Open access peer-reviewed chapter

Palladium (II) Oxide Nanostructures as Promising Materials for Gas Sensors

By Alexander M. Samoylov, Stanislav V. Ryabtsev, Vasily N. Popov and Petre Badica

Submitted: July 7th 2017Reviewed: November 9th 2017Published: April 18th 2018

DOI: 10.5772/intechopen.72323

Downloaded: 1042


One of the most important environment monitoring problems is the detection of oxidizing gases in the ambient air. Negative impact of noxious oxidizing gases (ozone and nitrogen oxides) on human health, sensitive vegetation, and ecosystems is very serious. For this reason, palladium (II) oxide nanostructures have been employed for oxidizing gas detection. Thin and ultrathin films of palladium (II) oxide were prepared by thermal oxidation at dry oxygen of previously formed pure palladium layers on polished poly-Al2O3, SiO2/Si (100), optical quality quartz, and amorphous carbon/KCl substrates. At ozone and nitrogen dioxide detection, PdO films prepared by oxidation at T = 870 K have demonstrated good values of sensitivity, signal stability, operation speed, and reproducibility of sensor response. In comparison with other materials, palladium (II) oxide thin and ultrathin films have some advantages at gas sensor fabrication. Firstly, for oxidizing gas detection, PdO films with p-type conductivity are more perspective than the material with n-type conductivity. Secondly, at ambient conditions, palladium (II) oxide is insoluble in water and does not react with it. These facts are favorable for the fabrication of gas detectors because they make possible to minimize the air humidity influence on PdO sensor response values. Thirdly, the synthesis procedure of PdO films is rather simple and is compatible with planar processes of microelectronic industry.


  • palladium (II) oxide
  • nanostructure
  • gas sensor
  • ozone
  • nitrogen dioxide

1. Introduction

Nowadays, the detection of oxidizing gases in the ambient air is one of the most important environment monitoring problems for industrialized countries. During the last 25 years, the steady increase in concentration of nitrogen dioxide and tropospheric (low level) ozone is observed. As it is known, three out of six common air pollutants (also called “criteria pollutants”) are oxidizing gases: sulfur dioxide, nitrogen oxides, and tropospheric ozone [1, 2]. One part of ecologists is sure that increase in the content of low-level ozone in atmospheric air is caused mainly by an intensification of industrial production, motor and air transport. Undoubtedly, ozone gas is applied in many fields such as food, pharmaceutical, textile, and chemical industries, water treatment, and purification of gases. However, there is an opinion that emergence of tropospheric ozone in ambient air is a consequence of the “greenhouse” effect [3].

Under sunlight, the interaction of ozone, nitrogen oxides, and volatile hydrocarbons can produce many toxic organic compounds (Figure 1). By the action of sunlight, oxygen atoms freed from nitrogen dioxide attack oxygen molecules to make ozone. Nitrogen oxide can combine with ozone to reform nitrogen dioxide, and the cycle repeats.

Figure 1.

The chemical reactions of tropospheric ozone under sunlight.

Moreover, at interaction with ozone, the ultraviolet component of sunlight leads to the formation of excess quantity of the reactive oxygen species (ROS): oxygen ions, free radicals, and peroxides. In living bodies, even the trace amounts of ROC can provoke an oxidative stress. For human, the oxidative stress is a reason for atherosclerosis, hypertension, Alzheimer’s disease, diabetes, and geromorphism [4, 5, 6, 7, 8, 9]. The negative impact on human health of an aspiration of noxious oxidizing gases (ozone and nitrogen oxides) is more serious, particularly for children, the elderly, and people who suffer from lung diseases [12]. Nitrogen oxides and tropospheric ozone can also have harmful effects on sensitive vegetation and ecosystems [10, 11, 12, 13].

