A Polymer‐Based Chemiresistive Gas Sensor for Selective Detection of Ammonia Gas

Breath analysis is a non‐invasive tool used in medical diagnosis. However, the current generation of breath analyzers is expensive, time‐consuming, and requires sample gas separation. In this work, a simple, yet effective, low‐cost ammonia gas sensor based on poly(2‐acrylamido‐2‐methyl‐1‐propanesulfonic acid) is presented for non‐invasive medical diagnosis. The designed sensor has a broad detection range to ammonia gas up to 1000 ppm with a limit of detection of 30 ppb. This is a robust sensor, which functions at high relative humidity (RH) (>90%) and exhibits consistent electrical responses under different test conditions. The result of a blind test validates the sensor's selective response to ammonia in the presence of other gases such as carbon dioxide. Furthermore, it is viable to integrate this sensor into a mask for real‐time ammonia gas detection accurately. Overall, this study demonstrates the feasibility of developing a simple, non‐invasive, and cost‐effective sensor for real‐time monitoring of ammonia gas with high potential in applications such as medical diagnostics, food safety, and environmental conditions.


Introduction
Medical diagnosis plays a vital role in assessing and treating the clinical conditions of patients.The traditional methods of DOI: 10.1002/adsr.202300125diagnosing diseases usually involve an invasive process of extracting biological samples, such as blood, biopsy, excision, and endoscopy. [1,2]Although these results are accurate and can help rule out certain conditions, these procedures are time-consuming, expensive (>$1000 for liver biopsy), labor-intensive, and may lead to other health complications. [3]or instance, liver biopsy is considered the "gold standard" for diagnosing liverbased diseases such as hepatitis, cirrhosis, and Wilson's disease. [4]However, this procedure can lead to complications such as bleeding, infection, and, in some cases infection via bacteria entering the abdominal cavity or bloodstream. [5]Alternative techniques such as either minimally invasive or non-invasive diagnosis procedures are the focus of current research.Among different non-invasive procedures such as X-ray, imaging, and ultrasound, exhaled breath analysis is an emerging versatile tool used for medical detection of Helicobacter pylori bacteria and for diagnosing diseases such as SIBO (small intestinal bacterial overgrowth) with test kits being available for patients to diagnose the disease at home. [6]Exhaled breath consists of 78% nitrogen (N 2 ), 16% oxygen, 4% carbon dioxide (CO 2 ) gas and the remaining are different volatile organic compounds (VOCs). [7]hese VOCs have been linked to multiple clinical conditions.For instance, acetone in breath above two parts per million (ppm) has been linked to diabetes, while increased CO 2 concentrations have been linked to chronic obstructive pulmonary disease. [8,9]mmonia gas (NH 3 ) is an important marker for determining healthy liver and kidney functions in the human body.The body regulates the breakdown of ammonia into urea via the urea cycle that then enters the bloodstream. [10]In mammals, 80% of the toxic nitrogenous compounds are eliminated from the body via urine. [11]Some of the remaining compounds in the bloodstream, diffuse into the salivary gland wherein, the urease enzyme hydrolyses urea to form ammonium ions and carbonic acid ions. [12]his in turn raises the pH of the breath and as long as the pH exceeds 9.24, ammonia gas can be detected directly through exhaled breath.A newer diagnostic approach for detecting H pylori infection utilizes this increase in pH to detect ammonia in breath. [13]t should be noted that for a healthy individual, typically the exhaled breath ammonia concentration varies between 0.05 and 1.5 ppm depending on parameters such as pulmonary functions, saliva, and the microflora in the said individual's respiratory system. [14]While a person suffering from chronic kidney disease would have an exhaled breath ammonia concentration of ≈15 ppm. [14]Current methods for assessing NH 3 concentration in breath include analytical techniques such as gas chromatography, photodetectors, and quartz crystal microbalance, However, while they produce reliable results, they are time-consuming, require sample preparation, and are not easily accessible.Another alternative is using gas sensors for medical monitoring. [11]Chemiresistive sensors are being researched due to their low cost, tunable sensing properties, versatility, and compatibility with different flexible substrates (paper, rubbers, and textiles). [15,16][22][23][24][25][26] For instance, Fujita et al. developed a paper-based NH 3 gas sensor integrated into a face mask. [21]The device could detect NH 3 gas at concentrations as low as two ppm in the presence of N 2 .While this is a low-cost sensor and accurately measures ammonia, the lower limit of detection is significantly higher than 0.1 ppm, which is pivotal for medical diagnostics.
Another approach used a graphene-based sensor doped with nanostructured zinc/copper oxide developed by Jagannathan et al. [27] The designed sensor exhibited a stable response at higher RH levels, with good selectivity toward NH 3 . [27]The sensor showed improved response/recovery time of 4/2 s, respectively.However, the limit of detection of this sensor at room temperature was 5 ppm.Considering that for a healthy individual, the NH 3 gas concentrations via mouth-exhaled and nose-exhaled tests vary between 0.05 and 1.5 ppm, [14] the sensor developed by Jagannathan et al. may not be suitable for such medical diagnostic applications.It is worth noting that a person suffering from a disease such as chronic kidney disease would have an exhaled ammonia breath concentration of ≈15 ppm. [14]n this study, we proposed to design a simple polymer-based sensor at low cost with high sensitivity and selectivity that can operate at high humidity levels (>90% RH) to overcome some of the drawbacks of the existing NH 3 gas sensors.To this end, poly(2acrylamido-2-methyl-1-propanesulfonic acid) (PAMPSA) organic polymer was selected that has a high affinity to NH 3 .It has been used as hydrogels for drug delivery, [28] and as scaffolds for cartilage regeneration, [29] and for NH 3 absorption. [30]Due to the presence of sulfonate functional group (SO 3 − ) in PAMPSA, its electrical resistance is highly dependent on the pH and the presence of other charged or polar species.A chemiresistive sensor based on PAMPSA was developed and its performance for measuring ammonia at various concentrations was examined at different conditions.The sensor's selectivity, long-term stability, and the feasibility of using this sensor for breath testing and medical application were demonstrated.

