the 2 wt.% WSe₂-containing nanocomposite exhibits n-type conductivity, has the highest and fastest response to the gas analyte, and has low electrical noise compared with the pristine compounds. In contrast, the individual MXene exhibits p-type conductivity resulting from an increase in resistance when exposed to the analyte gas. The long-term stability and reproducibility of the signal of the composite material over a month when exposed to 40 ppm of ethanol as well as the ultra-fast response (9.7 s) and recovery (6.6 s) at RT were revealed, which the authors ascribe to the presence of multiple heterojunctions in Ti₃C₂T_x/WSe₂. The stability of the receptor material to elevated humidity was studied in the RH range from 5% to 80%: the response to 40 ppm of ethanol at 80% RH was shown to drop by half from −12% (at 5% RH) to −6.1% (at 80% RH). Gas-sensing tests with ethanol, methanol, acetone, hexane, benzene, and toluene, all in a 40 ppm concentration, were carried out to reveal the advantages of the obtained nanohybrid in VOC detection [153]. A Ti₃C₂T_x–2 wt.% WSe₂-based coating was shown to exhibit high selectivity toward oxygen-containing VOCs. The authors do not provide a specific explanation for this effect, relying on the conventional detection mechanism of n-semiconductor materials involving an electron overflow due to surface redox reactions, which does not explain in this case the dominant interaction of oxygen-containing VOCs with the nanohybrid surface compared with hydrocarbon VOCs.

In another study [154], in situ synthesis of vertically oriented SnS nanoflakes has been reported in the form of inflorescence on the surface of conductive Ti₃C₂T_x followed by alkaline treatment with 0.05 M NaOH to modify the MXene surface functional groups. Solvothermal synthesis of SnS/Ti₃C₂T_x composite material was performed in the SnCl₂·2H₂O–thiourea–citric acid–ethylene glycol system at different component ratios and at different alkaline treatment times. Based on SEM and TEM analysis data, the optimal amount of MXene was determined to obtain a sample with a large specific surface area and a uniform distribution of SnS nanoflakes on the Ti₃C₂T_x surface. The authors explain the vertical orientation of sulfide nanoparticles during synthesis by the fact that surface titanium atoms with vacant orbitals tend to be coordinated by electron donor atoms, in this case sulfur. XRD data confirm the formation of SnS/Ti₃C₂T_x nanocomposite; however, it was found that alkaline treatment reduces the crystallinity of the material and increases the number of lattice defects (dislocations and microstrains). According to the XPS analysis, it was found that the alkaline treatment enhances the Sn⁴⁺ percentage from ca. 69.9% (before treatment) to ca. 83.8% (after treatment), which can indirectly indicate the successful modification of the surface with oxygen-containing functional groups. The obtained coatings were studied versus 1−10 ppm of NH₃ at 25 °C and 20 rel.% of humidity [154]. The receptor material was found to exhibit typical n-type semiconductor conductivity based on the decrease in resistance when exposed to NH₃. For the SnS/Ti₃C₂T_x nanocomposite-based sensor, a significant increase in sensitivity compared with pristine SnS and Ti₃C₂T_x was observed due to the larger specific surface area of vertically oriented sulfide particles and high N-S bond energy. In particular, the response value of the sample synthesized by additional alkaline treatment to 250 ppb of NH₃ exceeded that of the sample without such a treatment even to 10 ppm of NH₃. A selectivity test of the obtained coatings allowed the authors to distinguish ammonia: the response to 5 ppm of NH₃ was significantly higher than that to 5 ppm of methanol, ethanol, acetone, isopropanol, NO₂, and CO. Still, increasing the humidity level to 40 rel.% has not affected the gas-sensing performance of the coatings, although with a further rise in humidity, the stability of the signal and its magnitude are notably reduced. This effect was explained, firstly, by a lower effective concentration of NH₃ due to its dissolution in water vapors and, secondly, by a decrease in the number of vacant adsorption centers for the target gas [154].

Thus, the researchers attribute in general the higher sensitivity of gas sensors based on MXene/transition metal chalcogenides nanocomposites to both the appearance of heterojunctions at the interface and the oriented growth of chalcogenides on the MXene sheet surface, which contributes to the specific surface area growth. All the above mechanisms of gas-analyte detection are mainly reduced to the electron-hole model of p- or n-type semiconductors, which demonstrates the predominant role of chalcogenides when examining the sensory properties.

Table 3. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and semiconducting metal chalcogenides (MoS₂, SnS, WSe₂).

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
NO2	50 ppm	RT	60	10 ppm	Ti3C2Tx/MoS2 (20 wt.%)	HF (25%), RT, 48 h, + MoS2: hydrothermal treatment of dispersion containing multilayer Ti3C2Tx MXene and MoS2 synthesized from MoO3, KSCN, and HCl, 170 °C, 8 h	[149]
NO2	50 ppm	RT	55	0.023 ppm	Ti3C2Tx/TiO2/MoS2	LiF + HCl, RT, 48 h, delamination: US, 1.5 h, N2, +TiO2/MoS2: hydrothermal synthesis of MoS2 in the Na2MoO4·2H2O–CN2H4S reaction system on the surface of delaminated Ti3C2Tx, 200 °C, 24 h	[151]
NO2	100 ppm	27 °C, RH = 43%	−173.3	500 ppb	MoS2/Ti3C2Tx/TiO2	LiF + HCl, RT, 48 h, delamination: US, 1.5 h, N2, +TiO2/MoS2: hydrothermal synthesis of MoS2 in the Na2MoO4·2H2O–CN2H4S–citric acid reaction system on the surface of delaminated Ti3C2Tx, 200 °C, 24 h	[152]
Ethanol	40 ppm	RT, RH = 5%	−12	1 ppm	Ti3C2Tx/WSe2 (2 wt.%)	HF (30%), RT, 24 h, + WSe2: electrostatic assembly upon mixing dispersions of delaminated Ti3C2Tx and WSe2 nanoflakes having surface modified with cetyltrimethylammonium cation, 60 °C, 2 h	[153]

Table 3. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and semiconducting metal chalcogenides (MoS₂, SnS, WSe₂).