For these reasons, various types of the binary, ternary and quaternary metal-oxide semiconductors have been widely applied for oxidizing gas detection. In most cases, for this purpose, the n-type semiconductors such as SnO2 [14, 15, 16, 17, 18, 19, 20], ZnO [21, 22, 23, 24, 25, 26], WO3 [27, 28, 29, 30, 31, 32], In2O3 [33, 34], and TiO2 [35] are used traditionally. In recent years, the search of the materials, which would be capable to lower the detection limit of oxidizing gases, became more active.

The study of palladium (II) oxide nanostructures as materials for gas sensor fabrication was started only since 2014. The assumption to use palladium (II) oxide, which is a p-type semiconductor with the energy band gap ΔEg = 2.2–2.7 eV [36, 37, 38], as the material for the detection of toxic and highly inflammable gas in ambient air has not been accidental for some reasons [39, 40]. Firstly, for a long time, palladium and its compounds in (+2) oxidation state were exploited as very effective catalysts for oxidation reactions of hydrocarbons, including automobile catalytic converters and the catalytic combustion of methane in advanced gas turbines. In catalytic converters, the key processes are the complete oxidation of any hydrocarbon in the exhaust gas stream, the simultaneous oxidation of carbon monoxide, and reduction of nitrogen oxides. Secondly, due to the extremely high catalytic activity, palladium and palladium (II) oxide were applied as additives to improve gas-sensing performance of tin dioxide SnO2 to a wide range of gases [41, 42, 43]. Thirdly, the opinion that long recovery process and high stability could be referred to the main disadvantages of the oxidizing gas sensors based on tin dioxide has been expressed earlier [44, 45]. Fourthly, the metal oxide semiconductors with p-type conductivity are more perspective for oxidizing gas detection than the materials with n-type conductivity. In this case, the chemical adsorption of oxidizing gas molecule on p-type semiconductors surface leads to decrease in the sensor resistance that has simplified the detection process [35, 36]. Thus, at oxidizing gas detection, palladium (II) oxide nanostructures should demonstrate the increase in sensor response value in comparison with the traditional n-type conductivity materials using for the same purpose.


2. Fabrication of palladium (II) oxide nanostructures

Initially, the sensing properties to oxidizing gases of palladium (II) oxide nanostructures were tested on ultrathin and thin films at detection of ozone and nitrogen dioxide [46, 47]. The procedure of PdO thin and ultrathin films synthesis was realized by two stages. First, the initial palladium films (thickness 5–30 nm) were formed by thermal sublimation of palladium foil (purity is 99.99%) in high vacuum chamber evacuated to 5 × 10−7 Torr using a turbo molecular pump. In vacuum chamber, the condensation of Pd metal vapors was performed on different substrates: SiO2/Si (100), Si (100), optical quality quartz, and KCl (100) with buffer layer of amorphous carbon (Figure 2). The values of tungsten heater temperature in order to fabricate initial palladium films with average rate within interval 0.01–0.016 nm per second were determined as a result of Pd films cross-sections by high-resolution scanning and transmission electron microscopy (HR STEM) study.

Figure 2.

High-resolution TEM image of Pd/SiO2/Si (100) heterostructure cross-section prepared by focused ion beam (FIB) technique.

The substructure of initial palladium layers was studied by an X-ray analysis and the HEED method. As it is shown in Figure 3a and 3b, the initial Pd films were polycrystalline and highly dispersive with random orientation of grains irrespective of the substrate nature (SiO2/Si (100), optical quality quartz, and amorphous carbon/KCl). The analysis of bright-field (Figure 3c) and dark-field (Figure 3d) TEM images proves that palladium crystalline grains form a continuous coating without an axial texture with very low density of micropores. On bright-field image, the light contrast (Figure 3c) testifies to the decrease of film thickness at grain borders [48].

Figure 3.

Experimental results of initial Pd films (thickness ~10 nm) crystal structure study: (a) XRD patterns of Pd film deposited on Si (100) substrate; (b) HEED patterns of Pd film deposited on amorphous carbon/KCl substrate; (c) bright-field TEM image of Pd film deposited on amorphous carbon/KCl substrate; (d) dark-field TEM image of Pd film deposited on amorphous carbon/KCl substrate.