Sensor Mechanism
The sensing mechanism of this PAMPSA sensor is based on the interaction between the sulfonate functional groups with the charged ionic species under different pH. [30]At high humidity levels, polar gases such as NH 3 and CO 2 dissolved in water, leading to the formation of charged species, such as ammonium, bicarbonate, and carbonate ions, respectively as shown in Equations (1-4).The concentration of dissociated ions affects the electrical response of this sensor.
In the presence of high humidity (e.g., exhaled human breath), due to the hygroscopic nature of PAMPSA, it tends to absorb moisture.When NH 3 is introduced into the immediate environment of the PAMPSA sensor and comes in contact with this moisture present in the sensor system, the gas dissociates to form NH 4 + ions that interact with the SO 3 − of PAMPSA on the surface of the sensor through electrostatic interactions, while the remaining ions form a cluster around the polymer chains [30] as shown in Figure 1A.Compared to the author's previous work where in the interaction between the SO 3 − and NH 4 + ions was stabilized due to the presence of an aromatic ring and required external energy to break the interaction, the absence of the aromatic ring in the PAMPSA structure makes this interaction reversible without needing any external energy. [16] pH test was conducted, wherein PAMPSA solution was exposed to humidified NH 3 gas and approached the pH of NH 4 OH (pH 10.72, Figure 1B), indicating that all the dissociated NH 4 + ions interacted with the SO 3 − in the solution.This interaction between SO 3 − and NH 4 + ions increases the ionic conductivity that results in a decrease in the sensor's resistance.
In contrast, when CO 2 gas encounters moisture present in the sensor system, it dissociates to produce H 2 CO 3 .The interaction of these produced H 2 CO 3 molecules with SO 3 − is dependent on the pK a of both species.The pKa of PAMPSA is 1.9, [31] while the pK a of H 2 CO 3 is dependent on the pH of the immediate environment.This is because H 2 CO 3 has pK a1 of 6.37 and pK a2 of 10.32 and in an acidic medium pK a1 is more relevant.Based on the pK a PAMPSA is the stronger acid and when the PAMPSA solution was exposed to humidified CO 2 gas, the pH of the solution increased from 1.26 (PAMPSA initial pH) and stabilized at 1.59 (Figure 1C).The presence of the highly acidic SO 3 − ions prevents the dissociation of H 2 CO 3 into charged species (i.e., HCO 3 − and CO 3 2− ).As a result, the level of charged species in the system (i.e., SO 3 − ) is not greatly affected.The increase in pH of PAMPSA solution can be attributed to the increase in the number of neutral H 2 CO 3 molecules being added to the PAMPSA solution via CO 2 gas bubbling, which in turn increases the pH of the solution as shown in Figure 1C.The presence of these neutral H 2 CO 3 molecules impedes the mobility of the ionic charged species (SO 3 − ) and the sensor shows an increase in resistance in the presence of CO 2 gas.To further validate the interaction of the ions with the SO 3 − functional groups a FTIR analysis was conducted (Figure S1, Supporting Information) To confirm the sensor mechanism, the effect of pH on PAMPSA solution's conductivity was measured in the presence of both basic and acidic species.For this study, sodium hydroxide (NaOH) and lactic acid (LA) were selected to produce basic and acidic species, respectively.NaOH was selected to simulate a strong acid-base reaction, whereas LA has a similar activity as bicarbonates and carbonates on dissociation. [16]Therefore, the electrical response and pH of PAMPSA solution (0.1 m) were measured at various concentrations of NaOH and LA and the results were compared with a control sample, Milli-Q water (10 mL).The results in Figure 2A show that upon dropwise addition of 0.1 m NaOH solution over time, the pH of the PAMPSA solution increased, whereas the resistance response decreased from −17% to −50% as shown in Figure 2B.
The interaction of LA with PAMPSA is dependent on the pK a of both solutions.LA was selected mainly because it is a weak and non-volatile acidic compound and the anions generated have the same activity as bicarbonates and carbonates.The pK a of PAMPSA is 1.9 [31] and it is a stronger acid compared with lactic acid with 3.86 pK a .When LA was gradually added to the PAMPSA solution, due to the highly acidic nature of the SO 3 − ions, it prevented the protonation of LA.However, over time with an increase in the volume of LA in PAMPSA solution, the number of SO 3 − per volume decreases, which in turn increases the pH of the solution.As a result, over time the pH of PAMPSA solution increases from 1.26 to 1.85 with the addition of LA as shown in  The same steps were repeated with the control, while its pH showed an increase and decrease in the presence of NaOH and LA, respectively (Figure 2A,C), the conductivity of the control always increased (drop in resistance, Figure 2B,D).This indicates that compared with the control, PAMPSA solution can differentiate between acidic/basic charged species based on the change in resistivity and pH.

