Electrically Conductive Gold Films Formed By Sintering of Gold Nanoparticles at Room Temperature Initiated by Ozone

Understanding and controlling the sintering behaviour of gold nanoparticles is important in the �eld of ligand-protected nanoparticles for their use as precursors for thin �lm fabrication. Lowering the temperature of the sintering event of gold nanoparticles by facilitating desorption of the ligand through oxidation can provide compatibility of sintered gold nanoparticle thin �lms onto heat sensitive substrates. Here we examine the processes by which 1-butanethiol-protected gold nanoparticles sinter under an ozone-rich environment. Upon heating, an ozone-rich environment signi�cantly reduces the temperature of the sintering event when compared to sintering under ambient conditions. At room temperature, exposure to an ozone-rich environment induces sintering over a period of 2.5 hours. Upon exposure to ozone, the surface-bound butanethiyl ligands are oxidized to 1-butanesulfonic acid which facilitates sintering.


Introduction
Sintering of gold nanoparticles (AuNPs) is an attractive technique to form continuous, electrically conducting thin lms [1,2].Using this method, low cost printable electronic devices such as thin-lm transistors [2], eld effect transistors [3], and contacts [2,4] can be fabricated.Formulations that sinter AuNPs at relatively low temperatures provide great exibility and enable AuNP lms to be applied to low melting point polymers or other heat-sensitive substrates.AuNPs can sinter if there is su cient energy to overcome the activation energy barrier provided by the stabilizing ligands [5].One method to lower the activation energy barrier, and thus the temperature of the sintering event (T SE ), is to oxidize the capping ligands, which facilitates desorption from the gold surface.
The bond between gold and the thiolate stabilizing ligand may be described as a gold-thiyl interaction (where thiyls are species with the RS• structure) [6].These interactions can be degraded upon exposure to an oxidizer, to UV light, or at elevated temperatures [7][8][9][10][11][12].Alkanethiolate self-assembled monolayers (SAMs) on gold have been shown to oxidize to alkanesul nates and alkanesul des under ambient conditions [10,[13][14][15].Scanning tunnelling microscope images of air-oxidised decanethiol SAMs on Au(111) shows that after a two-week exposure to ambient conditions, some conversion to decanesulfonate occurs [16].Density functional theory calculations revealed that under oxidising conditions (surface oxide or ozone) thiolate groups on Au (111) surfaces might form sulfoxide derivatives (R 2 S = O), with sul nate (RS(= O)O − ) and sulfonate (RS(= O) 2 O − ) derivatives formed if active oxygens are further supplied [17].
Of particular relevance to the current work, exposure of thiol-bound SAMs on gold to ozone has been shown to oxidize the sulphur atoms [11,18,19].With regard to AuNPs, ozone can reduce the a nity of thiol-based ligands to the gold core in supported AuNPs, allowing for removal of the ligands by washing with water [20].X-ray photoelectron spectroscopy (XPS) analysis of the ozone treated AuNPs revealed that the sulphur atoms had been oxidized upon exposure [20].Previous work has also utilized nitrogen dioxide to oxidise the thiolate stabilizing ligands of AuNPs which subsequently sintered at room temperature [5].In this work, we examine the sintering of thiol-stabilized AuNPs using ozone, a common and readily generated gas.We show that in an ozone-rich environment, the ligands surrounding AuNPs are oxidised and signi cantly reduce T SE to produce gold lms.Importantly, we investigate the organic pro le of the desorbed ligand to determine the processes that occur upon sintering.These ndings shed a new light on advancing the use of gold lms for heat sensitive substrates.Experimental General 1-Butanethiol, Sodium 1-butanesulfonate, tetraoctylammonium bromide, sodium borohydride, methanol, acetonitrile, deuterated chloroform (CDCl 3 ) and deuterated dimethyl sulfoxide (DMSO-d 6 ) were purchased from Sigma-Aldrich and used as received.Toluene (ChemSupplyAustralia), chloroform (Rowe Scienti c) were used as received.1-Butanesulfonic acid [21] and dibutyl disul de [22] were prepared by literature procedures.Tetrachloroauric acid [23], and butanethiol-capped AuNPs (BT@AuNPs) [2] were prepared using literature procedures.The AuNPs were characterised by scanning electron microscopy (SEM), transmission electron microscopes (TEM), proton nuclear magnetic resonance spectroscopy ( 1 H NMR). 1 H NMR spectra were recorded using a Bruker NMR spectrometer operating at 400 MHz.Spectra were referenced using residual non-deuterated signals: Transmission electron microscopy (TEM) images were taken using a JEOL JEM-F200 FE-TEM operating at 200 kV and tted with a Gatan Rio 1816-4k x 4k camera.The TEM samples were prepared by evaporating diluted nanoparticle solution on the carbon-coated copper grid.The images were analysed using ImageJ software (https://imagej.nih.gov/ij/).SEM was performed at facilities at Western Sydney University.A Zeiss Merlin eld emission gun scanning electron microscope (FEGSEM) was utilised for imaging samples prepared on stubs.The FEGSEM was operated at 20kV accelerating voltage in Hivac mode at a working distance of approximately 3 mm.Both secondary and in-lens secondary detectors were utilised for imaging.High-resolution mass-spectrometry (HRMS) was performed using an Agilent 6510 Q-TOF using ow injection and in positive ion mode for [M + H] + , or negative ion mode [M-H] − where speci ed.

