Journal of Colloid and Interface Science 439 (2015) 21–27
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Vapor-phase preparation of gold nanocrystals by chloroauric acid pyrolysis Yiqin Chen a, Xuezeng Tian b, Wei Zeng a, Xupeng Zhu a, Hailong Hu c, Huigao Duan a,⇑ a
College of Physics and Microelectronics, State Key Laboratory for Chemo/Biosensing and Chemometrics, Hunan University, Hunan 410082, China Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China c School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link, 637371 Singapore, Singapore b
a r t i c l e
i n f o
Article history: Received 2 August 2014 Accepted 12 October 2014 Available online 22 October 2014 Keywords: Gold nanocrystals Vapor-phase deposition Chloroauric acid Pyrolysis Plasmon resonance
a b s t r a c t We report that gold nanocrystals can be prepared from vapor phase using chloroauric acid (HAuCl4) as the precursor. By tuning the vapor-phase deposition parameters, the size and space distribution of the gold nanocrystals can be well controlled on substrates. Systematic control experiments demonstrate that intermediate AuCl and AuCl3 products pyrolyzed from HAuCl4 play an essential role in this vapor-phase deposition process. Compared to conventional wet-chemical synthesis process, vapor-phase process enables direct deposition of gold nanoparticles on solid substrates with better coverage and uniformity, which may find applications in surface-enhanced Raman scattering and plasmon-enhanced photocatalysis. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Gold nanostructures have tremendous potential applications due to their unique electronic, optical and catalytic properties [1–6]. Most of their applications are strongly dependent on the size, shape and surface properties of gold nanostructures. Therefore, seeking for controllable methods to prepare gold nanostructures has been an active research area for the past decades. Generally, there are two main well-developed approaches in the preparation of gold nanostructures, i.e., top–down fabrication and bottom-up wet-chemical synthesis. Top–down technique provides freedom of design for planar patterns with high resolution and position accuracy for solid-state-device applications, but it usually involves lithographic process requiring costive vacuum facilities for large-scale fabrication [7,8]. Meanwhile, the fabricated nanostructures are polycrystalline with high surface roughness. Both grain boundaries and surface roughness result in the damping of hot electrons, which may degrade the electronic, optic and catalytic performances of gold nanostructures for applications [1,9,10]. On the contrary, wet-chemical synthesis could provide low-cost and large-scale preparation of monodispersed single-crystalline gold nanostructures with varied morphologies and surface chemistry for solution-based applications, but it requires an additional complicated assembling process for the ⇑ Corresponding author. E-mail address: [email protected] (H. Duan). http://dx.doi.org/10.1016/j.jcis.2014.10.017 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
specific applications on solid surface. As the nanoparticles tend to aggregate during the drying self-assembly process, it is difficult to obtain highly monodispersed nanoparticles with controlled space distribution on the solid surfaces, which limits the applications of wet-chemical synthesis in solid-state devices. Meanwhile, the surfactants, usually involved in the wet-chemical synthesis process, are difficult to completely remove from the particle surface, which have negative effects on the performance of gold nanoparticles in plasmonic and catalytic applications [11,12]. Therefore, developing alternative methods which have potential to overcome the limits of the existing top-down or wet-chemical techniques for preparing gold nanostructures is still interesting and required. Vapor-phase synthesis is a versatile, flexible and scalable technique to prepare nanostructures with high crystallinity and clean surface on various substrates [13–16]. Researchers have made many attempts to use vapor-phase method to prepare gold nanostructures. For example, Kim et al. demonstrated the preparation of high-quality single-crystalline gold nanowires via the vaportransport method using bulk gold pellets as the precursor [16,17]. However, this process requires temperature higher than 1000 °C to vaporize the bulk gold precursor. To lower the synthesis temperature, Igumenov et al. attempted to use metalorganic precursors to prepare gold nanoparticles [18]. In this case, to avoid the deposition of carbon substance, the deposition temperature should be as low as possible and thus it is difficult to obtain gold nanoparticles with high crystallinity.
22
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
In this work, we report a facile vapor-phase deposition process to prepare high-quality gold nanostructures with well-defined shapes using chloroauric acid (HAuCl4) as the precursor. It has been known that HAuCl4 is the most widely used reactant in wet-chemical synthesis of gold nanoparticles and can also be in situ decomposed into gold via pyrolysis or solution-based aerosol deposition process [19,20]. In our work, we found that HAuCl4 can be readily decomposed into evaporable products which can further transport to other solid substrates to form gold nanocrystals at relatively low temperature. We demonstrate that the size and space distribution of the gold nanocrystals can be tuned by controlling the deposition parameters in a tube-furnace system. With a systematic study and in-situ experiments, a possible mechanism is also proposed to explain this vapor-phase deposition process. With this unique technique, single-crystalline gold nanocrystals without any surface contaminations can be prepared in large scale on solid substrates, which have potential for plasmonic and catalytic applications. 2. Experimental section
(Thermo Scientific, ESCALAB-250Xi) was used to study the chemical valence of the materials. Specimens were analyzed in high vacuum environment (107 mbar, 25 °C) utilizing Al Ka at 20 kV with a power of 200 W. 2.2.3. In-situ TEM The sample for in-situ TEM experiment was prepared with HAuCl4 source temperature at 500 °C but TEM grid was put at room temperature in 9 Torr pressure. In-situ TEM experiment was carried out using a JEOL JSM-2010F Transmission Electron Microscope (200 kV accelerating voltage) which is integrated with a digital heating system. 2.2.4. The sample preparation procedures for Raman measurements We compared the Raman scattering signals between a 50 nm thick gold film and the sample prepared by our vapor-phase deposition method. We immersed two samples into 105 M Rhodamine-6G (R6G) aqueous solution for 12 h, then rinsed two substrates with ethanol completely and spontaneously dried them at room temperature. The rinsing process ensured the formation of only a monolayer of R6G over the surface of Au.