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
NO2	50 ppm	RT	60	10 ppm	Ti3C2Tx/MoS2 (20 wt.%)	HF (25%), RT, 48 h, + MoS2: hydrothermal treatment of dispersion containing multilayer Ti3C2Tx MXene and MoS2 synthesized from MoO3, KSCN, and HCl, 170 °C, 8 h	[149]
NO2	50 ppm	RT	55	0.023 ppm	Ti3C2Tx/TiO2/MoS2	LiF + HCl, RT, 48 h, delamination: US, 1.5 h, N2, +TiO2/MoS2: hydrothermal synthesis of MoS2 in the Na2MoO4·2H2O–CN2H4S reaction system on the surface of delaminated Ti3C2Tx, 200 °C, 24 h	[151]
NO2	100 ppm	27 °C, RH = 43%	−173.3	500 ppb	MoS2/Ti3C2Tx/TiO2	LiF + HCl, RT, 48 h, delamination: US, 1.5 h, N2, +TiO2/MoS2: hydrothermal synthesis of MoS2 in the Na2MoO4·2H2O–CN2H4S–citric acid reaction system on the surface of delaminated Ti3C2Tx, 200 °C, 24 h	[152]
Ethanol	40 ppm	RT, RH = 5%	−12	1 ppm	Ti3C2Tx/WSe2 (2 wt.%)	HF (30%), RT, 24 h, + WSe2: electrostatic assembly upon mixing dispersions of delaminated Ti3C2Tx and WSe2 nanoflakes having surface modified with cetyltrimethylammonium cation, 60 °C, 2 h	[153]

4.3. Modification of Ti₃C₂T_x and Ti₂CT_x by Carbon Nanomaterials

Modification of semiconductor metal oxide-based receptor materials with carbon nanomaterials is quite common [155,156,157,158,159], primarily to reduce the detection temperature and improve the specific surface area. Studies on the efficiency of different 2D nanomaterial combinations, including MXenes (Ti₃C₂T_x or Ti₂CT_x) and reduced graphene oxide, for gas analyte detection with chemiresistive sensors are much less known, and the introduction of carbon nanotubes into composite materials is currently only covered for liquid sensors [160,161,162,163,164,165]. Table 4 briefly summarizes the data on the sensing properties of the composite materials, which include reduced graphene oxide (rGO) and titanium-containing MXenes.

The authors of the paper [166] attempted to expand the Ti₃C₂T_x MXene band gap by the introduction of rGO with the goal of obtaining hybrid fibers of 1.2 m, excluding metal cations (Ca²⁺, Al³⁺, etc.), containing binders. The fibers were obtained by wet-spinning Ti₃C₂T_x synthesized under the influence of the LiF + HCl system and graphene oxide prepared by a modified Hummers method. For this purpose, MXene and graphene oxide dispersions in dimethylformamide were injected through a needle into a coagulating solution of n-hexane-dichloromethane to produce gel fibers, followed by thermal treatment at 200 °C. According to Raman spectroscopy data, the ratio of D-peak, at 1350 cm⁻¹, corresponding to defective carbon sites, and G-peak, at 1580 cm⁻¹, corresponding to sp²-domains, intensities was found to increase after the nanocomposite heating, confirming the conversion of GO to rGO. XPS analysis revealed that oxygen was transferred from GO to Ti₃C₂T_x MXene under thermal reduction of the composite due to the apparent potential difference. Based on the relationship between the photon energy and the optical absorption coefficient, the authors found that the nanohybrid’s band gap width has expanded from 1.05 eV to 1.57 eV. The sensing performance of the obtained nanocomposite was evaluated for ammonia, acetone, hydrogen sulfide, sulfur dioxide, xylene, and benzene at 50 ppm concentration [166]. It was found that the response to ammonia at RT was ca. 6.8%, while the response to other gases did not exceed 1%. At the same time, the nanocomposite demonstrated a ca. 4.3% response at a relatively low ammonia concentration of 10 ppm (Figure 36). The authors ascribed the ammonia selectivity to the expansion of the MXene band gap width as well as to the significant adsorption energy compared with other gases, which was confirmed by a computational DFT study referring to [87,167]. The desorption rate in this case has not been discussed.

Tran et al. synthesized a reduced graphene oxide-based nanocomposite (rGO/Ti₃C₂T_x) using the conventional route of MAX-phase etching with concentrated hydrofluoric acid and the introduction of hydrazine-reduced graphene oxide under ultrasound exposure in an ice bath [168]. The derived, highly concentrated suspensions of the nanocomposite were cast onto Si substrates by dropping and dried at 60 °C. According to SEM analysis, the rGO flakes in the obtained composite cover most of the MXene sides and even enter the interlayer space; at the same time, local rGO aggregates are also found. TEM data confirmed the conclusion on the close contact of the reduced graphene oxide with the MXene plane, while SAED data indicated the Ti₃C₂T_x hexagonal structure. Based on the XPS data, the qualitative composition of the nanocomposite was confirmed, while a significant decrease in the concentration of fluoride ions on the MXene surface during the rGO introduction was observed. The gas-sensing properties of the composite-based sensor were investigated for 10–100 ppm of NO₂. For instance, a high response of 37% at 50 ppm of NO₂ was found. The sensor exhibited excellent repeatability of signal when exposed to five cycles of NO₂ with no drop in response and complete recovery. NO₂ selectivity was established when exposed to 10% methane and 100 ppm of toluene. Given that the kinetic response curves for these gases were quite poor, the authors report responses equal to 20% in the case of CH₄ and 25% in the case of C₇H₈. The sensitivity to NO₂ has been again explained by the high specific surface area of the nanocomposite, its heavy functionalization, and the efficient process of electron transfer across the rGO/Ti₃C₂T_x boundaries [168].

As already mentioned in Section 4.1.1 [104], the introduction of rGO into the Ti₃C₂T_x/TiO₂ nanocomposite, having MXene as crumpled spheres oxidized under specified conditions with a controlled number of titanium defects and derived by ultrasonic pyrolysis, leads to an increase in the NO₂ response with the detection limit down to 10 ppb at RT. The authors show that the real reason for gas-sensing improvement is the large number of titanium atom defects caused by oxidation and the creation of Ti₃C₂T_x/rGO heterojunctions. DFT calculations were performed to confirm their hypothesis. It was shown that, compared with the original Ti₃C₂T_x, the adsorption energy of Ti₃C₂T_x MXene with Ti–O vacancies is significantly reduced for gas molecules, and the charge transfer goes in a more free manner. Therefore, the gas-sensing characteristics of such materials are significantly improved [104].