Prepared Pd nanostructures on different substrates were annealed at dry oxygen atmosphere for 1 h for layers with thickness 5–15 nm and for 2 h for layers with thickness 30 ± 5 nm at temperatures Tox = 510, 570, 670, 770, 870, and 1070 K. The dehumidification of oxygen at pressure 120–130 kPa (1.2–1.3 Bar) was carried out by gas flow passage through gas bubbler with concentrated sulfuric acid and further through silica tube full of ground zeolite [48].


3. Phase composition and crystal structure of palladium (II) oxide nanostructures

X-ray diffraction (XRD) patterns of samples prepared by oxidation of Pd films on SiO2/Si (100) wafers at dry oxygen atmosphere at Tox = 510, 570, 770, 870, and 970 K are shown in Figure 4.

Figure 4.

X-ray diffraction patterns of palladium film deposited on SiO2/Si (100) substrate after oxidation in dry oxygen at different temperatures:aTox = 510 K;bTox = 570 K;cTox = 770 K;dTox = 870 K; andeTox = 970 K.

It is necessary to note that in Figure 4, the values of XRD reflex intensities are presented in a logarithmic scale because the intensity of Si (400) peak practically exceeds the intensity of palladium and palladium (II) oxide peaks by two orders of magnitude owing to a small thickness of the prepared films. The comparison of the as-grown Pd films XRD patterns with XRD patterns of Pd film after the annealing at Tox = 510 K (Figure 4a) did not reveal any quality changes. The increase in intensities of palladium reflexes was found only.

Thus, it has been established that the annealing of Pd layers at Tox < 570 K (Figure 4a) did not result in the change of their phase constitution. The annealing at Tox = 570 K resulted in the formation of two phase films (Figure 4b). XRD patterns have shown the presence of Pd with PdO simultaneously.

According to XRD results, the rise of the oxidation temperature up to Tox = 770 K and Tox = 870 K led to the formation of the homogenous polycrystalline PdO films. It has been determined (Figure 4c and 4d) that palladium (II) oxide films were characterized by tetragonal crystal lattice (space group P42/mmcand PtS structure type). XRD patterns show (Figure 4c4e) that the peaks became sharper and higher with the oxidizing temperature increasing from Tox = 770 K up to Tox = 970 K. Moreover, the peaks of palladium (II) oxide prepared by oxidation at Tox = 970 K are much sharper and higher than those for films oxidized at Tox = 870 K. This fact can be interpreted as one of the evidences of the crystalline perfection enhancement of palladium (II) oxide films and the grain size enlargement with the increase in the oxidation temperature.

High energy electron diffraction (HEED) technique was used as an alternative method to study PdO film phase composition (Figure 5).

Figure 5.

HEED patterns (a) and bright-field TEM image (b) of PdO film after oxidation atTox = 870 K.

Table 1 compares the results of X-ray analysis (layers on SiO2/Si substrates), the HEED method (layers on optical quartz and Al2O3 substrates), and TEM micro diffraction (layers on amorphous carbon/KCl). An examination of the data presented in Table 1 shows that the X-ray analysis, HEED method, and TEM micro diffraction gave the identical results for the films oxidized at temperatures Tox = 510, 570, 770, 870, and 970 K. The results of these two methods confirm that: (1) the annealing of Pd films at Tox = 510 K does not induce the changes in their phase composition; (2) after annealing of Pd films at the Tox = 570 K, the partial oxidation takes place and gives two phase samples – a mixture of Pd and PdO; and (3) after the annealing of palladium layers at Tox = 770–970 K, the total oxidation gives homogeneous PdO films [48].

Oxidation temperature Tox, KPhase composition
X-ray analysisHEEDTEM microdiffraction
570Pd + PdOPd + PdOPd + PdO

Table 1.

Results of phase composition study of Pd films after oxidation at T = 500–970 K.


4. Electrical properties of palladium (II) oxide nanostructures

The type of conductivity of PdO films synthesized at Tox = 770–970 K was determined by the Seebeck effect study and by calculation of the electromotive force Eemf values:


where Sis thermo-power (Seebeck coefficient), and Tis the temperature gradient. Copper-constantan thermocouples have been used to measure a temperature difference.