Evaluating PAMPSA Sensor Response to CO 2 Gas
Human breath consists of multiple gases/VOCs, and the previous iterations of breath sample analysis involved isolating/separating the different gases before testing the target analyte. [11]In this study, the PAMPSA sensor response to humidified CO 2 gas was evaluated.Three sensors were exposed to a wide range of CO 2 concentrations (10 3 -10 6 ppm) that were measured accurately with a commercial sensor (ExplorIR-W-100) (Figure S2, Supporting Information) in the chamber.The sensor response to CO 2 was a function of the concentration of the interacting groups (SO 3 − and H 2 CO 3 ).As the concentration of CO 2 gas increases, the number of H 2 CO 3 molecules in the sen- sor surroundings increases making the pH of the environment acidic (<7 pH).As PAMPSA has a higher pKa (1.9> 6.37), the presence of highly acidic SO 3 − ions prevents the dissociation of H 2 CO 3 into charged species (i.e., HCO 3 − and CO 3 2− ).As a result, the level of charged species in the system (i.e., SO 3 − ) is not greatly affected.As shown in Figure 3, the slight increase in resistance (≈9% at 100% CO 2 ) could be attributed to the reduced mobility of the free-charged species in the matrix in the presence of neutral H 2 CO 3 molecules.This low sensitivity toward CO 2 is of great value for designing a real-time gas sensor that can selectively detect ammonia gas in the presence of CO 2 for applications such as exhaled breath analysis.In such cases, pre-treating the breath sample would not be necessary and the sensor can analyze the exhaled breath in real-time.