Wang et al. proposed the use of a ternary composite based on rGO nanosheets, nitrogen-doped Ti₃C₂T_x MXene (N-MXene), and titanium dioxide for formaldehyde detection at RT [169]. Nitrogen doping of Ti₃C₂T_x MXene synthesized with LiF/HCl was carried out by a hydrothermal method using urea as a nitrogen source at 180 °C. Under similar conditions, doping of the obtained N-MXene with titanium dioxide was carried out by a hydrothermal method using titanium tetrabutoxide and KCl solution. The introduction of rGO into the N-MXene/TiO₂ system was performed by ultrasonic treatment of the dispersion containing N-MXene/TiO₂ and rGO under a nitrogen atmosphere. SEM studies of the N-MXene/TiO₂/rGO showed that the sample has a specific layered structure with a small TiO₂ content, mainly at the edge areas. Surface functionalization was studied by IR spectroscopy. It was found that OH group-related valence vibrations of adsorbed water at 3420 cm⁻¹, terminal O–H groups at 1390 cm⁻¹, –NH groups at 3120 cm⁻¹, and Ti–O groups at 665 cm⁻¹ were detected. XRD data showed a shift of the diffraction peak from the (002) plane to low angles, from 9° to 6.7°, after N-doping of MXene, confirming that nitrogen atoms contribute to the increased interlayer distance. Under gas-sensing tests of the obtained material, it was established that there was no response to 20 ppm of formaldehyde at 20 °C in almost dry air (RH = 2.1%). However, with increasing the humidity level to the RH = 37–62% range, the response could be recorded with a low noise level [169]. The optimum RH value was found to be 54% because of the fast response and complete recovery of the sensor resistance when the target gas was removed from the cell. The average sensor response was 132% and 26%, or 20 ppm and 4 ppm of HCHO, respectively. The N-MXene/TiO₂/rGO-based sensor showed a high selectivity to formaldehyde when compared with sulfur dioxide, carbon dioxide, ethanol, acetone, and benzene, all of which appeared at 20 ppm concentration, and 4 ppm of hydrogen sulfide; the response to competitive gases did not exceed 3.5%. The authors proposed a mechanism for increasing sensitivity to formaldehyde with increasing atmospheric humidity [169]. According to their concepts, HCHO vapors easily dissolve in water to form hydrated formaldehyde, HCHO∙H₂O. The resulting hydrate weakens the dissociation of adsorbed water on the TiO₂ surface. These processes lead to resistance rising because, according to the authors, the dissociation of H₂O molecules on the metal oxide surface produces ionic-type conductive behavior. The hydration reaction is reversible due to the long and weak hydrogen bonds in HCHO∙H₂O [169].

A synthesis method of 3D Ti₃C₂T_x/rGO/SnO₂ aerogel as a formaldehyde detector was developed by the authors in [170]. The given nanocomposite was obtained by a hydrothermal method using freshly prepared Ti₃C₂T_x under hydrofluoric acid etching and GO produced by Hummer’s method with the addition of SnCl₄∙2H₂O, NH₄F, urea, and ammonium citrate. SEM and TEM images indicated the composite’s 3D structure, where Ti₃C₂T_x and rGO are represented by linked nanosheets, while SnO₂ appears as spheres distributed on top of the sheets. The resulting material is characterized by an extremely high specific surface area of 103.2 m²/g, as confirmed by BET data. The sensor based on synthesized Ti₃C₂T_x /rGO/SnO₂ was studied after exposure to 10 ppm of formaldehyde [168]. Nanocomposite was found to be an n-type semiconductor and exhibited a high response/recovery rate of 2.9 s and 2.2 s, which seems to be a sequence of a large number of surface-active centers accelerating the adsorption/desorption of the target gas. As the concentration of formaldehyde grew from 10 ppm to 200 ppm, the sensor response increased from ca. 55% to ca. 275%. The authors also compared the response to formaldehyde, which was found to be superior to those toward ethanol, methanol, acetone, ammonia, and toluene at the same concentration of 10 ppm. This sensor performance was explained by the high conductivity of the composite and the lowest bond dissociation energy of formaldehyde, 364 kJ/mol, in the series of gases studied, which contributes to a rapid reaction with chemisorbed oxygen on the surface. It was demonstrated that humidity vapors ranging from 20 rel.% to 60 rel.% have no significant effect on the response, but further increasing to 85 rel.% causes the response to drop by a quarter. To explain the observed sensor, the authors followed a typical detection mechanism in metal oxides dealing with redox reactions between the analyte gas and chemisorbed oxygen on the surface of a 3D composite and charge transfer into the EDL of a semiconductor. Again, the p-n-heterojunctions in this composite play a positive role in facilitating charge carrier separation and expansion of the EDL width. The high selectivity towards formaldehyde was supported by DFT calculations. It was found that the composite has the strongest affinity for HCHO [170].

Using a procedure similar to [170], a 3D Ti₃C₂T_x/rGO/CuO aerogel was created [171] by one-step hydrothermal treatment of a solution containing copper chloride, graphene oxide, MXene, ammonium fluoride, urea, and sodium citrate, followed by freeze-drying. The aerogel was mixed with ethanol to form a homogeneous paste that was applied with a brush to an Al₂O₃ substrate equipped with Au electrodes. SEM images revealed that Ti₃C₂T_x/rGO/CuO consists of a porous network similar to the original rGO with the copper oxide nanoparticles evenly distributed on its surface, while MXene sheets are placed on the rGO sheets. XRD patterns confirmed the presence of CuO and rGO plane reflexes only, which the authors attribute to an excessively low MXene concentration. Gas-sensing tests of the obtained 3D aerogel coatings showed that the maximum response value was achieved at 25 °C. High selectivity to acetone, with a response to 100 ppm concentration equal to ca. 50.9%, was argued because the response to other test gases, including benzene, thymethylamine, methanol, toluene, and triethylamine, did not exceed 15%. The authors showed that humidity significantly affected the sensor characteristics of the composite: the sensor response went gradually down from 52% to 43% with an increase in humidity from 20 rel.% to 60 rel.%. This effect is associated with the screening of surfaces by H₂O molecules. The mechanism of acetone detection proposed by the authors (Figure 37) is fully based on the typical concepts characteristic of p-semiconductors, such as CuO. The improved gas-sensing performance is justified by a p-p-heterojunction at the rGO-CuO interface and the presence of Schottky barriers at the Ti₃C₂T_x-CuO interface. Furthermore, the high response of the aerogel is explained by its large specific surface area.