2.1. Vapor-phase deposition process To quantify the amount of gold element in the experiment each time, one gram HAuCl44H20 (purity > 99.999%, Sinopharm Chemical Reagent Co., Ltd) was dissolved in 40 ml ethanol absolute, and gold element concentration was 74 mM. For each deposition, 80 ll chloroauric acid solution was loaded it into quartz boat. Ethanol solution was feasible to evaporate as vapor phase under the condition of low pressure and mild heating before the decomposition of chloroauric acid. Note that HAuCl4 solution enables accurate loading of the HAuCl4 source, which is essential to the vapor-phase deposition process. The tube-furnace system was purchased from Hefei Kejing Corp. HAuCl4 source together with SiO2 (300 nm thickness on Si) substrates were put into 60-mm-diameter quartz tube equipped with a rotary pump and carrying gas (N2). HAuCl4 precursor was loaded in the upstream and the substrates were put in the downstream. The control of in-tube work pressure was realized by the dynamic equilibrium between injecting N2 and pump rate. Heating procedure is programmed by micro-computer integrated in the furnace system. For example, the tube furnace was first heated from room temperature to a certain high temperature (e.g. 600 °C) with a constant heating rate (e.g. 40 °C/min), then kept at the high temperature for a certain time, and finally cooled down to room temperature naturally by turning off the heating units. 2.2. Characterization 2.2.1. SEM and HRTEM The morphology of gold nanoparticles was characterized using Scanning Electron Microscopy (Hitachi S-4800 & FEI NanoLab Helios 600i) with 10 kV accelerating voltage and work distance of 8 mm. The as-prepared gold nanoparticles were on SiO2 substrates. To prepare TEM sample, we scratched the gold nanoparticles off the substrates to ethanol solution using the edge of a fresh silicon wafer. With ultrasonic bathing, we made the gold nanoparticles dispersed and then dripped the solution onto a copper TEM grid coated with an ultrathin carbon film (10 nm). The high resolution TEM metrology was done by a FEI Titan G2 machine with an accelerating voltage of 300 kV. 2.2.2. EDS and XPS A HORIBA energy dispersive X-ray spectroscopy attached to HITACHI scanning electron microscopy operated at 20 kV was characterized to target sample. X-ray photoelectron spectroscopy
2.2.5. Raman scattering measurements Raman scattering spectra of two selected samples were measured on a HORIBA JOBIN YVON HR800 UV Raman Spectroscopy imaging system. The excitation source was 532 nm. Raman signals were collected using a 50x objective (numerical aperture, N.A. = 0.5) under laser power intensity of 0.5 mW and with integration time of 60 s. 2.2.6. Dark-field imaging and scattering measurements Dark Field imaging and the scattering spectroscopy of single gold nanocrystal were measured on a WITec Confocal Raman Imaging Microscope System alpha 300R. A halogen lamp (15V, 3150 K) with 100 W power intensity was used as white light source and illuminated to the sample via an dark-field objective lens (N.A. = 0.75, 100). All of the samples were measured with the integration time of 20 s. The calculation of dark-field scattering spectrum was referred to the formula as follow:
Iscat ðkÞ ¼
Iparticle Isub Ihalogen Idark
where Iscat ðkÞ presents the real scattering spectra of single nanoparticle; the spectrum of Iparticle, Isub, was separately collected from then original particle and surrounding substrate. Ihalogen, Idark is the spectrum of halogen light source and background count of measurement system, respectively. 3. Results and discussion 3.1. Preparation and morphology characterizations Fig. 1a shows the schematic of the experimental setup for the vapor-phase deposition process. Fig. 1b shows a scanning electron microscopy (SEM) image of the resultant monodispersed gold nanocrystals with well-defined shapes on a SiO2 (300 nm thick on Si substrate) surface. We observed that most of the obtained nanocrystals are octahedral, as shown by the enlarged SEM images in Fig. 1b. The trend to form octahedra is because the gold nanoparticles tend to expose their {1 1 1} crystal facets due to the energy minimization during the growth process [21]. To further investigate the morphologic uniformity of the synthesized gold nanocrystals, we studied their plasmon resonance properties using singleparticle dark-field scattering measurements [22,23]. Fig. 1c shows the corresponding dark-field optical micrograph of the nanocrys-
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
23
Fig. 1. (a) Schematic of the experimental system. (b) SEM image of gold nanocrystals (170 nm size). (c) and (d) The corresponding optical dark-field image and dark-field scattering spectra of the labelled six gold nanoparticles. Experimental parameters: In 660 mTorr pressure, substrate and source were heated from 25 °C to 600 °C during 15 min with a distance of 6 cm, and kept at 600 °C for 2 h.
