2D Materials for Environmental Sensing: Detecting Pollution at the Atomic Scale, Research from Parul University’s Micro-Nano Research and Development Center

A conventional sensor detects a pollutant when enough molecules accumulate to change something measurable. Most of the material doing the detecting is buried inside the sensor, where no pollutant will ever reach it. Only the outer surface participates, and the rest is structural.

A material one atom thick has no inside. Every atom sits on the surface, exposed, and a single molecule landing anywhere on it perturbs the electronic state of the whole sheet. That is the physical basis of two-dimensional materials for environmental sensing, and it is why they can detect contaminants at concentrations that defeat conventional approaches.

Researchers at Parul University’s Micro-Nano Research and Development Center have published a review mapping where this field has reached, covering both pollution detection and water purification, in the Journal of Water Process Engineering. This article sets out the science the review surveys: which materials matter, how they actually sense, what they can do for water, and the problems still standing between the laboratory and deployment.

The advantage of a two-dimensional material is not simply that it is small. It is that the ratio of surface to volume reaches its physical limit, and that changes the sensing mechanism itself.

In a bulk sensing material, an adsorbed molecule affects a thin surface layer while the bulk beneath continues behaving as it did. The signal is diluted by everything that did not participate. In a monolayer there is nothing that does not participate. Charge transferred between an adsorbed molecule and the sheet alters the carrier concentration across the entire material, and since electrical resistance depends directly on carrier concentration, that transfer becomes a measurable electrical signal.

Three properties follow from this and together define why the field exists.

• Extreme surface area per unit mass and surface area to volume ratio: which means more binding sites available per gram of material as well as per unit volume, mattering as much for adsorption-based water treatment and for sensing.
• Direct electronic response to adsorption: a molecule does not need to react or accumulate. The charge gets transferred when it comes in contact, and it is enough to shift the material’s electrical behaviour, which is what enables detection at very low concentrations.
• Tunable surface chemistry: the exposed surface can be chemically modified to bind particular targets preferentially, which is how selectivity is engineered into a material that is otherwise sensitive to almost everything.

A material one atom thick has no inside. Every atom is exposed, so a single molecule landing anywhere perturbs the whole sheet.

The review covers several families of two-dimensional material, each with a different combination of properties and therefore a different set of applications. They are not interchangeable.

A single layer of carbon atoms is organised in a hexagonal lattice, and is the material that opened the field. The property that makes it stand out is its exceptionally high charge carrier mobility, meaning electrons move through it with very little resistance, which makes its electrical response to adsorbed molecules both large and fast. The limitation is the mirror of its strength: pristine graphene has no bandgap. This constrains some electronic applications and means it responds to a great many species without discriminating strongly between them. Hence, much of the sensing work is about adding selectivity to an indiscriminately sensitive material.

Compounds pairing a transition metal such as molybdenum or tungsten with a chalcogen such as sulphur or selenium, molybdenum disulphide being the most studied. Unlike graphene these materials do have a band gap, and in monolayer form that band gap is direct, which makes them semiconductors rather than semimetals. That distinction matters because a semiconductor can be switched, giving a much larger relative change in conductance on adsorption and making these materials well suited to transistor-based sensing architectures.

A more recent family of two-dimensional transition metal carbides and nitrides, produced by chemically etching layers out of a parent material. Two properties make them particularly interesting for water applications. They are metallically conductive, which is unusual for a material with this much surface chemistry. And their surfaces carry terminating groups such as oxygen, hydroxyl and fluorine, which are hydrophilic and provide abundant binding sites for ionic and polar contaminants. A conductive material that disperses readily in water and binds heavy metal ions is close to purpose-built for water sensing and treatment.

A metal-free polymeric semiconductor built from carbon and nitrogen, notable for a band gap that places it in the visible range rather than the ultraviolet. That makes it photocatalytically active under ordinary sunlight, which is the property most photocatalysts lack. It is also chemically stable, thermally robust and made from abundant elements, so it is inexpensive. Its weakness is rapid recombination of photogenerated charge carriers, which is why it is so often used in composites rather than alone.