Zhou et al. reported the creation of a chemiresistive gas sensor based on N-Ti₃C₂T_x/PEI/rGO polymer composite, where PEI is polyethyleneimine, to detect CO₂ in a wide range of concentrations, 4–3000 ppm at RT [172]. The methodology for the synthesis of MXene and its nitrogen doping is similar to that described in [169]. The synthesis of the composite was performed by mixing N-Ti₃C₂T_x, PEI solution, and rGO under ultrasound for 2 h. SEM images of the obtained products indicated that the nitrogen-doped MXene has a laminar structure wrapped in PEI with a small number of TiO₂ particles located predominantly at its boundaries. XRD patterns displayed a rather weak (110) reflection of Ti₃C₂T_x, which the authors attributed to the covering effect of the amorphous PEI. A high degree of interaction between the polymer and N-MXene was confirmed by a strong electrostatic interaction due to different zeta potentials (PEI: +3.95 mV; N-Ti₃C₂T_x: −16.7 mV). It was found that the obtained composite film had a maximum response value with a full recovery at 62 rel.% of humidity under RT conditions; when detecting 600 ppm of CO₂, the response reached almost 9% at RH = 60%. When measuring the sensor’s response to input of other gases (SO₂, H₂S, HCHO, CO, NH₃), it was confirmed to have a high selectivity to CO₂. The authors argued that the large number of amino groups in PEI provides a stable and strong interaction with CO₂ molecules, especially in a humid atmosphere. The proposed sensor demonstrated a superior lower limit of CO₂ detection, equal to 8 ppm, than its commercial infrared counterpart, whose detection limit is 50 ppm. The high response in a humid atmosphere and the high selectivity of the N-Ti₃C₂Tx/PEI/rGO-based coating are further attributed to the proton conductivity of the composite [172]. The primary and secondary amino groups of PEI promote the acid-base reaction with the formation of carbamates, while the tertiary amines supply the CO₂ dissolution with carbonic acid formation through a base-catalyzed hydration. This feature yields a faster recovery of CO₂ molecules than one in the acid-base reaction due to the branched PEI structure, which inhibits the acid-base reaction due to the steric effect. In a humid atmosphere, the hydrophilic N-doped MXene is covered with H₂O molecules to protonate the polymer, which, as a result, facilitates a higher response rate. At the same time, rGO promotes charge transfer and accumulation. Following CO₂ exposure, the amount of free amino groups at the composite is reduced due to the noted reactions, which support lower proton mobility, resulting finally in a significant rise in resistance [172].

Summing up, it can be noted that the researchers introduced carbon materials into MXene to target the following issues:

(1)

To increase the specific surface area, which directly affects the amount of sorbed analyte gas;

(2)

To create p-p- and p-n-heterojunctions, which contribute to a more complete charge separation and hence a higher response;

(3)

To create defects in the MXene structure under controlled oxidation conditions that promote adsorption centers for analytes;

(4)

To change the prevailing conductivity type from p- or n-type to ionic with the introduction of polymer matrices.

Many authors provide DFT calculations to explain the high selectivity for a particular analyte, which confirm the adequate adsorption and desorption energies. However, they do not emphasize or take into account that simulations are performed in ideal conditions without chemisorbed oxygen on the sensor surface, which casts doubt on this method of confirming gas selectivity mechanisms, especially for the works that employ metal oxide semiconductors as dopants. Therefore, a set of synthetic and research methods for the study of the influence factors on the chemoresistance properties of composite sensors based on MXene and carbon materials needs further improvement and extension.

Table 4. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and reduced graphene oxide.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
Ammonia	50 ppm	RT	+6.77	10 ppm	rGO/Ti3C2Tx (40 wt.%)	LiF + HCl, 50 °C, 24 h, +rGO: wet-spinning of composite fibers from dispersion of Ti3C2Tx and GO in DMF with subsequent heat treatment, 200 °C, Ar	[166]
NO2	50 ppm	RT	+37	10 ppm	rGO/Ti3C2Tx (33 wt.%)	HF (50%), 50 °C, 24 h, +rGO: ultrasonic treatment of ethanol dispersion containing Ti3C2Tx and rGO (hydrazine reduction), 0 °C, 40 min	[168]
NO2	5 ppm	RT	−19.854	10 ppb	Ti3C2Tx/TiO2/rGO	Mixing of Ti3C2Tx and GO dispersions (mass ratio 2:1)	[104]
Formaldehyde	50 ppm	RT, RH = 54%	+325	2 ppm	N-Ti3C2Tx/TiO2/rGO	LiF + HCl (6 M), RT, 2 h, nitrogen doping: hydrothermal treatment of ethanol dispersion with urea addition, 180 °C, 24 h, + TiO2/rGO: (1) hydrothermal treatment of «N-Ti3C2Tx–Ti(OC4H9)4–KCl–C2H5OH–H2O» system, 180 °C, 24 h, (2) Ultrasonic treatment of suspension containing N-Ti3C2Tx/TiO2 and rGO, N2	[169]
Formaldehyde	200 ppm	RT, RH < 30%	−275	10 ppm	Ti3C2Tx/rGO/SnO2	HF (10%), RT, 6 h, + rGO/SnO2: hydrothermal treatment of suspension containing GO, Ti3C2Tx, SnCl4∙2H2O, NH4F, urea, and ammonium citrate, 140 °C, 24 h, washing with HCldil, freezing	[170]
Acetone	200 ppm	25°C, RH~20%	+50.9	10 ppm	Ti3C2Tx/rGO/CuO	HF (10%), RT, 6 h, delamination: DMSO, US, + rGO/CuO: hydrothermal treatment of suspension containing GO, Ti3C2Tx, CuCl2∙2H2O, NH4F, urea, and ammonium citrate, 140 °C, 24 h, washing with HCldil, freezing, heat treatment, 300 °C, 2 h, Ar	[171]
CO2	600 ppm	RT, RH = 60%	+9	8 ppm	N-Ti3C2Tx/PEI/rGO	LiF + HCl (6 M), RT, 2 h, nitrogen doping: hydrothermal treatment of ethanol dispersion with urea addition, 180 °C, 24 h, +PEI/rGO: ultrasonic treatment of an ethanol suspension containing N-Ti3C2Tx, rGO (prepared by glucose reduction in alkaline medium), and PEI, 2 h, N2	[172]

