Sub-ppm electrical detection of hydrogen sulfide gas at room temperature based on printed copper acetate–gold nanoparticle composite films

Abstract

This paper presents the sub-ppm level electrical detection of H₂S gas at room temperature using printed copper acetate–gold nanoparticle composite films. The excellent sensitivity of these films towards H₂S can be attributed to the catalytic activity of gold nanoparticles in combination with the plasma oxidation of copper acetate films.

---

Hydrogen sulfide (H₂S) is a toxic gas which poses a threat to human health. According to the National Institute of Occupational Safety and Health, the H₂S concentration immediately dangerous to life is 100 ppm.¹ On the other hand, the recommended exposure limit is 10 ppm for a maximum duration of 10 min.¹ Contemporary H₂S sensor devices can be divided into three major categories: semiconductor metal oxide sensors, electrochemical sensors, and optical sensors.² Current research aims for the development of robust and cost-effective H₂S sensors with enhanced sensitivity and stability. In addition, sensors should be able to operate consistently in harsh environmental conditions. This can be accomplished through emerging materials science and modern processing technologies. Nano-materials are of particular interest in gas sensing applications because of their high surface to volume ratio. Furthermore, noble metals have been used as oxidation catalysts for enhancing the reaction on a gas sensor surface by means of a superior oxygen dissociation catalytic ability.³ It has been reported in the literature that the sensitivity of the sensing layer towards the analyte gas can be improved by doping the sensing material with metals, such as Pt, Au, Pd and Ag.^4–6 Copper acetate (CuAc) has been recently introduced as an easily processable material which is sensitive to H₂S gas.^7–9 CuAc films have been shown to directly react with H₂S gas to form copper sulfide (CuS).⁹ This resulted in a significant and irreversible change in resistance of the film at room temperature with relatively low (1–20 ppm) H₂S concentrations.^7–9 The large change in resistance is attributed to a direct conversion of highly insulating CuAc (R > 1–100 GΩ) to a p-type semiconducting CuS (R ∼ 10–100 Ω). Chemiresistor-type sensors based on CuAc nanoparticles showing more than eight orders of magnitude change in resistance when exposed to H₂S on ppm level have been previously demonstrated.^7–9 Furthermore it has been shown that robust CuAc-based H₂S sensors can be fabricated on low-cost and flexible substrates by using mass-manufacturing technologies such as inkjet-printing. These inkjet-printed sensors have been successfully employed for quantitative detection of H₂S (1 to 20 ppm).^7,8 In addition, good repeatability, long-term stability, negligible humidity effect (at RH < 80%) and selectivity of printed CuAc-based H₂S sensors have been reported earlier.^7,8 However, the practical applicability of these sensors is still somewhat limited by rather slow response times and deficient sensitivity (to sub-ppm level detection).

This paper reports a CuAc-based chemiresistor-type sensor configuration printed on paper substrate capable of a fast response time and sub-ppm sensitivity towards H₂S gas at room temperature. In addition, these sensors can be regenerated for multiple uses. These improvements in sensing properties were achieved by the oxidative conversion of solid-state CuAc nanoparticle films to Cu₂O (and its cross defect structure Cu₃O₂) nanocrystals using plasma treatment and by incorporation of gold nanoparticles (AuNP) in the printed film. The changes in material properties of CuAc film due to plasma oxidation and the resulting improvement in performance toward H₂S gas detection is explained through detailed analysis of surface chemistry and morphology of the sensing film.

The effect of plasma treatment on the electrical response (change in resistance) of printed CuAc-based sensors was first tested without incorporation of AuNP in the sensing film. Printed integrated electrodes used in these sensors were fabricated using silver nanoparticle ink. Fig. 1A shows typical electric responses of the sensors exposed to 1 ppm H₂S gas concentration at room temperature with and without plasma treatment. The onset time of detection, (the time that it takes for the sensor to give detectable response to the presence of H₂S gas)¹⁰ t_on, and the response time to 90% of full scale (full scale from 100 MΩ to saturation level excluding onset time of detection) t₉₀, were determined from the response curves.¹¹

Fig. 1 The electrical response (change in resistance, R) of printed CuAc-based H₂S sensors with and without plasma treatment using either (A) silver or (B) gold/AuNP electrodes. The sensors were exposed to 1 ppm H₂S gas at room temperature with 45% relative humidity.

