Lead-free perovskite sensor reads ammonia at 1 ppm, and its makers say they now know why

A team led by Myungkwan Song of the Energy & Environment Materials Research Division at the Korea Institute of Materials Science has reported an ammonia sensor built on a lead-free perovskite that reliably detects the gas at 1 part per million ^[1][6]. The more consequential result is the second one: the group says it has explained, for the first time, the mechanism by which such a sensor produces its signal, which turns a device tuning exercise into a materials design rule ^[2][12].

The active layer is formamidinium antimony bromide, FA3Sb2Br9, an antimony-based composition used in place of the lead compositions that dominate perovskite gas sensing ^[4][5]. According to KIMS, lead content is what has limited commercialization of otherwise sensitive perovskite ammonia sensors, on both health and environmental grounds ^[5]. Swapping the toxic B-site cation and keeping the sensitivity is the whole point of the exercise.

The reported numbers: at 100 ppm the device responded in 13 seconds, with an electrical signal increase of up to 235% ^[7]. Signal rose consistently with concentration, which is what makes a quantitative reading possible rather than a threshold alarm ^[8]. Responses to methane, carbon monoxide, nitrogen oxides, hydrogen and methanol were substantially lower ^[9]. Hydrogen selectivity is not a footnote here, since the intended deployments are places where hydrogen and ammonia coexist. After two months of storage the sensor retained roughly 97% of its initial performance ^[10], and it is made by spin-coating from solution, which the institute frames as suited to large-area fabrication and lower-cost manufacturing ^[11].

The mechanism claim is the part with legs. Conventional accounts attribute ammonia sensor signals to molecules adsorbing on the surface of the sensing film ^[12]. The team instead found that ammonia reversibly intercalates into the perovskite lattice and induces p-type doping, raising conductivity and generating the signal ^[13]. That is a different design target: bulk lattice chemistry and reversible guest transport rather than surface area and site density ^[14]. The work was done with groups led by Hyung Woo Lee at Pusan National University, Youngho Kang at Incheon National University and Jincheol Kim at Macquarie University, and is published in Small Structures ^[3][15].

Why an operator should care: ammonia is being pushed as a hydrogen carrier because hydrogen itself is awkward to store and move, and exposure at only several tens of ppm can affect worker health ^[16][17]. A 1 ppm floor sits at least an order of magnitude below that harm threshold, which is the margin you need for leak detection rather than incident confirmation ^[18]. KIMS names ammonia-fuelled ships and power plants, production and storage and transport facilities, and fertilizer and chemical plants as targets ^[19].

What to watch is everything the announcement leaves out. There is a response time but no recovery time, no humidity or temperature cross-sensitivity data, and no continuous-operation figure; the only durability number is two months of storage retention ^[20]. Reversible intercalation is a promising mechanism precisely because reversibility is the thing that fails first under repeated high-concentration exposure. The tests that matter next are wet-air baselines, cycling to saturation and back, and behaviour in a hydrogen-rich atmosphere rather than a hydrogen challenge test.