Table 4. Sensory characteristics of nanocomposite coatings containing Ti₃C₂T_x/Ti₂CT_x MXenes and reduced graphene oxide.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
Ammonia	50 ppm	RT	+6.77	10 ppm	rGO/Ti3C2Tx (40 wt.%)	LiF + HCl, 50 °C, 24 h, +rGO: wet-spinning of composite fibers from dispersion of Ti3C2Tx and GO in DMF with subsequent heat treatment, 200 °C, Ar	[166]
NO2	50 ppm	RT	+37	10 ppm	rGO/Ti3C2Tx (33 wt.%)	HF (50%), 50 °C, 24 h, +rGO: ultrasonic treatment of ethanol dispersion containing Ti3C2Tx and rGO (hydrazine reduction), 0 °C, 40 min	[168]
NO2	5 ppm	RT	−19.854	10 ppb	Ti3C2Tx/TiO2/rGO	Mixing of Ti3C2Tx and GO dispersions (mass ratio 2:1)	[104]
Formaldehyde	50 ppm	RT, RH = 54%	+325	2 ppm	N-Ti3C2Tx/TiO2/rGO	LiF + HCl (6 M), RT, 2 h, nitrogen doping: hydrothermal treatment of ethanol dispersion with urea addition, 180 °C, 24 h, + TiO2/rGO: (1) hydrothermal treatment of «N-Ti3C2Tx–Ti(OC4H9)4–KCl–C2H5OH–H2O» system, 180 °C, 24 h, (2) Ultrasonic treatment of suspension containing N-Ti3C2Tx/TiO2 and rGO, N2	[169]
Formaldehyde	200 ppm	RT, RH < 30%	−275	10 ppm	Ti3C2Tx/rGO/SnO2	HF (10%), RT, 6 h, + rGO/SnO2: hydrothermal treatment of suspension containing GO, Ti3C2Tx, SnCl4∙2H2O, NH4F, urea, and ammonium citrate, 140 °C, 24 h, washing with HCldil, freezing	[170]
Acetone	200 ppm	25°C, RH~20%	+50.9	10 ppm	Ti3C2Tx/rGO/CuO	HF (10%), RT, 6 h, delamination: DMSO, US, + rGO/CuO: hydrothermal treatment of suspension containing GO, Ti3C2Tx, CuCl2∙2H2O, NH4F, urea, and ammonium citrate, 140 °C, 24 h, washing with HCldil, freezing, heat treatment, 300 °C, 2 h, Ar	[171]
CO2	600 ppm	RT, RH = 60%	+9	8 ppm	N-Ti3C2Tx/PEI/rGO	LiF + HCl (6 M), RT, 2 h, nitrogen doping: hydrothermal treatment of ethanol dispersion with urea addition, 180 °C, 24 h, +PEI/rGO: ultrasonic treatment of an ethanol suspension containing N-Ti3C2Tx, rGO (prepared by glucose reduction in alkaline medium), and PEI, 2 h, N2	[172]

4.4. Modifying Ti₃C₂T_x and Ti₂CT_x with Noble Metal Nanoparticles

The decoration of receptor materials with noble metal nanoparticles is an effective method for the sensitization of semiconducting metal oxides [173]. There is a series of works on obtaining composite materials for chemiresistive gas sensors containing Ti₃C₂T_x MXene whose surface has been decorated with highly dispersed Pt, Pd, Ag, and Au particles. Table 5 briefly summarizes the sensory properties of these materials.

Thus, Zhu and colleagues [174] considered the possibility of designing flexible and light-weight electronic devices, including high-performance H₂ sensors operating at RT with the help of Ti₃C₂T_x, MXene, and Pd nanoclusters. For this purpose, a PVP (polyvinylpyrrolidone)-stabilized suspension of palladium clusters was introduced into the dispersion of layered Ti₃C₂T_x, which enabled the metal particles to fix on the MXene surface. With support from the EDX method, the mass fraction of Pd in Ti₃C₂T_x/Pd films was found to be ca. 9.4%, while cross-sectional mapping ensured C and O atoms would distribute throughout the material volume but Ti, F, and Pd atoms would appear mainly on the surface. The increase in the MXene interlayer distance, as shown by the shift of (002) reflection for the Ti₃C₂T_x phase from 6.5° to 5.5° on the composite XRD pattern, indicates the introduction of Pd nanoclusters between the MXene planes. The XPS spectra proved that the palladium is partially oxidized, which is not a hindrance to the detection of hydrogen because the latter has strong reducing properties. The sensory properties of Ti₃C₂T_x/Pd coatings were studied upon exposure to 0.5–40% of hydrogen in air [174]. An optimal material composition with maximum hydrogen sensitivity was found, while the pristine Ti₃C₂T_x layer yielded almost no response to H₂. It was revealed that the output current of the Ti₃C₂T_x /Pd composite increased under hydrogen interaction to be equivalent to a decrease in resistance, indicating an n-type electronic conductivity. The responses to 0.5–40% of hydrogen were ca. 7.4–41.7%. The response and recovery times of the sensors were found to show a tendency to rise with increasing H₂ concentration. The Ti₃C₂T_x/Pd film exhibited excellent cyclic stability and reproducibility. The gas-sensing properties of the Ti₃C₂T_x/Pd film placed on the flexible substrate were also investigated upon mechanical bending at various angles (θ = 30°, 60°, 90°, 120°), and after successive “bending-straightening” cycles from 0° to 180°. The sensor displayed good stability in terms of response and response/recovery time. Furthermore, the authors supported the long-term stability of the sensitive coating via its measuring for 20 days: the response value and response time to 4% of H₂ were almost the same as when the experiment for analyte detection began. It is worth noting that the sensor showed an insignificant response to gases other than hydrogen, including methanol (1000 ppm), acetone (1000 ppm), and ammonia (500 ppm). These tests proved the selectivity of such a material for H₂. The authors tried to explain the Ti₃C₂T_x/Pd resistance drop upon H₂ exposure (Figure 38) and assigned this mechanism to the formation of palladium hydrides (PdH_x) during hydrogen adsorption, which have a lower work function of approx. 3.2 eV than that of metallic Pd, of approx. 5.3 eV, resulting in a reverse electron transfer from palladium hydride clusters to the MXene [174]. The reverse process occurs during H₂ desorption. The authors assembled a hydrogen leak detection system with a Ti₃C₂T_x/Pd film and LED lamp. The stable operation of the lamp when exposed to 4% H₂ for tens of seconds confirms the successful H₂ leak alarm.

Phuong Doan et al. [175] tried to solve two tasks at once: to develop a highly sensitive and selective hydrogen sensor and a solid-state hydrogen accumulator. For this purpose, a Na₂PdCl₄ solution with polyvinylpyrrolidone was mixed with an ethylene glycol suspension of multilayer Ti₃C₂T_x MXene obtained by Ti₃AlC₂ MAX-phase etching with concentrated hydrofluoric acid. Then, the Pd nanoparticles were synthesized via heating the resulting reaction system at 160 °C for 2 h. According to the XPS analysis, it was found that the surface of the multilayer MXene after the palladium decoration was functionalized with –OH or –O groups. Based on SEM data, the authors identified the best sample in terms of density and palladium nanoparticle distribution, suitable for gas-sensing studies. For the selected Ti₃C₂T_x/Pd sample, satisfactory responses to 10–100 ppm of H₂ are noted; in particular, the chemiresistive response to 100 ppm of H₂ was 56%.