A puckered layered form of phosphorus with high carrier mobility and, unusually, a band gap that varies with the number of layers, from a narrow gap in bulk form to a much wider one in a monolayer. That tunability is scientifically attractive because it allows the material’s optical and electronic response to be selected during fabrication. Its serious limitation is stability: black phosphorus degrades on exposure to air and moisture, which is a substantial obstacle for a material intended to operate in environmental conditions, and protecting it without blocking the surface that does the sensing remains an open problem.

Air quality monitoring is the application where the electronic sensitivity of two-dimensional materials pays off most directly. The ruling architecture is the chemiresistive sensor, which is conceptually simple: the material is suspended between two electrodes, gas molecules absorb on it, charge transfer changes its resistance, and that change is measured.

The 2D materials are effective here, as the resistance change is large and relative to the amount of gas involved, because adsorption affects the entire conducting path rather than a surface layer of it. This permits detection of gases such as nitrogen dioxide, ammonia and hydrogen sulphide at very low concentrations and of volatile organic compounds that conventional sensors detect only when heavily present.

The practical significance is a matter of scale and cost. Conventional reference-grade air monitoring uses large, expensive instruments deployed at a small number of fixed stations, which is how national networks under bodies such as the Central Pollution Control Board are built. A sensor that is small, low-powered and inexpensive can be deployed in far greater numbers, and a dense network of adequate sensors produces a spatial picture of pollution that a handful of excellent instruments cannot.

Two honest caveats belong here, because they are the reason such sensors have not simply replaced conventional instruments. Sensitivity is not selectivity: a material that responds strongly to many gases will report a change without identifying what caused it, and distinguishing between species in a real atmospheric mixture is considerably harder than detecting one in a laboratory. And humidity and temperature affect adsorption, so a sensor that performs well in controlled conditions can drift substantially in the field.

Water is where the review is centred, and it is the application where two-dimensional materials do two different jobs. Water quality and treatment in India fall under the Ministry of Jal Shakti, with discharge standards set through the Central Pollution Control Board.

Detection in water usually works electrochemically rather than by simple resistance change. The two-dimensional material forms or coats an electrode, the target species interacts with its surface, and the resulting change in electrical behaviour is measured as the electrode potential is varied. Heavy metal ions such as lead, cadmium and mercury are a major target, along with pesticides, pharmaceutical residues and organic pollutants.

MXenes are particularly suited to this combination of requirements, since they are conductive enough to serve as an electrode material while carrying surface groups that bind metal ions strongly. Selectivity is again the hard part, and it is generally engineered by functionalising the surface with groups that bind the intended target preferentially.

The removal of contaminants rather than detection has three mechanisms that recur across the literature:

• Adsorption: contaminants are clung to the material’s surface and are removed with it. The huge surface area of 2D material means high capacity per gram, and surface functional groups determine what binds and how strongly.
• Membrane separation: stacked two-dimensional sheets form a laminate whose interlayer spacing acts as a molecular sieve. Because that spacing can be adjusted during fabrication, the size of what passes through can be controlled, which is a level of tunability conventional polymer membranes do not offer.
• Photocatalytic degradation: rather than capturing a pollutant, the material uses light to break it apart. Visible-light-active materials such as graphitic carbon nitride are central here, since a photocatalyst that works under sunlight is far more practical than one requiring ultraviolet illumination.

The distinction between the first two and the third is worth drawing, because it determines what happens afterwards. Adsorption and membrane separation concentrate a pollutant rather than destroying it, which leaves a loaded material to be regenerated or disposed of. Photocatalysis breaks the pollutant down into simpler products, which avoids that problem but demands light, time and a catalyst that survives repeated use. This is the same set of trade-offs explored in photocatalytic dye degradation research presented at the Center.

A pristine two-dimensional material is rarely the best version of itself for a given application. The review covers the modification strategies that turn a promising material into a usable one, and the same three approaches recur across the field.

• Surface functionalisation: attaching chemical groups to the surface to control what binds to it. This is the primary route to selectivity, converting a material that responds to everything into one that responds preferentially to a chosen target, and it can also improve dispersibility in water and stability in air.
• Heterostructure engineering: stacking two different two-dimensional materials so that their electronic structures interact at the interface. In photocatalysis this is used to separate photogenerated electrons and holes across the junction, which suppresses the recombination that otherwise wastes most absorbed light. It is the same principle behind composite photocatalyst design generally.
• Hybrid composites: combining two-dimensional materials with nanoparticles, polymers or other supports to obtain properties none of the components has alone, and often to solve a practical problem such as recovering a powdered material from treated water.