tals in Fig. 1b. The high-contrast colorful dark-field scattering image indicates that the plasmon resonance of these observed gold nanocrystals locates in the visible region. Further dark-field scattering spectra of the selected octahedra are shown in Fig. 1d. All of the measured particles show two plasmon resonance peaks locating at 570 nm and 675 nm, which well match the green and orange colors in the dark-field image and also indicate that the synthesized nanocrystals have consistent morphologies. The different scattering intensities are supposed to result from the different postures of the octahedra lying on the substrate. To prove the synthesized products are metallic gold nanocrystals with high crystallinity, we investigated the obtained nanoparticles using high-resolution transmission electron microscopy (HRTEM). Fig. 2a shows the SEM image of the octahedral nanoparticles used for the HRTEM characterization. The average edge length of the octahedron was 100 nm for this sample, and the yield of octahedral particles is over 75% (the statistics of more than 100 particles). Other shapes include decahedral, diamond-like and undefined geometry. We transferred these nanocrystals onto TEM grid with an ultrathin supporting carbon film of 10 nm for the HRTEM study. Fig. 2b presents the TEM image of a typical octahedron, and the presumable posture of this particle on the carbon film is given by the inset schematic. From the HRTEM image of a corner of this particle, as shown in Fig. 2c, we can clearly see the lattice fringes with a interplanar spacing of 2.35 angstrom ascribed to {1 1 1} plane of face-center cubic (fcc) gold. Further selected-area electron diffraction (SAED) pattern of this particle, shown in Fig. 2d, suggests the high crystallinity of the nanocrystal. The above TEM results undoubtedly confirm that the vapor-phase prepared nanocrystals from HAuCl4 are metallic gold nanoparticles. More HRTEM results of gold nanocrystals with various sizes and shapes prepared at different parameters are provided in the supplementary information (Figure S1).
The size and the space distribution of the nanocrystals can be controlled by tuning the deposition parameters, such as growth temperature, source-to-substrate distance, flow rate of the carrying gas, which are similar with chemical vapor deposition process for other nanostructures [24–26]. As a proof of concept, we demonstrated that the size of gold nanocrystals can be finely tuned with different substrate-to-source distance, as shown in Fig. 3. When the substrates were put farther from the HAuCl4 source in the downstream, the obtained gold nanocrystals became smaller. The average diameter of the nanocrystals decreased from 67 nm to 10 nm when the substrate-to-source distance increased from 3 cm to 6 cm, as indicated by the statistical diagrams. The observed result can be attributed to the different precursor concentration at the different substrate positions [27,28]. It should be noted that though the obtained nanocrystals were highly monodispersed with a uniform size distribution for each substrate, we did not observe clear trend of the particle density when varying the substrate-tosource distance, which might be caused by the instability of the gas flow in the tube furnace during the deposition process. Meanwhile, for the influence of different deposition temperature, we found that the resultant gold nanoparticles became smaller when decreasing the deposition temperature. The details can be seen in Figure S2 in the supplementary information. 3.2. Mechanism study As a common precursor, HAuCl4 is widely used for the synthesis of gold nanoparticles, but the fact that it undergoes the decomposition into evaporable products at low temperature (as low as 100 °C according to our experimental results shown in Figure S3 in the supplementary information) has never been clearly studied. A further look at the mechanism of this vapor-phase deposition process is not only of significance for the greater control of the
24
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
(a)
(b)
20 nm
200 nm
(d)
(c)
(220)
(002)
(020) 2 nm
(111)
Fig. 2. (a) SEM image of gold nanocrystals (100 nm) which were used for the TEM characterization. Inset shows an enlarged SEM image of a single gold octahedral nanocrystal. (b) TEM image of a typical gold nanocrystal. The inset gives the schematic side view of the gold octahedron. (c) High-resolution TEM micrograph at the corner of gold nanocrystal (marked in Fig. 2(b). (d) The selected-area electron diffraction pattern of the nanocrystal. Experimental parameters: In 600 mTorr pressure, substrate and source were heated from 25 °C to 500 °C during 12.5 min with a distance of 6 cm, and kept at 500 °C for 2 h.
60 50 40 30 20
σ=3.7 nm
70 60 50 40 30 20 10
10 0
80
0 10 20 30 40 50 60 70 80 90 100
Diameter (nm)
0
100
90
Avg=34.7 nm
80 70
Avg=22.5 nm
90
Avg=10.9 nm
σ=1.1 nm
80
σ=1.1 nm
60 50 40 30 20 10
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
0
Frequency (%)
70
(d)
100
90
Frequency (%)
Frequency (%)
80
100
Avg=67.3 nm σ=9.3 nm
Frequency (%)
100 90
(c)
(b)
(a)
70 60 50 40 30 20 10
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
0
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
Fig. 3. SEM images of gold nanocrystals prepared with different source-to-substrate distances. (a) 3 cm, (b) 4 cm, (c) 5 cm, and (d)p6ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cm. Statistic histograms of the size of gold ffi nanocrystals on the SiO2 substrates are shown in the bottom panel. Over 100 nanoparticles were counted for statistics (d ¼ 2 0:385A). A is the area of the nanoparticles, which is obtained by ImageJ software [29]. All of the samples were prepared at 500 °C and 600 mTorr. All scale bars are 200 nm.