The sensor is reported to demonstrate distinct responses even after 90 days of storage in air. Selectivity was tested against 100 ppm of methane, ammonia, and nitrogen dioxide. The response to all these analytes did not exceed 5%. It should be noted that no baseline recovery was demonstrated after NO₂ exposure. The authors explain the good sensor selectivity versus H₂ by the creation of palladium hydrides, whose particles can contribute to hydrogen atoms’ delivery to the neighboring MXene planes through the spill-over mechanism with the formation of TiH₂. Meanwhile, such a functionalization leads to increased sample resistance. This hydrogen transfer is also enhanced by surface oxygen-containing groups. When discussing this issue, the authors refer to the data in [176] indicating that hydrogen atom diffusion is intensified when the concentration of oxygen groups grows.

The purpose of another work [177] was to study the gas-sensing characteristics of VOC sensors based on field-effect transistors employing Ti₃C₂T_x/Pt composite as receptor material. To decorate MXene planes with single platinum atoms, a solution of H₂PtCl₆·6H₂O was added to its dispersion, followed by stirring the reaction system in an Ar atmosphere for 30 min. The authors point out that the process of MXene delamination supports the appearance of titanium vacancies on the Ti₃C₂T_x nanosheet surface, which serve as active centers for anchoring single Pt atoms due to their strong ability to capture alternative atoms and good reducing potential. It was shown that the Ti₃C₂T_x/Pt sample obtained by such a method had a high response to 10 ppm of triethylamine. However, rather long response and recovery times were observed, suggesting a strong interaction between the analyte molecules and the material surface. Tests revealed that the sensor resistor baseline shifts to higher values by 55% and 172% after 20 days of exposure to 20 rel.% and 80 rel.% of surrounding humidity, and the response goes down by 60% and 89%, respectively (Figure 39). On the contrary, exposure to dry air (RH < 5%) almost does not change the resistance, which indicates that the coating is stable enough in a low-humidity environment. The selectivity was studied with respect to 10 ppm of analytes of different chemical natures, including acetone, ethanol, formaldehyde, benzene, hydrogen sulfide, ammonia, carbon monoxide, and dioxide. It is demonstrated that the triethylamine response is three times higher when compared with that of other gases. Overall, following the results given in [177], it can be stated that single Pt atom doping enhances the response of Ti₃C₂T_x MXene to triethylamine vapor. To explain the detection mechanism, the authors assume that triethylamine is oxidized at the material surface while interacting with surface-adsorbed oxygen. As a result, additional electrons are released, according to Reaction (26). Because the Mxene is considered a p-type conductor, these surface reactions lead to a lower carrier density of holes and reduced conductivity.

2(C₂H₅)₃N + 43 O⁻_(ads) → 12CO₂ + 2NO₂ + 15H₂O + 43 ē

(26)

The enhancement of the response under introducing single Pt atoms matures from, according to the authors, the effect of chemical sensitization and advancing the adsorption of analyte/oxygen molecules.

The study by Xu and colleagues [178] was aimed at designing a hydrogen sulfide sensor capable of operating under real humidity conditions with the help of delaminated Ti₃C₂T_x modified with Ag nanoparticles. To synthesize the Ti₃C₂T_x/Ag composite, a silver nitrate solution was introduced into the few-layer MXene dispersion, followed by ultrasonic treatment. TEM analysis revealed that silver nanoparticles have an average size in the range of 28.2–30.8 nm. The XRD data showed that the decoration of MXene by silver nanoparticles causes an obvious oxidation of the MXene surface; the shift of the (002) plane reflection from 7.4° to 6.3° indicates widening the interlayer space. D and G bands of graphitized carbon have been found in the Raman spectra. Measurement of the CVC (current-voltage characteristics) showed that higher additions of silver result in a reduction of the sample’s conductivity. The authors suggested that this phenomenon belonged to the oxidation of MXene with the reduction of silver on its surface. All samples showed p-type semiconductor characteristics. Studies with exposure to hydrogen sulfide revealed that the most sensitive sample of Ti₃C₂T_x/Ag yields the chemiresistive response to 10 ppm of H₂S equal to 60%, which is more than 10 times higher than that of undecorated MXene [178]. The study of selectivity for 1 ppm of NO₂, SO₂, H₂, acetone, ethanol, CO, ammonia, and formaldehyde showed that the Ag addition promotes the detection of all the gases under test. The authors hypothesized that the water molecules being adsorbed on the surface would reduce the number of charge carriers at the MXene surface, which would additionally cause the response to advance [178]. The observed resistivity growth in the detection of all test analytes, or p-type response, is explained by the fact that (1) gas adsorption prevents a charge carrier transfer and reduces carrier density, and (2) gas adsorption might increase the interlayer distance and thereby prevent an interlayer electron transfer. The enhanced response of the Ti₃C₂T_x/Ag sample to hydrogen sulfide is due primarily to the chemical sensitization of MXene by silver nanoparticles and the reduction of partially oxidized silver with the release of electrons, their recombination with holes, and a corresponding increase in resistance.

Wu et al. [179] investigated the effect of Ti₃C₂T_x decoration by silver on its gas sensitivity towards ammonia. The synthesis of Ti₃C₂T_x/Ag composite was performed by the self-reduction of AgNO₃ in the solvent over the reducing MXene surface via the introduction of the silver nitrate solution into the dispersion of multilayer MXene, to be followed by stirring for 30 min. The sensors were printed using an electrohydrodynamic jet printer with Ti₃C₂T_x/Ag-based functional inks. It was found that the Ti₃C₂T_x/Ag and multilayer Ti₃C₂T_x sensor materials were more sensitive to NH₃ after tetramethylammonium hydroxide intercalation than the few-layer MXene-based sensors. The authors explain this with three reasons [179]:

−

XPS and IR results showed that these materials are characterized by a large number of defects considered to be effective adsorption centers;

−

The increased content of intercalated water, which has an essential role in ammonia adsorption due to its strong affinity, was proven experimentally for these samples and will be further confirmed with the help of DFT calculations;

−

These materials had a hierarchical, accordion-like structure, which is characterized by a high porosity and a specific surface area.