Experimental values of Eemf have proved the p-type conductivity for all homogeneous palladium (II) oxide films (Figure 6). The fact of p-type conductivity of PdO bulk samples was reported in the previous publications [36, 37]. The values of thermo power (Seebeck coefficient) Shave been calculated using the Eq. (1). Depending on thickness of palladium oxide films and oxidation temperature, the Seebeck coefficient values changed within the limits from +120 to +220 μV/K. The relative error at thermo-power measurement did not exceed 7%.

Figure 6.

Electromotive forceEemf dependence upon the temperature gradient for PdO film prepared by oxidation atTox = 870 K (thickness ~ 30 nm).

In view of p-type conductivity, palladium (II) oxide films are characterized with the cation deficiency regarding the stoichiometric 1:1 ratio. Thus, for PdO, the Kröger-Vink defect reactions can be written as follows:

  1. with the cations in deficiency on the lattice sites:


  2. with the anions in excess on the interstitial sites:


The results obtained in the present work correlate with the capacitance voltage characteristics of PdO films on silicon [49]. Previously it was found that within the band gap of PdO films, one single energy state is realized only [49]. Therefore, only one type of point defects, which have generated holes, dominates in palladium (II) oxide films. The experimental study of the point defects nature will be the subject of further investigations.


5. Gas sensor properties of palladium (II) oxide nanostructures

Ozone and nitrogen dioxide sensitivity has been measured using the specially fabricated test samples of gas sensors based on thin and ultrathin PdO films oxidized at Tox = 870 K. During sensor response measurements of PdO, ultrathin and thin films prepared by oxidation at Tox = 873 K, synthetic air, calibrated gas mixtures with fixed nitrogen dioxide concentration, and standardized ozone generator produced by Optec were used. The ozone gas was generated by oxidizing oxygen molecules of synthetic air (SA) by a pen-ray ultraviolet (UV) lamp calibrated to give the O3 concentration range between 0.03 ppb and 800 ppb. The synthetic air containing ozone was blown directly on the sensor placed on the top of holder within the test chamber. The operating temperature Td of the sensor ranging from room temperature to 670 K was controlled by chromel-alumel thermocouple. The measurement started after the sample resistance achieved a steady value [50, 51].

Sensor response Swas determined as the ratio of the sensor resistance in synthetic air R0 to the sensor resistance in gas R:


The measurements of NO2 and O3 concentration were performed in flow path conditions with the rates of 300 cm3 per minute and 2.4 dm3 per minute, respectively. The gas flow rate was measured by controllers produced by Bronkhorst.

As it possible to see in Figures 7 and 8, at rather low operation temperature Td, the sensors based on thin (Td = 490 K) and ultrathin (Td = 448 K) PdO films show good sensitivity to rather low concentrations of ozone. Figures 7 and 8 show that at process of ozone quantitative detection at SA atmosphere within concentration interval 100–250 ppb, palladium (II) oxide films have demonstrated high values of sensor response, signal stability, and reproducibility of sensor response also. This fact was proved by the results of multiple measurement cycles with the same O3 concentrations. PdO films with thickness of about 35 nm are characterized by higher values of sensor response (on the average in 7–8 times) in comparison with ultrathin films at the same ozone concentrations (Figure 9). It is possible to explain this fact that the contribution in integrated conductivity of near-surface layers with high defects density is essentially higher for ultrathin PdO films than for films with thickness of about 35 nm (Figure 9). It is necessary to emphasis that the established feature demands a detailed study.

Figure 7.

Time dependence of PdO ultrathin film (thickness ~ 10 nm) sensor resistanceRat ozone different concentrations (operation temperatureTd = 448 K).

Figure 8.

Time dependence of PdO thin film (thickness ~ 35 nm) sensor resistanceRat ozone different concentrations (operation temperatureTd = 490 K).

Figure 9.