Resistance measurements of AuNP lms.
Suspensions of BT@AuNPs in chloroform (10 mg/ml) were drop cast onto DropSens (Metrohm) interdigitated gold electrodes forming lms of AuNPs, which were then heated within a modi ed Linkam THMS600 temperature control stage.A Linkam TMS 94 controller maintained a heating rate of 10°C min − 1 from room temperature to 250°C.A Rigol DM3058E digital multimeter and a PT100 (RS PRO) RTD sensor, 2mm x 5mm Class B thermocouple measured the temperature on the gold electrode.A Rigol DM3068 digital multimeter (maximum resistance of 100 MΩ) measured the Electrical resistance of the electrode.A LabView program was used to interface with and control the multimeters, and to acquire the temperature and electrical resistance.Ozone was generated using a Hailea HLO-300 Ozonizer at 300 mg/h in a ow of 3.5 L/min.Warning.Ozone was destructive to several electronic components including thermocouples when exposed for extended periods (up to 20 hours).

Analysis of ozone treated AuNPs.
BT@AuNPs were placed in a 5 mL side arm tube attached to a condenser cooled to -0.5°C and tted with a drying tube.The out ow from an ozone generator was passed through dry silica gel beads tightly packed in a condenser cooled to -0.5°C to remove moisture and then directed into the side arm tube for 48 hrs.Organic residues were then collected by rinsing the interior of the condenser and reaction tube rst with CDCl 3 (with sonication) and then DMSO-d 6 .The solutions were ltered through cellulose bre (Kimwipe) to remove elemental gold and analysed using 1 H NMR spectroscopy and HRMS.
Films of AuNPs were prepared by drop-casting a suspension of BT@AuNPs in chloroform onto interdigitated gold electrodes.Upon heating at 10°C/min in air, the lms sintered at ~ 190°C to form a conductive gold lm.The sintering event is associated with a change in resistance from > 1 MΩ to < 100 Ω (Fig. 1).These results are consistent with our earlier studies on BT@AuNPs [24].In contrast, heating of the lms in an ozone-rich atmosphere caused the lms to sinter at ~ 80°C, which is signi cantly lower than the T SE of the AuNPs sintered in air.
The signi cant decrease in the T SE observed upon heating at 10°C/min prompted experiments to examine the effect of exposure to ozone at room temperature.Films of metallic gold were formed from AuNP lms exposed to a stream of ozone (Fig. 2).To probe this behaviour further, lms of AuNPs were formed by dropcasting AuNP suspensions onto interdigitated electrodes and the resistance measured upon exposure to ozone.Figure 3 shows the resistance of BT@AuNPs at room temperature upon exposure to air and an ozone rich environment over 15 h.Under an atmosphere of air, the resistance remained stable at 2.5 MΩ.In contrast, under an ozone atmosphere the resistance of the BT@AuNPs decreased markedly to ~ 300 Ω after ~ 2.5 h.The slightly greater resistance of the room temperature sintered AuNPs (300 Ω) compared to the thermally sintered AuNPs (~ 15 Ω) may be attributed to residual organic material remaining after the decomposition of butanethiol (see below), leading to a less dense lm [25].
Low temperature sintering has been reported in our earlier work using a chemically synthesised oxidant but the nature of the reactions leading to the sintering event was not explored [5].Here we examine the residue surrounding the gold lm formed after exposure to the ozone atmosphere using 1 H NMR spectroscopy and HRMS data.After reaction with ozone, which induced sintering, the reaction vessel was rinsed with CDCl 3 and then DMSO-d 6 .The CDCl 3 fraction contained very little material of which none could be characterised by 1 H NMR spectroscopy.The DMSO-d 6 fraction contained a signi cant amount of organic material.