A gold nanoparticle-decorated multilayer Ti₃C₂T_x MXene was considered by the authors in their study [180] as a receptor material for the detection of low formaldehyde concentrations. In this case, the Ti₃C₂T_x sample synthesized with a mixture of lithium fluoride and hydrochloric acid was ultrasonically treated and then mixed with HAuCl₄·3H₂O for the Au³⁺ reduction reaction and the formation of nanoclusters on the MXene surface. SEM images revealed that Au nanoparticles, about 30 nm in diameter, were uniformly distributed on the Ti₃C₂T_x sheets. The exceptional Ti₃C₂T_x-Au selectivity towards oxygen-containing organic compounds (ethanol, formaldehyde, and acetone) compared with hydrocarbon and aromatic analytes (toluene, hexane, and ethylene) was established and can be explained by dipole scattering from methyl groups that reduces the overall electron transfer and decreases the response. Among all oxygen-containing analytes, formaldehyde yielded the highest chemiresistive response; in particular, the response to 20 ppm of its concentration was 3%. The response to 200 ppm of water vapor, equal to 1.28%, was found to be significantly lower compared with that to 20 ppm of HCHO, indicating quite selective formaldehyde detection.

In summary, it can be concluded that the decoration of MXenes with noble metal nanoparticles/clusters considerably advances the sensitivity to the analyte gases, for which the latter exhibit a high catalytic activity. Most researchers connect this fact with: (1) the emerging spill-over effect; (2) increasing the number of adsorption centers for oxygen molecules and analyte gases; and (3) intercalating noble metal atoms between MXene sheets that promote an inter-layer electron transfer. Most researchers also decorated the MXene surface with noble metal atoms in situ, which additionally caused a reduction in conductivity as a result of MXene surface partial oxidation accompanying the appearance of heterojunctions. However, this aspect of the detection mechanism was not considered in any of the given articles. Thus, the understanding of the synergistic effect caused by the noble metal introduction into the MXene-based composite structure and contributing to the improved sensing characteristics of their sensors is at an early stage and needs further exploration.

Table 5. Sensory characteristics of coatings based on nanocomposites containing Ti₃C₂T_x/Ti₂CT_x MXenes and noble metal nanoparticles.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
H2	0.5%	RT	−7.4 *	0.5%	Ti2CTx/Pd	LiF + HCl (9M), 40 °C, 24 h, delamination: US, 30 min; +Pd: mixing dispersions of MXene and Pd nanoclusters, separation by filtration	[174]
H2	100 ppm	RT	+56	10 ppm	Ti3C2Tx/Pd	HF (50%), 50 °C, 24 h, +Pd: ML-Ti3C2Tx suspension in ethylene glycol was mixed with Na2PdCl4 solution with polyvinylpyrrolidone, heating at 160 °C, 2 h	[175]
Triethylamine	10 ppm	RT	+6 *	0.05 ppm	Ti3C2Tx/Pt	+Pt: H2PtCl6·6H2O solution was added to the few-layer Ti3C2Tx MXene dispersion, stirring for 30 min, Ar	[177]
H2S	1 ppm	RT	+6 *	50 ppb	Ti3C2Tx/Ag	+Ag: silver nitrate solution was introduced into the few-layer Ti3C2Tx dispersion, followed by ultrasonic treatment	[178]
Formaldehyde	20 ppm	RT	+3	1 ppm	Ti3C2Tx/Au	LiF + HCl (9M), RT, 24 h, delamination: US, 1 h; +Au: HAuCl4 solution was added to the few-layer Ti3C2Tx dispersion, stirring	[180]

* S = (I − I₀)/I₀ × 100 (%) [174,177,178].

Table 5. Sensory characteristics of coatings based on nanocomposites containing Ti₃C₂T_x/Ti₂CT_x MXenes and noble metal nanoparticles.

Gas Analyte	Concentration	Detection Conditions	Response (ΔR·100%/R0), %	Detection Limit	Receptor Material	Synthetic Features	Ref.
H2	0.5%	RT	−7.4 *	0.5%	Ti2CTx/Pd	LiF + HCl (9M), 40 °C, 24 h, delamination: US, 30 min; +Pd: mixing dispersions of MXene and Pd nanoclusters, separation by filtration	[174]
H2	100 ppm	RT	+56	10 ppm	Ti3C2Tx/Pd	HF (50%), 50 °C, 24 h, +Pd: ML-Ti3C2Tx suspension in ethylene glycol was mixed with Na2PdCl4 solution with polyvinylpyrrolidone, heating at 160 °C, 2 h	[175]
Triethylamine	10 ppm	RT	+6 *	0.05 ppm	Ti3C2Tx/Pt	+Pt: H2PtCl6·6H2O solution was added to the few-layer Ti3C2Tx MXene dispersion, stirring for 30 min, Ar	[177]
H2S	1 ppm	RT	+6 *	50 ppb	Ti3C2Tx/Ag	+Ag: silver nitrate solution was introduced into the few-layer Ti3C2Tx dispersion, followed by ultrasonic treatment	[178]
Formaldehyde	20 ppm	RT	+3	1 ppm	Ti3C2Tx/Au	LiF + HCl (9M), RT, 24 h, delamination: US, 1 h; +Au: HAuCl4 solution was added to the few-layer Ti3C2Tx dispersion, stirring	[180]

* S = (I − I₀)/I₀ × 100 (%) [174,177,178].

4.5. Modifying Ti₃C₂T_x and Ti₂CT_x with Polymers

The development of composite materials containing MXenes and polymers is quite common, primarily for obtaining materials for liquid sensors. This is justified by the necessity to promote porosity and to design hierarchical structures, which are extremely important for MXene-based materials. These issues were thoroughly discussed in another review [181]. In addition, the introduction of polymer components leads to high elasticity and durability of the MXene-based layers, which are required for their practical implementation in flexible Internet-of-Things units. The presence of certain functional groups, acidic or basic, on the polymer surface improves the selectivity of the sensors for specific gaseous analytes.

The number of studies on the fabrication and analysis of chemiresistive gas sensors employing MXenes and polymers is not high. Most of them are mainly aimed at detecting ammonia [182,183,184,185,186,187], methanol [188], and NO_x [189,190]. Almost all of the composites under study were developed with conductive polymers, such as polyaniline [182,183], PEDOT:PSS [184,185,188], polypyrrole [186], and polyethyleneimine [172,190].