Dependence of PdO ultrathin and thin film sensor responseSat ozone different concentrations: 1—Ultrathin film (thickness ~ 10 nm, operation temperatureTd = 448 K) and 2—Thin film (thickness ~ 35 nm, operation temperatureTd = 490 K).

It has been established that PdO thin and ultrathin film sensors gave the stable signal, and the resistance values reliably returned to the baseline at SA atmosphere [50, 51]. It is necessary to note that the recovery period is quite long (600–700 s). It is necessary to note that the similar sensor behavior is typical for other materials used oxidizing gas detection. Usually in this case, the long recovery period is explained by the absence of oxidizing gas immediate interaction with oxygen molecules adsorbed on sensor material surface. At reducing gas detection, the direct interaction with oxygen molecules takes place; therefore, the recovery time is quite short. Moreover, the recovery time depends significantly on the operating temperature.

The sensitivity of palladium (II) oxide ultrathin films to nitrogen dioxide (another toxic oxidizing gas) has also been tested (Figure 10). As it can be seen in Figure 10, at the process of NO2 quantitative detection within concentration interval 500 ppb–200 ppm, PdO ultrathin films have demonstrated good values of sensor response, signal stability, and reproducibility of sensor response [51]. It is necessary to note that the recovery period at NO2 detection is longer than that at O3 detection (Figures 7, 8, and 10).

Figure 10.

Time dependence of PdO ultrathin film (thickness ~ 10 nm) sensor resistanceRat nitrogen dioxide different concentrations (operation temperatureTd = 448 K).

During the determination of ozone (concentration ϕ = 100 ppb) and nitrogen dioxide (concentration ϕ = 10 ppm), the temperature dependences of PdO ultrathin film sensor response Sare presented in Figure 11.

Figure 11.

Dependence of PdO ultrathin film (thickness ~ 10 nm) sensor responseSupon the operation temperatureTd at detection of ozone (O3 concentration 0.1 ppm) and nitrogen dioxide (NO2 concentration 10 ppm).

It is found that within interval of operation temperature 323 < Td < 623 K, the maximum values of response Shave been observed at Td = 448 K (NO2 detection) and at Td = 490 K (O3 detection).As it can be seen in Figure 11, approximately equal values of sensor response Sof palladium (II) oxide films are realized at different concentration of oxidizing gases: φ(O3) = 0.1 ppm and φ(NO2) = 10 ppm.


6. Discussion

Data presented in Table 2 show that physical properties (molecular mass, electric dipole moment) and chemical properties (structure of molecule, high oxidative activity) of detected gases are very similar. According to experimental evidence from microwave spectroscopy, ozone and nitrogen dioxide are bent molecules with C2v symmetry (Table 2). O3 and NO2 are the polar molecules with a dipole moment of 0.66 and 0.39 D, respectively.

MoleculeMolar mass M, g × mol−1Space group symmetryMagnetic PropertiesMagnetic susceptibility χ × 106, cm3/molDipole moment μ × 1030, C·mPEL, ppm
O348.00C2vDiamagnetic+ mg/m3)
NO246.0055C2vParamagnetic+ mg/m3)

Table 2.

Physicochemical properties and permissible exposure limit (PEL) of ozone and nitrogen dioxide [52, 53, 54, 55, 56].

Data in Table 2 show that ozone and nitrogen dioxide molecules essentially differ with magnetic properties only. Ozone is diamagnetic, which means that its electrons are all paired. Unlike ozone, the ground electronic state of nitrogen dioxide is a doublet state. Owing to since nitrogen atom has one unpaired electron NO2 molecule is paramagnetic.

Nevertheless, at detection of ozone and nitrogen dioxide, the temperature that has matched the maximum values of sensor response differs only 25°. Thus, there is prerequisite for the increase in selectivity of palladium (II) oxide sensors at O3 and NO2 detection after studying in detail that oxidation procedure conditions influence on microstructure and stoichiometry deviation.