The 1 H NMR spectrum of the DMSO-d 6 fraction contained signals that are consistent with the spectrum of 1-butanesulfonic acid (Fig. S3).A triplet at 0.85 ppm is assigned the CH 3 group, a sextet and a quintet at 1.31 and 1.54 ppm, respectively, are assigned to the two central CH 2 groups.A triplet at 2.43 ppm is assigned to the CH 2 group adjacent to the SO 3 H group and is identical to a spectrum of synthesised 1butanesulfonic acid.Importantly, we found no evidence of dibutyl disul de formation upon ozoneinduced sintering in either the 1 H NMR spectrum (Fig. S4) or HRMS.Upon sintering in air, hydrogen, nitrogen or argon atmospheres, surface-bound thiyl ligands leave exclusively as their corresponding disul de compounds [24].Other possible sulfur-containing compounds such as 1-butanethiol were excluded by comparison of the 1 H NMR spectra (Fig. S4).
The CDCl 3 and DMSO-d 6 fractions were both examined by mass spectrometry in positive and negative ion modes.In negative ion mode, the major peak was observed at m/z 137.028, which corresponds to the sulfonate ion with formula CH 3 (CH 2 ) 3 SO 3 − .Other oxygen-and sulphur-containing compounds such as sulfoxide and sul nates (105 and 121 m/z respectively) were not detected.These ndings are consistent with work examining UV-induced photooxidation of thiol SAMs on gold (over various time periods) using XPS where only the corresponding sulfonate ions were detected [26].
Considering the resistance data together with post-sintering analysis, it is apparent that when the BT@AuNPs are exposed to an ozone-rich environment, the butanethiol ligands undergo oxidation to butanesulfonic acid thus facilitating desorption of the ligand from the gold surface and inducing the sintering event (Fig. 4.) SEM images were collected of thermally-induced and ozone-induced sintered BT@AuNPs (Fig. 5) as well as pristine BT@AuNPs (Fig. S2).The SEM images of the thermally-induced, sintered gold lms are consistent with previous reports, showing densi cation and large grain barriers [24].SEM images of the room temperature ozone-induced sintered gold lms show ner grain size with agglomeration of particles and some residual material.Grains ranging from 600-1000 nm are apparent in the thermallyinduced sintered structures while the ozone-induced lms have smaller grains ≤ 200 nm.

Conclusion
Conductive gold lms have been prepared by sintering BT@AuNPs under an ozone-rich atmosphere.
Resistance measurements of BT@AuNPs showed that exposure to the ozone-containing atmosphere during heating signi cantly reduced the T SE compared to sintering under ambient conditions by ~ 80°C.
Furthermore, electrically conductive gold lms were formed at room temperature when BT@AuNPs were exposed to ozone for ~ 2.5 hours.
Examination of the AuNPs post-sintering revealed that the butanethiyl ligands undergo oxidation to form the corresponding butanesulfonic acid, which is a poor stabilizing ligand.We found no evidence for dibutyl disul de (the major product of sintering under ambient conditions) in the post-sintering residue, indicating that the oxidation process is further promoted by ozone.The gold lms prepared by the new room temperature ozone-induced sintering process showed a different morphology to those sintered by thermal activation under ambient conditions (observed by SEM) with the former producing ner grain sizes.

Declarations Figures
Resistance Resistance data of BT@AuNPs at room temperature in air and an ozone rich environment over 15h Schematic depicting the removal of butanethiyl ligands from the surface of gold nanoparticles in an ozone rich environment Figure 5 SEM images of gold lms obtained by (left) heating BT@AuNPs in to 250 °C at 10 °C/min, and (right) exposing BT@AuNPs to an ozone-rich atmosphere for 6 h at room temperature

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download. Supplementaryv4.pdf