Thus, a high-performance ammonia sensor based on a composite of hollow polyaniline nanospheres with Ti₃C₂T_x (2–25 wt%) MXene, which was deposited directly on a flexible PET substrate during polymerization, is reported in [182]. The response, ΔR/R₀, to 10 ppm NH₃ of the PANI/Ti₃C₂T_x composite with an optimum MXene content of 15 wt.% was found to be 3.7 at room temperature. The degradation of the material under a 20° cyclic bending, which leads to a decrease in the response value, is only noticeable after 500 cycles.

In the paper of Li et al. [183], a flexible hybrid ammonia sensor was developed using the PANI/Ti₃C₂T_x composite for agricultural tasks. It is shown that the material exhibits an appropriate performance even at a relative humidity of 20–80% in the practically important temperature range of 10–40 °C. The authors gave an explanation of the high gas-sensing properties due to the formation of the Schottky barrier in the material under study and the improvement of the polymer protonation degree.

For the PEDOT:PSS/Ti₃C₂T_x composite, the sensitivity to ammonia [184,185] and to such an important VOC as methanol [188] was also observed. Nitrogen doping and partial MXene oxidation were found to increase the interlayer distance and consequently improve the NH₃ detection kinetic parameters of the 100 wt.% N-Ti₃C₂T_x–50 wt.% PEDOT:PSS composite [184]. At a mass ratio of PEDOT:PSS/Ti₃C₂T_x equal to 4:1, a high sensitivity of the hybrid material at room temperature to 300 ppm methanol was also shown as a result of the increased interlayer distance in MXene; for comparison, the response to 300 ppm acetone and ethanol is more than 5 times lower.

For the sensor based on a composite of polypyrrole (PPy) and Ti₃C₂T_x, derived by polymerization in the presence of MXene plates, it is also reported to have enhanced sensitivity to ammonia, making it promising for implementation in the Internet of Things (IoT) [186]. The authors attribute this primarily to advancing the specific surface area, the high electron mobility of MXene, and the formation of hydrogen bonding between the polymer and Ti₃C₂T_x functional groups.

Thus, the use of composite materials based on Ti₃C₂T_x MXene with conductive polymers allowed one to enhance the observed responses not only by making materials with a developed surface, which provides a higher number of adsorption centers and facilitates gas access to the MXene surface, but also by generating Schottky barriers at the components’ interfaces. Moreover, a number of experiments have shown a high influence of ionic conductivity on the detection mechanism in a humid atmosphere, which is characteristic for electrically conductive polymers. This can be extremely promising for the manufacturing of sensors aimed at detecting diseases taking place in vivo through exhaled breath because this environment is highly enriched with humidity vapors.

5. Conclusions

Analysis of publications on the design of chemiresistive sensor materials based on carbide MXene phases, Ti₃C₂T_x and Ti₂CT_x, shows a high potential of these materials for sensors operated at RT and indicates extensive R&D efforts applied by the research community in this field to level the unfavorable properties of MXenes as receptor materials. Among others, the low stability against oxidation and the extreme sensitivity to H₂O vapors can be partially compensated in these MXene structures by a chemical modification of their surface with bulk organic fragments, often of a hydrophobic nature such as sodium ascorbate, trimethylacetic acid anhydride, or perfluoroalkyl silane. In addition to creating steric hindrances for the diffusion of water molecules to the MXene surface, this also allows one to expand the interlayer space, i.e., to enlarge (i) the diffusion space for analyte gases, (ii) the specific surface area, and (iii) the number of active adsorption centers, and consequently to improve the response and kinetic characteristics. Doping MXenes with heteroatoms, such as nitrogen or sulfur, as well as replacing functional groups with –Cl, –Se, etc., also provides an opportunity to extend the distance between MXene 2D layers, which gives some options for detection selectivity because of the geometry factor.

The approaches used to develop MXene composites with carbon nanomaterials, polymers, noble metal nanoparticles, and especially semiconducting metal oxides are shown to be very effective. In these hybrid materials, the operation temperature is reduced almost to RT. That is a drastic difference from, say, metal oxide-based gas sensors, which are currently the most commercialized ones. Moreover, the response value and response/recovery times of pristine MXenes are substantially improved via enhancing the material porosity and appearing heterojunctions at the interface of involved components. In the case of MXenes decorated with a minor amount of metal oxides, the latter atoms could embed into the inter-layer space and be not just localized on the surface of MXenes that facilitates its access to analyte molecules from a gas phase, increasing the number of adsorption centers.

The detection mechanisms suggested by the authors for Mxene-derived composites are not always uncontroversial. However, in recent years, the following factors have appeared to be among the primary reasons explaining the enhancement of their sensory properties:

(1)

An increase in the specific surface area of nanocomposites, thus raising the diffusion rate of gases and advancing the number of adsorption centers;

(2)

Retention of polar functional groups (–O, –OH, –F, etc.) on the MXene surface within the composites and the appearance of titanium defects that improve gas adsorption, in particular for analytes capable of forming hydrogen bonds;

(3)

High electric conductivity and charge carrier mobility, typical for MXenes, which allow one to raise in some cases the energy efficiency of sensors based on composites with semiconductor metal oxides;

(4)

Composed of heterojunctions that support an effective separation of charge carriers at the interfaces and improve the chemiresistive response;

(5)

Growing the number of dissociative adsorption centers with the introduction of noble metals, which exhibit a high catalytic activity in various surface reactions.

One of the most effective ways to reduce the response/recovery time is UV irradiation, optionally pulsed, to promote primarily analyte desorption, thereby accelerating this process. Still, in some composites such as Ti₃C₂T_x/TiO₂ and Ti₃C₂T_x/ZnO, the UV-activation also improves sensitivity and reduces the detection limit to the ppb range of analyte concentrations.

Despite the observed progress in the development of Ti₃C₂T_x/Ti₂CT_x-based receptor materials for RT chemiresistive gas sensors, there are some challenges, common to all sensor materials and specific to MXenes:

(1)

High sensitivity to humidity hinders the detection of other analyte gases;

(2)

Urgent need to detect individual components in complex gas atmospheres, i.e., under interreference with the presence of other gases that facilitate a sensor response;

(3)

High, uncontrollable ability of MXenes to reduce when synthesizing gas sensor nanocomposites.

One of the promising areas where Ti₃C₂T_x- and Ti₂CT_x-based materials can be widely applied is the development of multisensor units [191] that allow one to distinguish even chemically akin analytes due to the mathematical signal processing of vector signals. A combination of printed technologies [192] for applying sensor coatings to local miniaturized areas of the multisensor and flexible substrates [193] convenient for embedding gas sensors into portable electronics can be extremely beneficial for this purpose.