Fabrication Batch Research Articles

Integrating inverse design and partially etched platform: an ultra-compact polarization splitter and rotator as an example

Silicon photonics devices benefit greatly from a partially etched platform and inverse design. Herein, we propose a bi-layer polarization splitter and rotator with a topology pattern and demonstrate it on a silicon-on-insulator platform. Our device exhibits a significantly reduced physical footprint of only 2µm×6µm, compared to traditional directional couplers and tapered waveguides. The device accomplishes the functions of polarization conversion and separation in such a compact design without redundant tapered or bending waveguides. The tested minimum insertion loss with the fabrication batch reaches 0.57 and 0.67 dB for TE and TM modes, respectively. The TE mode demonstrates a wider bandwidth and lower ILs than the TM modes, averaging around 1 dB from 1530 to 1565 nm. The M modes exhibit approximately 2 dB ILs at the same wavelength range, decreasing to about 1 dB between 1565 and 1580 nm. Improved designs and fabrication conditions strongly suggest the potential for further performance enhancement in the device. This successful initiative validates the exceptional performance resulting from the integration of the partially etched platform and inverse design, providing valuable insights for future photonic integrated device designs.

Controlling the Architecture of Freeze-Dried Collagen Scaffolds with Ultrasound-Induced Nucleation.

Collagen is a naturally occurring polymer that can be freeze-dried to create 3D porous scaffold architectures for potential application in tissue engineering. The process comprises the freezing of water in an aqueous slurry followed by sublimation of the ice via a pre-determined temperature-pressure regime and these parameters determine the arrangement, shape and size of the ice crystals. However, ice nucleation is a stochastic process, and this has significant and inherent limitations on the ability to control scaffold structures both within and between the fabrication batches. In this paper, we demonstrate that it is possible to overcome the disadvantages of the stochastic process via the use of low-frequency ultrasound (40 kHz) to trigger nucleation, on-demand, in type I insoluble bovine collagen slurries. The application of ultrasound was found to define the nucleation temperature of collagen slurries, precisely tailoring the pore architecture and providing important new structural and mechanistic insights. The parameter space includes reduction in average pore size and narrowing of pore size distributions while maintaining the percolation diameter. A set of core principles are identified that highlight the huge potential of ultrasound to finely tune the scaffold architecture and revolutionise the reproducibility of the scaffold fabrication protocol.

Preparation of high-performance large-area perovskite solar cells by atomic layer deposition of metal oxide buffer layer

Perovskite solar cells have been widely recognized as the most promising new-type photovoltaic device due to its power conversion efficiency rapidly increasing from 3.8% to over 26% in merely fifteen years. However, the high performances are achieved mainly on small area cells with an active area lower than 0.1 cm<sup>2</sup>. When enlarging the active area of perovskite solar cells, the efficiency falls dramatically. So, how to reduce the gap between performances of small area cells and large area cells gradually becomes a critical point in the path towards the commercialization of perovskite photovoltaic technology. Herein, a strategy of pre-growing thin layer of TiO<sub>2</sub> on a rough FTO substrate by atomic layer deposition method before spin-coating SnO<sub>2</sub> nanoparticles is proposed. Due to the inherent conformal film growth mode of atomic layer deposition, the FTO substrate can be completely covered by TiO<sub>2</sub>, thus preventing the direct contact between local protrusions of FTO and perovskite layer and impeding the current leakage phenomenon, which can be verified by the measurements from X-ray photoelectron spectroscopy, scanning electron microscopy, and atomic force microscopy, and further proved by the dark current measurement. By using this method, the repeatability and consistency of the small area cell fabrication technology on the same substrate are improved obviously. The improved electron transport process revealed by photoluminescence results and incident light management process revealed by external quantum efficiency results also brings about better solar cell performances. More importantly, highly efficient 0.5 cm<sup>2</sup> large area perovskite solar cells are fabricated through optimization of TiO<sub>2</sub> thickness. When growing 200 cycles TiO<sub>2</sub> (~9 nm in thickness) by using atomic layer deposition technology, the champion large area perovskite solar cell possesses a power conversion efficiency as high as 24.8% (certified 24.65%). The device performances also show excellent repeatability between different fabrication batches. The perovskite solar cell with TiO<sub>2</sub> buffer layer grown by the atomic layer deposition method can still retain over 95% of its initial efficiency after having been stored in a nitrogen atmosphere for 1500 h. The technique proposed in this paper can be helpful in manufacturing perovskite solar cell modules in the realistic photovoltaic market and can be extended to the large area fabrication of other perovskite optoelectronic devices such as light emitting diode, laser and detector.