The common thread is that performance is engineered rather than discovered. A material’s intrinsic properties set the starting point, and modification determines whether it becomes useful. This is why a review of the field is a review of methods as much as of materials.

The forward-looking portion of the review concerns what happens when these materials are combined with the digital infrastructure now available around them, and the argument is about scale rather than novelty.

A sensor that is small, low-powered and cheap can be deployed in numbers, connected, and left to report continuously. That is the Internet of Things applied to environmental monitoring: many distributed nodes producing a continuous stream of readings rather than periodic manual sampling at a few locations. Wearable versions extend the same idea to personal exposure, measuring what an individual actually encounters rather than what a distant monitoring station records.

Artificial intelligence enters because that arrangement produces more data than anyone will read. Pattern recognition across a distributed network can identify pollution events, distinguish genuine signals from sensor drift, and support faster decisions than manual analysis allows. There is also a technically interesting role in the selectivity problem: an array of sensors that each respond differently but imperfectly to a range of species can, taken together and interpreted by a trained model, identify what is present more reliably than any single sensor in the array.

The direction the review points toward is therefore not a better individual sensor but a different monitoring architecture, in which sensitivity comes from the materials, coverage from low cost, and interpretation from computation.

The most useful section of any honest review is the one about what does not yet work, and this is where the gap between published performance and practical technology is set out.

• Scalable manufacturing: many two-dimensional materials are produced in laboratory quantities by methods that do not scale, and material made at scale is frequently less uniform than material made carefully in small batches.
• Reproducibility: performance that depends on precise layer count, defect density or surface termination is difficult to reproduce consistently, which complicates both research comparison and manufacture.
• Stability: materials must survive extended operation in humid, chemically complex and variable conditions. Black phosphorus is the clearest case, but degradation, fouling and drift affect the field broadly.
• Selectivity in real matrices: a sensor validated against a single contaminant in clean laboratory water faces a very different problem in actual environmental water containing competing species.
• Regeneration and disposal: an adsorbent that captures a pollutant must be regenerated economically or disposed of safely, and a treatment that produces contaminated waste has moved the problem rather than solved it.
• Cost and commercialisation: a material that works but costs more than the conventional method it replaces will not be adopted, whatever its performance.

None of these is a reason for pessimism about the field, and stating them is what makes the optimistic parts credible. They are the specific problems that determine whether laboratory results become deployed technology, and identifying them clearly is one of the functions a review is for.

The review, titled “Nanoengineered Two-Dimensional Materials for Environmental Sensing: From Air Quality Monitoring to Water Purification”, is authored by Mohit Tannarana, Anand Joshi and Vidhi Pathak of the Micro-Nano Research and Development Center at Parul University. It appears in the Journal of Water Process Engineering, published by Elsevier and indexed in both SCIE and Scopus. In the Journal Citation Reports release of June 2026, based on 2025 citation data, the journal carries an impact factor of 6.6 and is classified as Q1.

It is worth being precise about what kind of contribution this is, because review articles are frequently misunderstood. This paper does not report new experiments. It surveys a large body of already-published research, evaluates it, and organises it into a coherent account of where a field stands.

That is a distinct and demanding form of scholarship rather than a lesser one. A researcher entering a fast-moving area faces hundreds of papers of varying quality using inconsistent terminology and incompatible measurement conditions. A good review establishes what has actually been demonstrated as opposed to claimed, identifies where results conflict, and states plainly what remains unsolved. Reviews are consequently among the most heavily cited article types in the scientific literature, because they become the entry point through which others approach the subject.

For a research centre, a review in a Q1 journal signals something specific: that its researchers know a field well enough to be trusted to map it. Peer review at this level is selective, and the paper sits alongside the Center’s other work in materials characterisation and nanomaterials, including its spectroscopy and materials characterisation programme. Research of this kind sits within the national research effort supported by bodies including the Department of Science and Technology and the Department of Scientific and Industrial Research.