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
preparation of gold nanocrystals using this method, but also provides insights to understand other HAuCl4-related reaction process. To study the mechanism, we did both ex-situ SEM and insitu TEM experiments to understand the mass-transport and nucleation-growth process in this vapor-phase deposition process. The first thing we tried to do was to determine the intermediate vapor products during the deposition process. Metallic gold can be first ruled out as an intermediate vapor product due to its melting temperature is higher than 1000 °C while our deposition temperature can be as low as 100 °C in supplementary information (see Figure S3). It is known that HAuCl4 can be easily decomposed into AuCl3 and HCl under thermal treatment with a temperature as low as 110 °C [19]. To explain the vapor-phase process, we further assume that the decomposed AuCl3 is able to vaporize and transport to distant regions, so the reactive equation can be written as: D
HAuCl4 ðsÞ ! AuCl3 ðgÞ þ HClðgÞ:
ð1Þ
To verify our assumption, we deposited some product onto a SiO2 substrate at room temperature by putting the substrate at the outside of the heating zone. The source temperature was set to be 500 °C. A typical SEM image of the product was provided in Figure S4a in supplementary information, which shows the film-like morphology on the substrate. With the help of energy dispersive X-ray spectroscopy (EDX or EDS), we found both gold and chlorine elements exist in the film (see Figure S4b in supplementary information). We further studied the chemical structure of as-deposited film by X-ray photoelectron spectroscopy (XPS). The peak of 91.1 eV near in Au (III) 4f5/2 and 88.6 eV near in Au (I) 4f5/2 was clearly shown in Figure S4c in supplementary information, respectively [30,31]. Au (I) existed in the as-deposited film owing to the source temperature of 500 °C. Under high-temperature environment, vaporized AuCl3 as stated in Eq. (1) simultaneously might occur the reaction as follow: D
AuCl3 ðgÞ ! AuClðgÞ þ Cl2 ðgÞ:
ð2Þ
Hence, the existence of Au (I) in film was originated from further decomposition of vaporized AuCl3. Vapor-phase AuCl3 and AuCl were together transferred with carried gas flow and deposited onto distant substrate at room temperature. Therefore, as-deposited film was a mixture of AuCl3 and AuCl. With the above assumption, we further investigated the nucleation growth process of the gold
25
nanocrystals. A possible growth mechanism is shown in Fig. 4a and can be briefly described as follows. Firstly, a homogeneous AuCl3/AuCl film is rapidly deposited onto the substrate during the pyrolysis process at a low pressure. Under continuous thermal treatment, the AuCl3/AuCl decompose into Au to form ultrasmall monodispersed gold seeds embedded in AuCl3/AuCl film. These gold seeds further grow to be gold nanoparticles due to the continuous decomposition of the remaining AuCl3/AuCl. During the above processes, small gold nanoparticles may merge each other due to Ostwald ripening [32] or Rayleigh instability [33]. To further confirm the above hypothesis on growth mechanism, we carried out a series of thermal treatments on the AuCl3/AuCl precursor film obtained on a substrate put at the room temperature region. The results are shown by the SEM images in Fig. 4b– d. Fig. 4b shows the as-deposited film, and no obvious gold nanoparticles can be seen. After a thermal treatment at 100 °C for 30 min (Fig. 4c), the continuous AuCl3/AuCl film transferred to isolated droplets in which gold seeds were wrapped by surrounding AuCl3/AuCl product. Note that the gold nanoparticles are much brighter in the SEM images due to their higher secondary electron yields compared to AuCl3/AuCl. Further increasing the reaction time to 60 min resulted in larger gold nanocrystal seeds and less remaining AuCl3/AuCl, as shown in Fig. 4d. By elevating the thermal-treatment temperature to 400 °C, shape-defined gold nanocrystals with larger size and clean surface formed, as shown in Fig. 4e. These series of results clearly demonstrate the scenarios of deposition of AuCl3/AuCl precursor film, formation of tiny gold seeds, and growth of the gold nanocrystals from the AuCl3/AuCl precursor, which provide sufficient and convincible evidences to verify the crystallization and growth mechanism as indicated by the schematics. From above results, we also surprisingly see that AuCl3/AuCl product can be slowly reduced to Au (0) at a temperature as low as 100 °C. It is known that disproportionation reaction could occur in AuCl and transfer into AuCl3 and Au (0) at relatively low temperature [34]. The possible reaction equation is given below: D
3AuClðsÞ ! AuCl3 ðsÞ þ 2AuðsÞ:
ð3Þ
Though the existing literature only report a lowest reaction temperature of 260 °C [19], lower reaction temperature could be possible at the nanometer scale [35]. Once tiny gold nanoparticles were
Fig. 4. (a) A series of intuitive schematics showing different stages of the gold nanocrystals in the vapor-phase deposition process. (b)–(e) Ex-situ SEM images of experimental results to show the different stages of the gold nanocrystals under a mild thermal treatment. All SEM images were done on the same sample. The AuCl3/AuCl film was deposited at 22 Torr pressure with source temperature of 500 °C and substrate temperature of 25 °C (i.e. room temperature). The scale bars are 200 nm.