Fabrication-Aware Joint Clustering in Freeform Space-Frames

The geometrical variability in the joints of large-scale, doubly-curved space-frame structures can have a substantial impact on the time and cost of their construction. This paper proposes a novel framework to assess the construction complexity of space-frame structures as a factor of the geometrical variability and fabrication of their joints, to promote the informed design of the fabrication process. The k-means algorithm was used to cluster space-frame joints into fabrication batches, providing an overview of the variability distribution. A novel initialisation method was developed that allows the algorithm to adapt to project-specific inputs, substantially improving cluster compactness. Overlaying the clustering results with the properties of different fabrication processes provides an accurate estimation of the construction complexity of alternative fabrication options. The method was applied to a large-scale case study to demonstrate the benefits in practice. Alternative fabrication scenarios were assessed in the early stages of the design development, leading to the informed design of the fabrication process and hence to the efficient construction of large-scale, complex structures.

ATMOSPHERIC CORROSION BEHAVIOR OF HOT-DIP GALVANIZED AND CONTINUOUS GALVANIZED STEEL

Five years of outdoor atmospheric corrosion tests of hot dip galvanized steel samples were performed, for products of continuous galvanizing and after fabrication batch galvanizing processes. For the purpose of comparison between corrosion performances of these two different galvanizing process products, an industrial coastal area (Gemlik-Bursa/TURKEY) was chosen for outdoor testing, which fits into C4 type corrosive atmosphere definition, according to ISO 9223. Samples were studied in laboratory with accelerated salt spray exposure test and electrochemical methods. Corrosion products formed on exposed samples and cross section of coatings are analyzed by SEM. Lead is observed to change the corrosion characteristics of the coatings with change in constituents of environments. In saline electrolytes, alloying of lead is found to accelerate corrosion rate. This metal deposits as cluster on top layer of the galvanized coatings and acts as strong cathodes with respect to the zinc and accelerates the corrosion rate. It was determined that differences in dip and continuous galvanization processes cause dramatic differences in the elemental composition, morphology and regional hardness values of coatings. In the comparison of corrosion resistance, lower performance of the dip galvanized coating, although it is much thicker, has been shown due to the differences mentioned above.

TEMPO-Functionalized Carbon Nanotubes for Solid-Contact Ion-Selective Electrodes with Largely Improved Potential Reproducibility and Stability.

Solid-contact ion-selective electrodes (SCISEs) can overcome essential limitations of their counterparts based on liquid contacts. However, attaining a highly reproducible and predictable E0, especially between different fabrication batches, turned out to be difficult even with the most established solid-contact materials, i.e., conducting polymers and large-surface-area conducting materials (e.g., carbon nanotubes), that otherwise possess excellent potential stability. An appropriate batch-to-batch E0 reproducibility of SCISEs besides aiding the rapid quality control of the electrode manufacturing process is at the core of their “calibration-free” application, which is perhaps the last major challenge for their routine use as single-use “disposable” or wearable potentiometric sensors. Therefore, here, we propose a new class of solid-contact material based on the covalent functionalization of multiwalled carbon nanotubes (MWCNTs) with a chemically stable redox molecule, (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO). This material combines the advantages of (i) the large double-layer capacitance of MWCNT layers, (ii) the adjustable redox couple ratio provided by the TEMPO moiety, (iii) the covalent confinement of the redox couple, and (iv) the hydrophobicity of the components to achieve the potential reproducibility and stability for demanding applications. The TEMPO-MWCNT-based SC potassium ion-selective electrodes (K+-SCISEs) showed excellent analytical performance and potential stability with no sign of an aqueous layer formation beneath the ion-selective membrane nor sensitivity toward O2, CO2, and light. A major convenience of the fabrication procedure is the E0 adjustment of the K+-SCISEs by the polarization of the TEMPO-MWCNT suspension prior to its use as solid contact. While most E0 reproducibility studies are limited to a single fabrication batch of SCISEs, the use of prepolarized TEMPO-MWCNT resulted also in an outstanding batch-to-batch potential reproducibility. We were also able to overcome the hydration-related potential drifts for the use of SCISEs without prior conditioning and to feature application for accurate K+ measurements in undiluted blood serum.