Plasma treatment of CuAc film prior to exposure to H₂S gas resulted in considerable improvements of both the t_on and the t₉₀ values of the sensor compared to untreated CuAc film. The t_on value decreased from 105 min to 3 min and t₉₀ from 65 min to 10 min (Fig. 1A). A slow response of the unmodified sensor is related to the time needed for the formation of sufficient CuS particle–particle contacts. During the conversion of CuAc to CuS upon exposure to H₂S, a conductive percolation path is formed between the electrodes due to the formation and growth of CuS crystals (ESI Fig. S1†). On the other hand, the plasma oxidation of CuAc film led to an increase of the particle size on the sensor surface (ESI Fig. S2†). The resulting decrease of the particle–particle distance explains the faster reaction of the sensor towards H₂S. In addition, the plasma oxidized film also showed p-type semiconductor characteristics. However, the film was still poorly conductive and the resistance remained over 100 MΩ. Cross sensitivity and selectivity of the sensor to other sulfur compounds was not tested. However, the several orders of magnitude change in the resistance of the sensors is based on the transformation of insulating CuAc to semiconducting CuS. Compared to H₂S, gaseous organo-sulfur compounds (e.g. dimethyl sulfides, thioalcohol) form complexes with CuAc rather than react with CuAc to form CuS.¹² Negligible change in the resistance is expected with the formation of these complexes, therefore the electrical response of the sensor can be considered to be highly selective to H₂S.

Fig. 1B shows typical responses of the CuAc-based sensors fabricated using inkjet-printed gold electrodes with non-conductive layers of AuNP between the conductive interdigitated electrodes. The fabrication process of these gold/AuNP electrodes is explained in detail in the ESI section (Fig. S3†). The sensor response with and without plasma treatment is shown for 1 ppm H₂S exposure at room temperature (Fig. 1B). The sensor without plasma treatment showed the t_on value of 35 min and the t₉₀ value of 29 min. Plasma treatment of CuAc film improved the response, the t_on value being only 1.4 min and t₉₀ around 4 min. Compared to the results obtained for sensors fabricated on silver electrodes excluding the AuNP layer (Fig. 1A), significant improvements were gained using gold/AuNP electrodes. Although the direct comparison between the two sensors is not possible due to the differences in the geometry of the electrodes (Fig. S3†) especially in the gap width (400 μm for the gold/AuNP electrodes, 100 μm for the silver electrodes), the improvements in the sensor response with the inclusion of AuNP demonstrate qualitatively the metal catalyst-enhanced effect. The metal catalyst-enhanced gas sensing of nanomaterials has been stated to be based on two prevailing mechanisms.^11,13 First, metal catalyst nanoparticles can effectively activate adsorbed oxygen and enrich the surface of the reactive film with oxygen adatoms (O⁻, O²⁻, O₂⁻). This leads to faster oxidation of gas molecules resulting in an enhanced sensing effect. The other mechanism improving the sensitivity involves the dissociation of gas molecules on the catalytic metal clusters and subsequent diffusion across or through the clusters to the film where the gas molecules can interact with the sensing material. In addition to having the fastest response among all the sensor configurations tested, the sensor with plasma treated CuAc film on gold/AuNP electrodes had the capability to detect sub-ppm concentrations of H₂S gas at room temperature (Fig. 2A). Around four orders of magnitude change in resistance was detected with H₂S concentration of 300 ppb with t_on < 60 min. For comparison, no measurable electrical response was obtained from either untreated CuAc film printed on gold/AuNP electrodes or plasma treated CuAc film on silver electrodes when exposed to 300 ppb of H₂S for three hours.

Fig. 2 (A) Electrical response of a plasma treated sensor fabricated on gold/AuNP electrodes towards 300 ppb H₂S gas at room temperature with 45% relative humidity. (B) Regeneration of sensor by plasma after exposure to 1 ppm H₂S.

The plasma treated sensor fabricated on gold/AuNP electrodes could also be regenerated by subsequent plasma treatments (Fig. 2B). After the initial exposure to H₂S gas (1 ppm) and subsequent decrease in resistance, the sensor was again plasma treated for 2 minutes resulting in an increase of resistance of the sensor over 100 MΩ. The second exposure of the sensor towards 1 ppm H₂S resulted again in almost seven orders of magnitude change in the resistance of the sensor. The same procedure was repeated for the third time and similar results were observed showing the regeneration and reusability of the sensor by the plasma treatment. The regeneration did not significantly affect the saturation level reached after H₂S exposure. However, the detection rate of the sensor decreased somewhat after each regeneration cycle, t₉₀ changing from 5 min (initial exposure) to 17 min after the first regeneration cycle and to 23 min after the second regeneration cycle.