26
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
(b)
(a)
(c)
(d)
Thin film Thick film Thick film Thin film
100 nm
(e)
25e C (222) (220) (200) (111)
50 nm
(f)
200e C
(220)
50 nm (220)
(g)
(111)
400e C
50 nm
C 600e
(h) (220) (111)
(111)
Fig. 5. In-situ TEM observation of the dynamic evolution of gold nanoparticles from the AuCl3/AuCl precursor film. (a)–(d) TEM micrographs show the morphologies of the formed gold nanoparticles after a series of thermal annealing under different temperatures. (e)–(h) The electron diffraction patterns of the samples corresponding to TEM images in (a)–(d), respectively.
30.0k
(a) Intensity (Counts)
25.0k
(b) 614 1362
20.0k 776 15.0k
1185 1311
1509 1650 1570
10.0k 5.0k
100 nm
0.0
600
800
1000
1200
1400
1600
Raman Shift (cm-1) Fig. 6. (a) A SEM image of gold nanocrystals (100 nm) on SiO2 substrate with a high particle density, and amount of sub-10-nm gaps (highlighted by red arrows) can be formed. (b) The Raman spectra of R6G molecules on high-density gold nanocrystals (blue curve) and a gold film (red curve). Both the source and substrate temperature for this sample were 500 °C. All of the other experimental parameters were the same as described in Fig. 3. The sample was away from the source with a distance of 3 cm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
formed in the AuCl3/AuCl film, they may serve as catalytic sites for particle growth via a self-catalysis growth model [36–38], resulting in the formation of larger nanocrystals at a relatively low temperature. To further demonstrate the formation of gold nanocrystals from the vapor-deposited AuCl3/AuCl film under thermal treatment, insitu TEM study was done. To do this, we first deposited an AuCl3/ AuCl film on the TEM grid with an ultrathin continuous carbon film of 10 nm, and the experimental condition for AuCl3/AuCl deposition was the same as that in Fig. 4. In order to minimize the electron-beam influence, the electron beam was highly dispersed for imaging and was shut off during heating. A representative TEM image of the as-deposited AuCl3/AuCl film is shown in Fig. 5a, from which we can see that the deposited film is almost homogeneous on the substrate and no obvious gold nanoparticles can be observed, similar with the result obtained from SEM image Fig. 4b. The corresponding electron diffraction pattern of the asdeposited film is given by Fig. 5e. Interestingly, from the diffraction pattern, we can see the diffraction rings of the gold, indicating the
existent of ultrasmall sub-nanometer gold nanocrystals in the deposited film. Fig. 5b–d show the evolution of the gold nanoparticles obtained at different annealing temperature and time. Even with a low temperature at 200 °C, amount of AuCl3/AuCl precursor transferred into gold nanoparticles with an average diameter less than 5 nm, as shown in Fig. 5b. It is worth noting that relatively larger gold nanoparticles formed at the original darker area (marked by ‘‘thick film’’ in Fig. 5a) than those in brighter area (marked by ‘‘thin film’’) due to the different thickness of initial AuCl3/AuCl film, indicating that the thickness of AuCl3/AuCl film was a key role to the size of gold nanoparticles. While elevating the annealing temperature, the size of the gold nanoparticles gradually increased and the particle density decreased, as shown in Fig. 5c and d. Further electron diffraction results, shown in Fig. 5f–h, prove the gradual crystallization of the gold nanoparticles. It should be noted that due to the different wetting properties between gold and carbon film, the final formed gold nanoparticles did not have well-defined shapes, which is different from what we observed on the SiO2 substrates. To investigate how the wetting properties of the substrates
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
affect the synthesis results, we further did the control experiments on carbon fiber and silicon substrate (see Figure S5 in supplementary information). All of the resultant gold nanoparticles on the two different substrates do not show well-defined shape. Therefore, in the in-situ TEM study, the unwell-defined shape of synthesized gold nanoparticles was attributed to the different wetting property of carbon film compared to SiO2 substrate. 