Enhanced Nitrite Detection by a Carbon Screen Printed Electrode Modified with Photochemically-Made AuNPs

Excessive nitrite amounts harm the environment and put public health at high risk. Therefore, accurate and sensitive detection of nitrite in surface and groundwater is mandatory for mitigating its adverse effects. Herein, a highly sensitive electrochemical sensor based on carbon screen-printed electrodes (CSPE) surface-modified with photochemically-made gold nanoparticles (AuNPs, ~12 nm) is proposed for nitrite detection. Scanning electron microscopy, cyclic voltammetry, and electrochemical impedance spectroscopy show that AuNPs uniformly coat the CSPE, increase its surface area, and contribute to oxidizing nitrite to much lower potential (+0.5 V vs. Ag/AgCl) and faster rate. Under optimized differential pulse voltammetry conditions, the CSPE/AuNPs-PEI electrode responds linearly (R2 > 0.99) to nitrite within a wide concentration range (0.01–4.0 µM), showing a sensitivity of 0.85 µA·µM−1·cm−2 and limit of detection as low as 2.5 nM. The CSPE/AuNPs-PEI electrode successfully detects nitrite in tap water and canned water of olives, showing no influence of those matrices. In addition, the electrode’s response is highly reproducible since a relative standard deviation lower than 10% is observed when the same electrode is operated in five consecutive measurements or when electrodes of different fabrication batches are evaluated.

Incorporating an outsourcing strategy and in-house quality assurance into the production-shipment decision making

To stay competitive in turbulent business environments, manufacturing firms' managers today constantly seek ways to reduce order response time, smooth production schedules, ensure the quality of their products, and lower overall making and shipping costs. This study incorporates an outsourcing strategy and in-house quality assurance into a production-shipment problem to address the aforementioned operational goals. The objectives are to simultaneously find the optimal fabrication batch size and frequency of delivery that minimize the system's relevant costs and reveal in-depth information regarding the impact of diverse system parameters on the optimal policy and system cost. This study develops a model and uses the optimization method to resolve the problem. The research results facilitate managerial decisions in such a real-life situation.

Global calibration models for temperature-modulated metal oxide gas sensors: A strategy to reduce calibration costs

Tolerances in the fabrication of metal oxide (MOX) gas sensors lead to inter-device variability in baseline and sensitivity, even for sensors of the same fabrication batch. This has traditionally forced the use of individual calibration models (ICMs) built specifically for each sensor unit, which requires an expensive and time-consuming calibration process and hinders sensor replacement. We propose Global calibration models (GCMs) built using the responses of multiple sensor units, and then applied to a new sensor unit that is not part of the calibration set. GCM have been already successfully applied to transfer calibration models between sensor arrays (electronic noses) for classification tasks. In this work, we investigate the use of such models for regression purposes in temperature-modulated sensors, aiming at the quantification of low concentrations of carbon monoxide (CO) in the presence of variable humidity levels (20–80% r.h. at 26 ± 1 °C). Using a laboratory dataset containing data from 6 replicas of the FIS SB-500–12 model, we evaluate the performance of global models built with data from 1 to 4 sensors when applied to unseen sensor units. Results show that the performance of global models improves with an increasing number of sensors in the calibration set, approaching the performance of individual calibration models (1.38 ± 0.15 ppm for GCM; 1.05 ± 0.24 ppm for ICM), and surpassing their performance only if few calibration conditions per sensor are available (2.09 ± 0.10 ppm for GCM;; 2.76 ± 0.22 ppm for ICM, if only 5 samples per sensor are used).

Influence of Microgel Fabrication Technique on Granular Hydrogel Properties.