Evaluating PAMPSA Sensor Response to NH 3 Gas
To demonstrate the viability of PAMPSA gas sensor for detecting ammonia concentration accurately in various conditions, we selected a scenario of using this sensor for breath testing.It is critical to differentiate NH 3 concentration for a healthy and patient within the range of 50-1500 parts-per-billion (ppb).We, therefore, measure the PAMPSA sensor's response to humidified NH 3 gas, its response time, long-term stability, and selectivity under different conditions.The PAMPSA loading concentration for the solidstate sensor testing was optimized at 20 weight/volume% (Figure S3A,B, Supporting Information).Figure 4A shows the triplicated response of the PAMPSA sensor over a wide range of NH 3 gas concentrations (10 −2 -10 3 ppm) while using N 2 and CO 2 as the interfering gas.With the increase in the NH 3 concentration, the resistance decreased proportionally.This is because the number of NH 4 + ions that interact with the SO 3 − ions on the sensor's surface increases with an increase in NH 3 gas concentration.The sensor's limit of detection to NH 3 was 30 ppb with mean (triplicated) response of −3.76%, indicating that the sensor could monitor NH 3 levels in a healthy person (≈0.05 -1.5 ppm) and a diseased person (≈1.5 -15 ppm).The sensor exhibited a linear trend from 50 ppb to 1000 ppm, in the presence of N 2 and CO 2 gas with a R 2 value of 0.9566 and 0.9564, respectively (Figure S4A,B, Supporting Information).The sensor's response to NH 3 gas was not affected by N 2 /CO 2 gas, showcasing that for breath analysis, the breath components do not need to be isolated/separated and that the sensor can accurately monitor NH 3 gas in real-time.
Next, we assessed the stand-alone recovery time of the PAMPSA sensor.Herein, the sensor was first exposed to NH 3 gas until the sensor resistance stabilized, before humidified N 2 gas was flushed into the sensor chamber, to get rid of any residue NH 3 gas, until the sensor recovered to its original baseline resistance.The recovery time of a sensor is defined as "the time taken by the sensor to recover to its baseline resistance when the target gas flow is switched off". [32]From Figure 4B we observe that the time required for the sensor to recover to its original baseline resistance is ≈3 min.
Next, we assessed the sensor long term ability to monitor humidified ammonia gas.In this study, three identical sensors were prepared and stored at room temperature (23 °C) for a period of 50 days, and their response to 100 ppm NH 3 was tested every 7 days under identical test conditions.The sensors' response to 100 ppm NH 3 over 50 days was within the response range of −49%-−57.5% based on the calibration curve (Figure 4A).This indicated that the designed sensors had high repeatability and reproducibility over 50 days and could be used for potential long term NH 3 sensing, as shown in Figure 4C.
The most common VOCs present in human breath are methanol, ethanol, and acetone. [33]To further test the selectivity of PAMPSA to different VOCs, an aromatic hydrocarbon (toluene) was selected. [18]In this test, the sensor selectivity to these VOCs and NH 3 was tested under different conditions.The concentration of VOCs was 50 ppm and NH 3 was 1 ppm, respec-tively.The VOCs were introduced into the sensor chamber for 5 min after which the sensor response was recorded.For each VOC, the sensor showed an increase in resistance, and for NH 3 , the sensor showed a decrease in resistance with a response as high as −27% for 1 ppm (Figure 4D).This resistance behavior can be attributed to the VOCs having a pH ≈7 with lower ion mobility and they do not generate any charged species.Hence, when it comes in contact with the sensor surface, it may lead to drying on top of the sensor surface, which leads to a drop in ionic conductivity, and we see an increase in resistance.A statistical analysis (ANOVA) was conducted to see whether the different test conditions influenced the overall sensor performance.The differences in the mean were not statistically significant (ns) as p> 0.05 (Table S1, Supporting Information), indicating that the external test conditions, such as interfering gas, high humidity and change in temperature (Figure S5, Supporting Information), did not influence the sensor's performance.

Blind Test
To reduce bias and validate the sensor's response to ammonia gas, a randomized control trial (blind test) was conducted.Herein, the ability of PAMPSA sensor to accurately predict NH 3 gas in unknown media was assessed.Figure 5A,C showed a decrease in resistance, indicating that NH 3 was present in these solutions.The calculated sensor responses to solution 1 and 3 were −55.4% and −38.6%, therefore, it was estimated to have 100 ppm (−49-−57.5%)and 10 ppm (−31.3-−37.6%)NH 3 , respectively in these solutions (Figure 4A).Based on the sensor's low sensitivity toward CO 2 gas, it was interpreted that solutions 2 and 4 had CO 2 gas (Figure 5B,D, respectively), whereas no gas media was present in solution 5 (Figure 5E).The comparison of these results with the actual formulation of the solutions (Figure 5F) demonstrated that PAMPSA sensor designed in this study can be used for measuring NH 3 concentration accurately.