3.3. SERS application A promising application area of gold nanostructures is to utilize their plasmonic properties for sensing applications. One of the advantages of this vapor-phase method is its potential to obtain gold nanocrystals in large scale with clean surface and tunable particle density, which may be applied in various applications such as surface-enhanced Raman spectroscopy (SERS) for molecule-scale detections. Fig. 6a shows a SEM image of high-density gold nanocrystals in a large area and the particle density is 116 counts/ lm2. The sample was prepared with an appropriate source-to-substrate distance. However, due to the high-density distribution of gold nanoparticles, adjacent particles might coalesce together, resulting in the formation of irregular gold nanoparticles. When using the samples as SERS substrates to detect R6G molecules with an excitation wavelength at 532 nm, the Raman scattering intensities were significantly enhanced with a factor of 103 compared to common gold film (50 nm thick) substrate. Such a significant enhancement factor can be attributed to the local surface plasmon resonance of these monodispersed high-quality gold nanocrystals. Furthermore, due to the high density of the nanocrystals, amount of small gaps less than 10 nm can be formed between the nanocrystals. These tiny gaps could significantly enhance the electric field of the incident light [39,40] to form so-called ‘‘hot spots’’ for Raman scattering. Further improvement of the enhancement factor may be achieved in the future by optimizing the particle density to obtain more sub-10-nm gaps. 4. Conclusion In summary, we reported a vapor-phase method to prepare gold nanocrystals by pyrolyzing a common HAuCl4 precursor at relatively low temperature without using any reducing agents and catalysts. With this method, we demonstrated that large-scale singlecrystalline gold nanocrystals with well-defined shapes can be synthesized on solid substrates. The systemic study of mechanism make us understood the process of crystallization. It is helpful to tune shape and size of crystals by feasible experiment optimum. Meanwhile, we can analogize that this vapor-phase method could also be used to prepare gold nanorods and nanowires by introducing appropriate assistant reactants, and similar process may be applied to HPtCl4 and HPdCl4 to prepare Pt and Pd nanostructures. This method provides a good candidate to obtain super-clean sample for improving the accuracy of molecular probing or to study reactive dynamics on metal nanocrystals surface in catalysis. Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant Nos. 11274107, 61204109), Program for New Century Excellent Talents in University (Grant No. NCET-130185), Foundation for the Author of National Excellent Doctoral Dissertation of China (Grant No. 201318), Interdisciplinary Program of Hunan University, and Open Project of State Key Laboratory of Solid State Microstructures in Nanjing University.
27
We are grateful to Yufang Ma for Raman scattering measurement and to Zhengmei Yang for helpful discussion. We also thank Wei Chen in National University of Defense Technology for SEM measurements. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.10.017. References [1] Y. Zhao, J. Chu, S.H. Li, W.W. Li, G. Liu, Y.C. Tian, H.Q. Yu, Electroanalysis 26 (2014) 656–663. [2] H. Hu, H. Duan, J.K.W. Yang, Z.X. Shen, ACS Nano 6 (2012) 10147–10155. [3] A. Gopalakrishnan, M. Chirumamilla, F. De Angelis, A. Toma, R.P. Zaccaria, R. Krahn, ACS Nano 8 (2014) 7986–7994. [4] W.L. Chen, F.C. Lin, Y.Y. Lee, F.C. Li, Y.M. Chang, J.S. Huang, ACS Nano 8 (2014) 9053–9062. [5] A. Corma, H. Garcia, Chem. Soc. Rev. 37 (2008) 2096–2126. [6] M. Valden, X. Lai, D.W. Goodman, Science 281 (1998) 1647–1650. [7] H. Duan, H. Hu, K. Karthik, Z.X. Shen, J.K.W. Yang, ACS Nano 5 (2011) 7593– 7600. [8] Y. Wang, M. Abb, S.A. Boden, J. Aizpurua, C.H. Groot, O.L. Muskens, Nano Lett. 13 (2013) 5647–5653. [9] P. Nagpal, N.C. Lindquist, S.H. Oh, D.J. Norris, Science 325 (2009) 594–597. [10] K.D. Schierbaum, U. Weimar, W. Göpel, Sens. Actuators B 3 (1991) 205–214. [11] W.E.A. Moerner, J. Phys. Chem. B 106 (2002) 910–927. [12] G. Chen, Y. Tan, B. Wu, G. Fu, N. Zheng, Chem. Commun. 48 (2012) 2758–2760. [13] P. Mohanty, I. Yoon, T. Kang, K. Seo, K.S.K. Varadwaj, W. Choi, Q.H. Park, J.P. Ahn, Y.D. Suh, H. Ihee, B.S. Kim, J. Am. Chem. Soc. 129 (2007) 9576–9577. [14] Y. Yoo, I. Yoon, H. Lee, J. Ahn, J.P. Ahn, B. Kim, ACS Nano 4 (2010) 2919–2927. [15] H. Choi, S.H. Park, J. Am. Chem. Soc. 126 (2004) 6248–6249. [16] Y. Yoo, K. Seo, S. Han, K.S.K. Varadwaj, H.Y. Kim, J.H. Ryu, H.M. Lee, J.P. Ahn, H. Ihee, B. Kim, Nano Lett. 10 (2010) 432–438. [17] I. Yoon, T. Kang, W. Choi, J. Kim, Y. Yoo, S.W. Joo, Q.H. Park, H. Ihee, B. Kim, J. Am. Chem. Soc. 131 (2009) 758–762. [18] R.G. Parkhomenko, N.B. Morozova, G.I. Zharkova, Y.V. Shubin, S.V. Trubin, V.V. Kriventsov, B.M. Kuchumov, T.P. Koretskaya, Chem. Vap. Deposition 18 (2012) 336–342. [19] R.G. Palgrave, I.P. Parkin, Chem. Mater. 19 (2007) 4639–4647. [20] H. Shi, L. Zhang, W. Cai, Mater. Res. Bull. 35 (2000) 1689–1695. [21] M. Grzelczak, J. Pérez-Juste, P. Mulvaney, Chem. Soc. Rev. 37 (2008) 1783– 1791. [22] J.J. Mock, M. Barbic, D.R. Smith, D.A. Schultz, S. Schultz, J. Chem. Phys. 116 (2002) 6755–6759. [23] K.H. Su, Q.H. Wei, X. Zhang, Nano Lett. 3 (2003) 1087–1090. [24] X. Li, W.C. Magnuson, A. Venugopal, J. An, J.W. Suk, B. Han, M. Borysiak, W. Cai, A. Velamakann, Y. Zhu, L. Fu, M.E. Vogel, E. Voelkl, L. Colombo, S.R. Ruoff, Nano Lett. 10 (2010) 4328–4334. [25] Z. Li, P. Wu, C. Wang, X. Fan, W. Zhang, X. Zhai, C. Zeng, Z. Li, J. Yang, J. Hou, ACS Nano 5 (2011) 3385–3390. [26] H. Peng, Z. Pan, L. Xu, X. Fan, N. Wang, C.S. Lee, S.T. Lee, Adv. Mater. 