A more detailed surface analysis was conducted to study the effects of plasma treatment on the chemistry and properties of the sensor films. The X-ray photoelectron spectroscopy (XPS) survey spectra (ESI Fig. S4†) and XPS high resolution spectra for carbon 1s (ESI Fig. S5†), oxygen 1s (ESI Fig. S6†), copper 2p (Fig. 3B) and sulfur 2p (Fig. 3C) peaks were measured. The plasma treatment had a marked effect on the relative amounts of carbon and copper elements, the former decreasing and the latter increasing (Fig. 3A). On the other hand, the relative amount of oxygen remained approximately the same. Trace amount of chlorine was also detected, most probably present in the paper substrate.

Fig. 3 (A) XPS elemental composition and atomic concentration (at%) of the detected elements. (B) XPS high resolution spectra of Copper 2p^3/2 peak for different samples. The inset shows the peak for ref on an adjusted scale. (C) XPS high resolution spectra of Sulfur 2p for the studied samples.

High resolution spectra of Cu 2p^3/2 showed that the peak had shifted to a lower binding energy as a result of the plasma treatment. In addition, the satellite peak observed for the untreated CuAc film (Fig. 3B inset) was absent. This indicates that the oxidation state of Cu had changed from +2 to +1.⁷ It has been previously shown that reduction of the copper acetate precursor by thermal oxidation (in air) leads to the formation of pure Cu metal and subsequent oxidation to Cu₂O.¹¹ In addition, low temperature oxidation (such as plasma) of Cu has been shown to lead to the formation of Cu₂O and its cross defect structure Cu₃O₂.¹⁴ The Cu/O ratio of 1.6 suggests that the Cu₃O₂ phase is the major oxidation product of CuAc after the plasma treatment.

After exposure to 1 ppm H₂S gas, sulfur was detected in abundance at the surface of the sensor with a clear S 2p peak visible in the high resolution spectrum (Fig. 3C). In addition, the amount of oxygen decreased compared to the plasma treated unexposed sensor (Fig. 3A). For comparison, XPS was also measured after exposure to 10 ppm H₂S gas. This led to even higher amounts of sulfur and more profound decrease of oxygen (Fig. 3A). In addition, a clear shift in the Cu 2p^3/2 peak position to a lower binding energy was observed (Fig. 3B), matching the previously observed transition of CuAc to CuS.³ For 1 ppm exposure, the peak position remained at the same binding energy as for the plasma treated unexposed sensor. It has been previously shown that Cu₂O particles suspended in Na₂S solution or exposed to low concentration of H₂S gas are immediately converted into Cu₂O/CuS core–shell structures at room temperature.^15,16

The core–shell structure is formed by a replacement reaction between the oxygen atom in Cu₂O lattice and free sulfur or sulfide ion (S⁻) in solution.¹⁷ A very thin shell layer of CuS could explain why copper–sulfur bond was not detected in XPS high resolution spectrum of copper after exposure to 1 ppm H₂S. In addition, a thin CuS layer in core–shell structure might also facilitate the easier conversion of CuS to Cu₂O by plasma treatment enabling the regeneration of the sensor compared to the situation where CuAc is fully converted to CuS.¹⁴

The catalytic oxidation of H₂S by the oxygen adatoms present at the surface has been shown to be the prominent factor for the observed electrical changes in the H₂S sensors based on (p-type) semiconducting copper oxide.^13,14,18,19 There are two main reaction pathways for catalytic oxidation of H₂S.²⁰ Generally, at higher temperatures, the oxidation of H₂S results in the formation of sulfur dioxide and subsequent release of electrons. The electrons recombine with holes in the p-type semiconductor surface, resulting in an increase of resistance of the system. The change in the resistance is reversible. At room temperature the main product of the oxidation of H₂S is elemental sulfur.²⁰ This leads to the spontaneous formation of a CuS layer on the surface of Cu₂O film causing an irreversible decrease in resistance as shown here.¹⁸ In addition, the presence of water has been shown to play an important role in the oxidation process by enhancing the performance of catalyst/promoter for the low temperature oxidation.²⁰