Evaluating the Sensor Response via a Simulated Breath Test (Demo)
The sensor was integrated into a face mask and then mounted on a human face model, to illustrate the potential of PAMPSA sensors for non-invasive medical diagnostic applications, (Figure S6A, Supporting Information).As exhaled human breath has high contents of water vapor, [34] each gas was bubbled through a water chamber to reach high humidity levels (> 90% RH).In this study, the sensor was exposed to a continuous stream of humidified N 2 /CO 2 gas (0.1 L min −1 ) and alternately was exposed to NH 3 gas periodically for less than 10 s (Figure S6B, Supporting Information).As shown in Figure 6, when NH 3 gas was introduced into the sensor's immediate environment, there was an instantaneous drop in the sensor's resistance, highlighted in yellow.The gas cycle was alternated between N 2 /CO 2 and N 2 /CO 2 /NH 3 and these alternating gas cycles were repeated six times to showcase the sensor's response and recovery in the presence and absence of NH 3 gas.This demo illustrates the sensor's ability for real-time non-invasive detection of ammonia via breath analysis.Due to the sensor's low sensitivity (0.03 ppm), high selectivity, and rapid detection of NH 3 gas, the potential applications of this sensor can be expanded to assessing food quality and safety and environmental monitoring of safe NH 3 gas levels.
The key parameter of this designed sensor and other chemiresistive NH 3 gas sensors for breath analysis reported in the literature are listed and compared in Table 1.Herein, parameters such as the limit of detection, detection range, response time, and effect of humidity on the sensors have been compared.This is because, for medical applications, the sensor should be sensitive, have a fast response time, and be able to function at high humidity levels (>85%). [21,34]
Sensor Fabrication: The electrodes were printed on a PET substrate using carbon black conductive ink via the EnvisionTEC 3D-bioplotter (Manufacturing Series, Germany) as shown in Figure S7 (Supporting Information).The printed interdigitated electrodes had a dimension of 25 mm × 20 mm, with each electrode consisting of two teeth.The carbon black ink was extruded through a 250 μm nozzle, and the printing parameters, including temperature, pressure, and speed of printing, were optimized at 21 °C, 3.2 bar, and 10 mm s −1 .The printed electrodes were dried in an oven at 45 °C overnight.The dried sensor was placed in a plasma chamber to enhance surface hydrophilicity and facilitate the spread of the PAMPSA solution on the substrate.The sensor was exposed to air plasma (Harrick Plasma, USA) for 5 min with a radio frequency power of 45 W to improve surface hydrophilicity (Figure S8A,B, Supporting Information).PAMPSA solution (20 w/v%) was prepared by mixing 2 g of PAMPSA in 10 mL of Milli-Q water and was stirred overnight at room temperature (23 °C).
To achieve an even coating of polymer on the surface of the sensor, spin coating (MuTech micro coater, Argentina) was employed.PAMPSA solution (100 μL) was deposited on top of the sensor and parameters such as velocity, acceleration, and time were optimized to 5000 rpm, 1000 rpms −1 and 30 s, respectively, and an even coating was achieved (Figure S9C, Supporting Information).The prepared spin-coated sensors were left to dry inside the fume hood overnight before the sensors could be used for further experiments, as shown in Figure S7 (Supporting Information).
PAMPSA Sensor Performance in the Gas Phase: The electrical resistance of sensors was measured using a multimeter (2450 SourceMeter, Keithley).To test the response of sensors in the presence of different gases and vapors, the sensor was sealed inside a 220 mL container with 1.5 mL water to maintain the humidity level constant.The gas flow rate from the cylinder was controlled using a mass flow controller (Omega FMA-2600A series, flow rate 0.5-500 mL min −1 for N 2 /NH 3 and 0.5-10 mL min −1 for CO 2 gas).Before each test, the sensor chamber was kept in the oven at 45 °C for 2 h to allow the sensor chamber to reach high humidity levels (> 90% RH) (Figure S10, Supporting Information).As human exhaled breath has high levels of humidity > 90% RH, for all sensor tests that were performed, gases were humidified before entering the sensor chamber.Humidified N 2 gas was flushed into the test chamber for at least 10 min to achieve a stable baseline resistance.The sensor response was calculated as shown in Equation ( 5): where R f denotes sensor resistance in the presence of target gas and R o represents sensor resistance in the presence of humidified N 2 gas.The same formula is used for change in the solution's resistance monitored during the pH test as shown in Equation (6).
Wherein, R o represents the solution resistance in the absence of added dissociated ions and the R f represents the solution resistance in the presence of added dissociated ions.PAMPSA Sensor Selectivity: To test the sensor selectivity via the simulated breath test, the sensor chamber was initially purged with 95% humidified N 2 gas (47.5 mL min −1 ) and 5% humidified CO 2 gas (2.5 mL min −1 ).Once a stable baseline resistance (R o ) was achieved, each target VOC was introduced periodically into the sensor chamber, and the PAMPSA sensor response was recorded.To evaluate selectivity at human body temperature, the temperature inside the sensor chamber was maintained at 37 °C, and the same steps were repeated.For standard conditions, the sensor was tested at room temperature (21 °C), and the chamber was purged with humidified N 2 gas.The selectivity test results under each test condition were triplicated using three different sensors.
Blind Test PAMPSA Sensor Validation: Five different solutions of different concentrations of NH 3 and CO 2 in water were prepared and stored in vials labelled 1, 2, 3, 4, and 5, respectively.The solution concentration was based on the static liquid gas distribution method (Equation S1, Supporting Information). [35]The NH 3 solution vials were prepared by mixing NH 4 OH (0.2 and 0.02 mL, respectively) in water (9.8 mL and 9.98 respectively), while the CO 2 solution vials were prepared by bubbling CO 2 gas (5 and 2 min, respectively) in water (10 mL).For this test five individual sensors were fabricated, solutions were prepared by one of the team members, and the test was conducted by another team member, who did not know the compositions of the solution.
Evaluating the Sensor Response via a Simulated Breath Test (Demo): A PAMPSA sensor was integrated into the M3 face mask by removing the mask filter and replacing it with the sensor, as illustrated in Figure S4A (Supporting Information).The mask was placed on a mannequin and was exposed to simulated breath (humidified N 2 /CO 2 gas) through the mannequin's mouth with trace amounts of NH 3 periodically (Figure S4B, Supporting Information).The flow rate was maintained at 0.1 L min −1 .The resistance was measured continuously.