13 (2001) 317–320. [27] J. Xu, X. Zhuang, P. Guo, Q. Zhang, W. Huang, Q. Wan, W. Hu, X. Wang, X. Zhu, C. Fan, Z. Yang, L. Tong, X. Duan, A. Pan, Nano Lett. 12 (2012) 5003–5007. [28] F. Gu, Z. Yang, H. Yu, J. Xu, P. Wang, L. Tong, A. Pan, J. Am. Chem. Soc. 133 (2011) 2037–2039. [29] S.M. Kim, K.K. Kim, Y.W. Jo, M.H. Park, S.J. Chae, D.L. Duong, C.W. Yang, J. Kong, Y.H. Lee, ACS Nano 5 (2011) 1236–1242. [30] H. Kitagawa, N. Kojima, T. Nakajima, J. Chem. Soc. Dalton Trans. (1991) 3121– 3125. [31] A.B. Smetana, K.J. Klabunde, C.M. Sorensen, A.A. Ponce, B. Mwale, J. Phys. Chem. B 110 (2006) 2155–2158. [32] T.W. Chang, M.R. Gartia, S. Seo, A. Hsiao, G.L. Liu, Nanotechnology 25 (2014) 145304–145312. [33] A.S.K. Hashmi, Chem. Rev. 107 (2007) 3180–3211. [34] M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chem. Lett. 2 (1987) 405–406. [35] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chem. Rev. 105 (2005) 1102– 1125. [36] I. Lisiecki, J. Phys. Chem. B 109 (2005) 12231–12244. [37] P.D. Cozzoli, M.L. Curri, C. Giannini, A. Agostiano, Small 2 (2006) 413–421. [38] L. Scarabelli, M. Coronado-Puchau, J.J. Giner-Casares, J. Langer, L.M. LizMarzán, ACS Nano 8 (2014) 5833–5842. [39] H. Duan, A.I. Fernández-Domínguez, M. Bosman, S.A. Maier, J.K.W. Yang, Nano Lett. 12 (2012) 1683–1689. [40] T.J. Collins, Biotechniques 43 (2007) S25–S30.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Vapor-phase preparation of gold nanocrystals by chloroauric acid pyrolysis Yiqin Chen a, Xuezeng Tian b, Wei Zeng a, Xupeng Zhu a, Hailong Hu c, Huigao Duan a,⇑ a
College of Physics and Microelectronics, State Key Laboratory for Chemo/Biosensing and Chemometrics, Hunan University, Hunan 410082, China Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China c School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link, 637371 Singapore, Singapore b
a r t i c l e
i n f o
Article history: Received 2 August 2014 Accepted 12 October 2014 Available online 22 October 2014 Keywords: Gold nanocrystals Vapor-phase deposition Chloroauric acid Pyrolysis Plasmon resonance
a b s t r a c t We report that gold nanocrystals can be prepared from vapor phase using chloroauric acid (HAuCl4) as the precursor. By tuning the vapor-phase deposition parameters, the size and space distribution of the gold nanocrystals can be well controlled on substrates. Systematic control experiments demonstrate that intermediate AuCl and AuCl3 products pyrolyzed from HAuCl4 play an essential role in this vapor-phase deposition process. Compared to conventional wet-chemical synthesis process, vapor-phase process enables direct deposition of gold nanoparticles on solid substrates with better coverage and uniformity, which may find applications in surface-enhanced Raman scattering and plasmon-enhanced photocatalysis. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Gold nanostructures have tremendous potential applications due to their unique electronic, optical and catalytic properties [1–6]. Most of their applications are strongly dependent on the size, shape and surface properties of gold nanostructures. Therefore, seeking for controllable methods to prepare gold nanostructures has been an active research area for the past decades. Generally, there are two main well-developed approaches in the preparation of gold nanostructures, i.e., top–down fabrication and bottom-up wet-chemical synthesis. Top–down technique provides freedom of design for planar patterns with high resolution and position accuracy for solid-state-device applications, but it usually involves lithographic process requiring costive vacuum facilities for large-scale fabrication [7,8]. Meanwhile, the fabricated nanostructures are polycrystalline with high surface roughness. Both grain boundaries and surface roughness result in the damping of hot electrons, which may degrade the electronic, optic and catalytic performances of gold nanostructures for applications [1,9,10]. On the contrary, wet-chemical synthesis could provide low-cost and large-scale preparation of monodispersed single-crystalline gold nanostructures with varied morphologies and surface chemistry for solution-based applications, but it requires an additional complicated assembling process for the ⇑ Corresponding author. E-mail address: [email protected] (H. Duan). http://dx.doi.org/10.1016/j.jcis.2014.10.017 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