Moreover, at ambient conditions, palladium (II) oxide is insoluble in water and does not react with it. As the bottom sediment, the palladium (II) hydroxide is formed only at interaction of soluble palladium (II) salt and alkali [52]. These facts are favorable for fabrication of gas detectors because they make possible to minimize the air humidity influence on PdO sensor response values.

To estimate such perspective for palladium (II) oxide nanostructures, we have allowed the speculative extrapolation of experimental data to the point that corresponds to zero ozone concentration (Figure 12). At ozone concentration φ= 10 ppb, which corresponds to 0.1 × PEL (permissible exposure limit), sensor response Swould be about 2 (S~ 2 is an open circle in Figure 12). The extrapolated sensitivity value at concentration φ(O3) = 0.1 × PEL arouses hope that palladium (II) oxide films will be used in fabrication of ozone sensors.

Figure 12.

Dependence of PdO sensor responseSupon the ozone concentration in synthetic air (operating temperatureTd = 490 K (220°C).

At interaction with PdO surface ozone molecules are more active than nitrogen dioxide ones. This interaction is accompanied by more essential increase in the hole density of palladium (II) oxide ultrathin films. In general case, the surface interaction of PdO nanostructures can be written within the framework of Kröger-Vink notation:

PdPd×+ OO×+ O3(gas)PdPd×+ OO×+ O″i+ 2h··+O2(gas)E5

According to Eq. (5), oxygen atom is integrated with palladium (II) oxide structure and O2 molecule is desorbed from the surface. As result of this reaction (5), two holes are formed.

From this point of view, it is possible to explain high efficiency of palladium (II) oxide films at ozone detection. The attempt to distinguish the real reason of PdO nanostructures’ higher sensitivity to ozone, it should be looked for ozone’s extremely high oxidizing ability. As it can be seen in Table 2, the PEL value of ozone is smaller than the similar characteristic of nitrogen dioxide by 50 times practically. This fact is the indirect evidence of ozone-exclusive oxidizing activity. The difference in sensitivity of palladium (II) oxide nanostructures at ozone and nitrogen dioxide detection will be a subject of the subsequent experiments and discussions. Now, it is possible to designate the direction of these future researches only. It is reasonable to assume that under ozone molecules impact, the metastable nanoclusters are formed on the surface of PdO, in which the oxidation states of palladium are higher than (II), for example, (III) or (IV).


7. Conclusion

The results of X-ray analysis, HEED, and HR TEM have demonstrated the possibility of the synthesis of homogeneous nanocrystalline thin and ultrathin films of palladium (II) oxide on different substrates. The very first examinations of sensitivity to different nitrogen dioxide and ozone concentration at rather low operating temperature have shown the high values of sensor response, signal stability, operation speed, and reproducibility of PdO films sensor response. The possibility of work at quite low temperatures will allow decreasing in the energy consumption of the analytical instruments. The detection of O3 and NO2 by palladium (II) oxide sensors can be applied in the fields of the human health and environment protection. Because the synthesis procedure is rather simple and compatible with planar processes of the microelectronic industry PdO nanostructures have a good perspective to be one of the main materials for commercial fabrication of oxidizing gases (ozone, nitrogen dioxide, chlorine etc.) sensors.

© 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Alexander M. Samoylov, Stanislav V. Ryabtsev, Vasily N. Popov and Petre Badica (April 18th 2018). Palladium (II) Oxide Nanostructures as Promising Materials for Gas Sensors, Novel Nanomaterials - Synthesis and Applications, George Z. Kyzas and Athanasios C. Mitropoulos, IntechOpen, DOI: 10.5772/intechopen.72323. Available from:

chapter statistics

1042total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Functionalized Carbon Nanomaterials in Drug Delivery: Emergent Perspectives from Application

By Nabanita Saikia

Related Book

First chapter

Degradation of Toxic Organic Contaminants by Graphene Cathode in an Electro‐Fenton System

By Liu Shuan, Zhao Xia, Jiang Xin, Zhao Xiaorong, Pu Jibin, Wang Liping and Zhou Kaihe

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us