Conclusions

Significant improvements on the onset time of detection and reaction kinetics were observed for the paper-based H₂S sensors based on a printed plasma-treated CuAc thin film. Further improvements were gained by incorporation of catalytic AuNP into the sensing layer. In addition, the combination of plasma treatment and the use of AuNP enabled the sub-ppm detection of H₂S and the regeneration of sensor by subsequent plasma treatment. The improvements on the H₂S gas sensing over the untreated CuAc film can be explained by two possible mechanisms. Firstly, the formation of a p-type semiconducting Cu₂O (and its cross defect structure Cu₃O₂) layer by the plasma oxidation of CuAc film facilitates the growth of particles and the formation to sufficient particle–particle contact before exposure to H₂S gas. Secondly, the oxidation of H₂S to sulfur by oxygen adatoms could lead to formation of Cu₂O (Cu₃O₂)/CuS core–shell structure where a thin layer of CuS facilitates the rapid formation of a conductive percolation path between electrodes. This leads to a decrease of resistance of the sensor with improved onset time and rate of detection compared to the untreated CuAc film. The above improvements enable the use of printed paper-based H₂S sensors which are suitable for mass manufacturing in applications requiring both fast as well as sub-ppm detection.

Acknowledgements

The authors would like to thank Imerys Minerals Ltd., UK, Paramelt B.V., NL, StoraEnso, FI, and Basf, DE for providing the materials used in the multilayer coated paper. Academy of Finland (grants 118650, 137093 and 141115), the European Regional Development Fund in South Finland and the Academy of Finland grant through the National Center of Excellence Programme are acknowledged for the financial support.

Notes and references

1. H. Xua, J. Wub, C. Chena, L. Zhangb and K. Yanga, Sens. Actuators, B, 2010, 143, 535–538.
2. S. Pandey, K. Kim and K. Tang, Trends Anal. Chem., 2012, 32, 87–99.
3. C. Wang, L. Yin, L. Zhang, D. Xiang and R. Gao, Sensors, 2010, 10, 2088–2106.
4. Y. Lu, J. Li, J. Han, H. Ng, C. Binder, C. Partridge and M. Meyyappan, Chem. Phys. Lett., 2004, 391, 344–348.
5. D. Wang, Z. Ma, S. Dai, J. Liu, Z. Nie, M. Engelhard, Q. Huo, C. Wang and R. Kou, J. Phys. Chem. C, 2008, 112, 13499–13509.
6. D. Haridas, K. Sreenivas and V. Gupta, Sens. Actuators, B, 2008, 133, 270–275.
7. J. Sarfraz, A. Määttänen, P. Ihalainen, M. Keppeler, M. Lindén and J. Peltonen, Sens. Actuators, B, 2012, 173, 868–873.
8. J. Sarfraz, P. Ihalainen, A. Määttänen, T. Gulin, J. Koskela, C.-E. Wilén, A. Kilpelä and J. Peltonen, Sens. Actuators, B, 2014, 191, 821–827.
9. S. Virji, R. Kaner and B. Weiller, Inorg. Chem., 2006, 45, 10467–10471.
10. K. Benkstein and S. Semancik, Sens. Actuators, B, 2006, 113, 445–453.
11. A. Kolmakov, D. Klenov, Y. Lilach, S. Stemmer and M. Moskovits, Nano Lett., 2005, 5, 667–673.
12. Gaylord Chemical CompanyDimethyl Sulfide Reaction Solvent guide, Bulletin # 203B, 2007.
13. M. Yin, C.-K. Wu, Y. Lou, C. Burda, J. T. Koberstein, Y. Zhu and S. O'Brien, J. Am. Chem. Soc., 2006, 127, 9506–9511.
14. K. Santra, C. K. Sarkar, M. K. Mukherjee and B. Ghosh, Thin Solid Films, 1992, 213, 226–229.
15. S. Jiao, L. Xu, K. Jiang and D. Xu, Adv. Mater., 2006, 18, 1174–1177.
16. P. Wijesundara, D. Perera, D. Jayasuriya, W. Siripala, L. De Silva, P. Samantilleke and M. Dharmadasa, Sol. Energy Mater. Sol. Cells, 2000, 61, 277–286.
17. H. Zhu, J. Wang and D. Wu, Inorg. Chem., 2009, 48, 7099–7104.
18. D. Zappa, E. Comini, R. Zamani, J. Arbiol, J. R. Morante and G. Sberveglieri, Sens. Actuators, B, 2013, 182, 7–15.
19. J. Chen, K. Wang, L. Hartman and W. Zhou, J. Phys. Chem. C, 2008, 112, 16017–16021.
20. A. Primavera, A. Trovarelli, P. Andreussi and G. Dolcetti, Appl. Catal., A, 1998, 173, 185–192.

---

Footnote

† Electronic supplementary information (ESI) available: X-ray diffraction data, atomic force microscopy images before and after plasma treatment of CuAc film, the printing of interdigitated gold electrodes with AuNP between the fingers and materials and methods. See DOI: 10.1039/c4ra17256f

---