Bulk hydrogels traditionally used for tissue engineering and drug delivery have numerous limitations, such as restricted injectability and a nanoscale porosity that reduces cell invasion and mass transport. An evolving approach to address these limitations is the fabrication of hydrogel microparticles (i.e., "microgels") that can be assembled into granular hydrogels. There are numerous methods to fabricate microgels; however, the influence of the fabrication technique on granular hydrogel properties is unexplored. Herein, we investigated the influence of three microgel fabrication techniques (microfluidic devices (MD), batch emulsions (BE), and mechanical fragmentation by extrusion (EF)) on the resulting granular hydrogel properties (e.g., mechanics, porosity, and injectability). Hyaluronic acid (HA) modified with various reactive groups (i.e., norbornenes (NorHA), pentenoates (HA-PA), and methacrylates (MeHA)) were used to form microgels with an average diameter of ∼100 μm. The MD method resulted in homogeneous spherical microgels, the BE method resulted in heterogeneous spherical microgels, and the EF method resulted in heterogeneous polygonal microgels. Across the various reactive groups, microgels fabricated with the MD and BE methods had lower functional group consumption when compared to microgels fabricated with the EF method. When microgels were jammed into granular hydrogels, the storage modulus (G') of EF granular hydrogels (∼1000-3000 Pa) was consistently an order of magnitude higher than G' for MD and BE granular hydrogels (∼50-200 Pa). Void space was comparable across all groups, although EF granular hydrogels exhibited an increased number of pores and decreased average pore size when compared to MD and BE granular hydrogels. Furthermore, granular hydrogel properties were tuned by varying the amount of cross-linker used during microgel fabrication. Lastly, granular hydrogels were injectable across formulations due to their general shear-thinning and self-healing properties. Taken together, this work thoroughly characterizes the influence of the microgel fabrication technique on granular hydrogel properties to inform the design of future systems for biomedical applications.

SENSEI: Direct-Detection Results on sub-GeV Dark Matter from a New Skipper CCD.

We present the first direct-detection search for sub-GeV dark matter using a new ∼2-gram high-resistivity Skipper CCD from a dedicated fabrication batch that was optimized for dark matter searches. Using 24days of data acquired in the MINOS cavern at the Fermi National Accelerator Laboratory, we measure the lowest rates in silicon detectors of events containing one, two, three, or four electrons, and achieve world-leading sensitivity for a large range of sub-GeV dark matter masses. Data taken with different thicknesses of the detector shield suggest a correlation between the rate of high-energy tracks and the rate of single-electron events previously classified as "dark current." We detail key characteristics of the new Skipper CCDs, which augur well for the planned construction of the ∼100-gram SENSEI experiment at SNOLAB.

Mechano-activated biomolecule release in regenerating load-bearing tissue microenvironments

Although mechanical loads are integral for musculoskeletal tissue homeostasis, overloading and traumatic events can result in tissue injury. Conventional drug delivery approaches for musculoskeletal tissue repair employ localized drug injections. However, rapid drug clearance and inadequate synchronization of molecule provision with healing progression render these methods ineffective. To overcome this, a programmable mechanoresponsive drug delivery system was developed that utilizes the mechanical environment of the tissue during rehabilitation (for example, during cartilage repair) to trigger biomolecule provision. For this, a suite of mechanically-activated microcapsules (MAMCs) with different rupture profiles was generated in a single fabrication batch via osmotic annealing of double emulsions. MAMC physical dimensions were found to dictate mechano-activation in 2D and 3D environments and their stability in vitro and in vivo, demonstrating the tunability of this system. In models of cartilage regeneration, MAMCs did not interfere with tissue growth and activated depending on the mechanical properties of the regenerating tissue. Finally, biologically active anti-inflammatory agents were encapsulated and released from MAMCs, which counteracted degradative cues and prevented the loss of matrix in living tissue environments. This unique technology has tremendous potential for implementation across a wide array of musculoskeletal conditions for enhanced repair of load-bearing tissues.

Humidity sensitivity of large area silicon sensors: Study and implications

The production of large area sensors is one of the main challenges that the ATLAS collaboration faces for the new Inner-Tracker full-silicon detector. During the prototype fabrication phase for the High Luminosity Large Hadron Collider upgrade, several ATLAS institutes observed indications of humidity sensitivity of large area sensors, even at relative humidities well below the dew point. Specifically, prototype Barrel and End-Cap silicon strip sensors fabricated in 6-inch wafers manifest a prompt decrease of the breakdown voltage when operating under high relative humidity, adversely affecting the performance of the sensors. In addition to the investigation of these prototype sensors, a specific fabrication batch with special passivation is also studied, allowing for a deeper understanding of the responsible mechanisms.This work presents an extensive study of this behaviour on large area sensors. The locations of the hotspots at the breakdown voltage at high humidity are revealed using different infrared thermography techniques. Several palliative treatments are attempted, proving the influence of sensor cleaning methods, as well as baking, on the device performance, but no improvement on the humidity sensitivity was achieved. Furthermore, a study of the incidence of the sensitivity in different batches is also presented, introducing a hypothesis of the origins of the humidity sensitivity associated to the sensor edge design, together with passivation thickness and conformity. Several actions to be taken during sensor production and assembly are extracted from this study, in order to minimize the impact of humidity sensitivity on the performance of large area silicon sensors for High Energy Physics experiments.