Figure 1 .
Figure 1.Schematic illustration of the sensor mechanism.A) The PAMPSA-coated sensor interacts with the ammonium ion on the surface of the sensor via electrostatic interaction.B) The change in pH of PAMPSA solution on exposure to humidified NH 3 gas.C) The change in pH of PAMPSA solution on exposure to humidified CO 2 gas.

Figure 2 .
Figure 2. PAMPSA solution interaction with acidic/basic charged species.A) pH and B) the changes in the solution's electrical resistivity due to the addition of NaOH to PAMPSA and water (control).C) Demonstrates the pH and D) the change in solution's electrical resistivity due to the addition of Lactic acid to PAMPSA and water (control), respectively.

Figure 2C .
Figure 2C.The addition of these neutral LA molecules reduces the mobility of free-charged species in the solution, and we observe an increase in solution resistance, Figure2D.The same steps were repeated with the control, while its pH showed an increase and decrease in the presence of NaOH and LA, respectively (Figure2A,C), the conductivity of the control always increased (drop in resistance, Figure2B,D).This indicates that compared with the control, PAMPSA solution can differentiate between acidic/basic charged species based on the change in resistivity and pH.

Figure 4 .
Figure 4.Evaluating PAMPSA sensor performance to NH 3 gas.A) Sensor response to different concentrations of NH 3 gas with N 2 and CO 2 gas as the interfering gas.B) Sensor response time to 1 ppm NH 3 and recovery in the presence of N 2 gas (green region: NH 3 purged into the system; purple region: sensor response recovery in the presence of N 2 gas).C) Sensor long-term stability and response to 100 ppm NH 3 gas.D) Sensor selectivity under different test conditions.Triplicates are reported as mean ± SD.ANOVA did not show any significant (p>0.05)difference among samples VOC/Gas: 1) Acetone, 2) Ethanol, 3) Methanol, 4) Toluene, 5) Ammonia.

Figure 5 .
Figure 5. PAMPSA sensor blind test.A) Sensor response to solution 1. B) Sensor response to solution 2. C) Sensor response to solution 3. D) Sensor response to solution 4. E) Sensor response to solution 5. F) Blind solutions compositions.

Figure 6 .
Figure 6.Evaluating the sensor response via a simulated breath test.A) PAMPSA sensor was integrated into the mask and mounted on a human/mannequin model (left).The sensor's response to alternating N 2 /CO 2 and N 2 /CO 2 /NH 3 gas cycles (yellow region indicates the presence of NH 3 gas).

Table 1 .
Comparison of the current work with other chemiresistive NH 3 sensors.