THERMOCOAX rhodium SPND sensitivity dispersion and validation of the sensitivity calculation model

This paper describes the neutron irradiation tests of 7 THERMOCOAX Self-Powered Neutron detectors (SPNDs) in the BR1 reactor at SCK•CEN. The SPNDs were fabricated according to the same specifications, but from different fabrication batches. The SPND signals were recorded during stepwise power-up of the reactor, proving the linearity of the SPND response within a wide thermal neutron flux range: from as low as 0.8·109 n/(cm2) to 2.6·1011 n/(cm2s). Intercomparison of the SPND signals shows a very small spread, confirming the repeatability of the THERMOCOAX fabrication processes. The overall neutron sensitivities of the seven SPNDs agreed within a 1% margin. The experimental data were analyzed in terms of prompt and various delayed responses. Prompt contributions to the signal are due to external gamma induced processes and to processes involving gamma rays emitted instantaneously upon neutron capture. The main contribution in a rhodium SPND is due to activation of the rhodium emitter and beta emission during decay of the activated rhodium and leads to a delayed response with a characteristic time of a few minutes. Activation and subsequent beta decay in other materials present in the SPND lead to additional minor delayed signal contributions. The partial SPND sensitivities due to all these processes were calculated using an MCNPX based model and were compared with experimental sensitivities based on the recorded data. The results were in fair agreement; for the overall SPND neutron sensitivity an agreement within a 1% margin was achieved.

Electronic design automation for increased robustness in inkjet-printed electronics

Printed electronics is of great interest for low-cost, large-area fabrication due to fast and scalable processes with low material consumption. However, notable resources are currently required for fine-tuning layout designs and process parameters to prevent unintended fabrication artifacts. This study proposes the use of electronic design automation to mitigate such artifacts by segmenting input layouts and arranging the resulting objects in separate conflict-free layers. To this end, two exemplary conflicts with relevance for inkjet printing are defined. The first addresses heterogeneous structures composed of low- and high-resolution features. The second is concerned with layouts requiring a high feature density. Experimental data is used to derive constraints for layer arrangement. Finally, mixed-integer-linear programming is used to formally model the problem and minimize the total number of layers under the aforementioned constraints. The results of printing test layouts with and without algorithmic treatment are characterized in terms of film morphology, device functionality, and fabrication yield. The data demonstrates a more consistent fabrication outcome and significantly increased yields for optimized fabrication batches compared to non-optimized ones. The final processing of the layouts requires no human intervention and can readily be transferred to a variety of ink-substrate systems.

Injection compression molding of transmission-type Fano resonance biochips for multiplex sensing applications

Nanostructure-based surface plasmon resonance biosensors have attracted considerable attention since the phenomenon of extraordinary light transmission in metallic nanohole arrays was discovered. However, the mass production of uniform metallic nanostructures with a low-cost, rapid, and high-throughput fabrication process remains a key issue for various multiplex sensing applications. We successfully utilized injection compression molding to mass fabricate transmission-type Fano resonance biochips with a feature size of 60nm for multiplex sensing applications. Two types of metallic nanostructures, aluminum nanoslits and capped aluminum nanoslits with 24 sensing arrays, were made on polycarbonate substrates. The bulk sensitivity and uniformity of the nanostructure arrays were tested. A Fano resonance with a full-width at half-maximum bandwidth of 8nm was observed in the visible light region for 470-nm-period capped aluminum nanoslit arrays. The refractive index sensitivity was 460nm/RIU, and a figure of merit of 58 was achieved. Moreover, aluminum nanoslits had a dip resonance with a bandwidth of 25nm and a refractive index sensitivity of 463nm/RIU. The coefficients of variation of the refractive index sensitivities for 24 arrays on a biochip and 10 biochips from different fabrication batches were both below 3%, which indicates that uniform nanostructures can be fabricated by injection compression molding and that the reproducibility is controllable. In addition, the multiplex sensing capability of the aluminum nanostructure arrays was demonstrated by simultaneously monitoring 24 water/glycerin solutions with a hyperspectral imaging system. Protein–protein interactions between bovine serum albumin and anti-bovine serum albumin demonstrated proof of concept of the biological detection capability of the chips. To benefit various multiplex sensing applications such as clinical disease diagnosis, drug screening, and protein biomarker discovery, 96-array aluminum nanoslit biochips the same size as a standard 96-well plate were successfully fabricated. Such a nanostructure-based plasmonic biochip produced by a low-cost, rapid, and high-throughput fabrication method can benefit commercial applications.

Comparison of Superconducting Nanowire Single-Photon Detectors Made of NbTiN and NbN Thin Films

We fabricated superconducting nanowire single-photon detectors (SNSPDs) made of NbTiN and NbN thin films in the same fabrication batch. Various parameters, such as critical temperature, resistivity, switching current, and kinetic inductance, were compared between the two types of SNSPDs. These data indicate that NbTiN SNSPD surpasses NbN SNSPD in many aspects, including a shorter recovery time due to smaller kinetic inductance, a lower timing jitter due to higher switching current, and better uniformity. Based on the measurement of 28 out of 121 NbTiN SNSPDs on a 2-in Si wafer, 71% of detectors show a system detection efficiency of over 50% at a dark count rate of 100 Hz, with the maximum efficiency being 77.5%.

A monolithically integrated torsional CMOS-MEMS relay

We report experimental demonstrations of a torsional microelectromechanical (MEM) relay fabricated using the CMOS-MEMS approach (or intra-CMOS) which exploits the full foundry inherent characteristics enabling drastic reduction of the fabrication costs and batch production. In particular, the relay is monolithically integrated in the back end of line of a commercial standard CMOS technology (AMS 0.35 μm) and released by means of a simple one-step mask-less wet etching. The fabricated torsional relay exhibits an extremely steep switching behaviour symmetrical about both contact sides with an on-state contact resistance in the k-range throughout the on-off cycling test.

A detailed study of magnetization reversal in individual Ni nanowires

Magnetic nanowires have emerged as essential components for a broad range of applications. In many cases, a key property of these components is the switching field, which is studied as a function of the angle between the field and the nanowire. We found remarkable differences of up to 100% between the switching fields of different nanowires from the same fabrication batch. Our experimental results and micromagnetic simulations indicate that the nanowires exhibit a single domain behavior and that the switching mechanism includes vortex domain wall motion across the nanowire. The differences between the switching fields are attributed to different cross-sections of the nanowires, as found by electron microscopy. While a circular cross-section yields the smallest switching field values, any deviation from this shape results in an increase of the switching field. The shape of the nanowires' cross-sections is thus a critical parameter that has not been previously taken into account.

Hyper-branch sensing polymer batch self-assembled on resonant micro-cantilevers with a coupling-reaction route

In order to develop a volume fabrication route for chemical sensors, hydrogen-bond acidic hyper-branched polymer is synthesized and directly self-assembled in-batch onto MEMS resonant micro-cantilevers as sensing layer. The batch self-assembly is via a facile coupling reaction between the SH group in the self-assembled monolayer (SAM) and one OH group in the hyper-branched polymer. During the self-assembly process of the hyper-branched polymer, the polyurethane-industry used isophorone diisocyanate (IPDI) is employed as the key coupling reagent, since the two NCO coupling groups in IPDI feature different reactivity. The NCO group with higher reactivity is firstly reacted with the SH group pre-grown at the cantilever surface. Then, another NCO group with lower reactivity reacts with one OH group in the hyper-branched polymer. In this way the hydrogen-bond acidic hyper-branch polymer can be batch grafted onto a run of micro-cantilevers for uniform sensing to targeted gas. Through this route, the resonant cantilever gas sensors can be batch produced for uniform detection of trace organophosphates (OPs). Based on the specific interaction between the OH sensing group and the PO group in OP, the targeted molecules can be captured by the hyper-branched sensing polymer, and the mass addition induced frequency shift of the resonant cantilever is output as sensing signal. Sensing experiment is implemented for the sensors taken from one fabrication batch, resulting in satisfactory consistency in sensing performance. Taking 5 sensors as example, the deviation of the sensing response to 2ppm dimethyl methylphosphonate (DMMP) is less than 1%. Besides, the sensor shows good repeatability/selectivity and good linear response to various concentrations (300ppb to 1.5ppm) of DMMP.