specific applications on solid surface. As the nanoparticles tend to aggregate during the drying self-assembly process, it is difficult to obtain highly monodispersed nanoparticles with controlled space distribution on the solid surfaces, which limits the applications of wet-chemical synthesis in solid-state devices. Meanwhile, the surfactants, usually involved in the wet-chemical synthesis process, are difficult to completely remove from the particle surface, which have negative effects on the performance of gold nanoparticles in plasmonic and catalytic applications [11,12]. Therefore, developing alternative methods which have potential to overcome the limits of the existing top-down or wet-chemical techniques for preparing gold nanostructures is still interesting and required. Vapor-phase synthesis is a versatile, flexible and scalable technique to prepare nanostructures with high crystallinity and clean surface on various substrates [13–16]. Researchers have made many attempts to use vapor-phase method to prepare gold nanostructures. For example, Kim et al. demonstrated the preparation of high-quality single-crystalline gold nanowires via the vaportransport method using bulk gold pellets as the precursor [16,17]. However, this process requires temperature higher than 1000 °C to vaporize the bulk gold precursor. To lower the synthesis temperature, Igumenov et al. attempted to use metalorganic precursors to prepare gold nanoparticles [18]. In this case, to avoid the deposition of carbon substance, the deposition temperature should be as low as possible and thus it is difficult to obtain gold nanoparticles with high crystallinity.
22
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
In this work, we report a facile vapor-phase deposition process to prepare high-quality gold nanostructures with well-defined shapes using chloroauric acid (HAuCl4) as the precursor. It has been known that HAuCl4 is the most widely used reactant in wet-chemical synthesis of gold nanoparticles and can also be in situ decomposed into gold via pyrolysis or solution-based aerosol deposition process [19,20]. In our work, we found that HAuCl4 can be readily decomposed into evaporable products which can further transport to other solid substrates to form gold nanocrystals at relatively low temperature. We demonstrate that the size and space distribution of the gold nanocrystals can be tuned by controlling the deposition parameters in a tube-furnace system. With a systematic study and in-situ experiments, a possible mechanism is also proposed to explain this vapor-phase deposition process. With this unique technique, single-crystalline gold nanocrystals without any surface contaminations can be prepared in large scale on solid substrates, which have potential for plasmonic and catalytic applications. 2. Experimental section
(Thermo Scientific, ESCALAB-250Xi) was used to study the chemical valence of the materials. Specimens were analyzed in high vacuum environment (107 mbar, 25 °C) utilizing Al Ka at 20 kV with a power of 200 W. 2.2.3. In-situ TEM The sample for in-situ TEM experiment was prepared with HAuCl4 source temperature at 500 °C but TEM grid was put at room temperature in 9 Torr pressure. In-situ TEM experiment was carried out using a JEOL JSM-2010F Transmission Electron Microscope (200 kV accelerating voltage) which is integrated with a digital heating system. 2.2.4. The sample preparation procedures for Raman measurements We compared the Raman scattering signals between a 50 nm thick gold film and the sample prepared by our vapor-phase deposition method. We immersed two samples into 105 M Rhodamine-6G (R6G) aqueous solution for 12 h, then rinsed two substrates with ethanol completely and spontaneously dried them at room temperature. The rinsing process ensured the formation of only a monolayer of R6G over the surface of Au.
2.1. Vapor-phase deposition process To quantify the amount of gold element in the experiment each time, one gram HAuCl44H20 (purity > 99.999%, Sinopharm Chemical Reagent Co., Ltd) was dissolved in 40 ml ethanol absolute, and gold element concentration was 74 mM. For each deposition, 80 ll chloroauric acid solution was loaded it into quartz boat. Ethanol solution was feasible to evaporate as vapor phase under the condition of low pressure and mild heating before the decomposition of chloroauric acid. Note that HAuCl4 solution enables accurate loading of the HAuCl4 source, which is essential to the vapor-phase deposition process. The tube-furnace system was purchased from Hefei Kejing Corp. HAuCl4 source together with SiO2 (300 nm thickness on Si) substrates were put into 60-mm-diameter quartz tube equipped with a rotary pump and carrying gas (N2). HAuCl4 precursor was loaded in the upstream and the substrates were put in the downstream. The control of in-tube work pressure was realized by the dynamic equilibrium between injecting N2 and pump rate. Heating procedure is programmed by micro-computer integrated in the furnace system. For example, the tube furnace was first heated from room temperature to a certain high temperature (e.g. 600 °C) with a constant heating rate (e.g. 40 °C/min), then kept at the high temperature for a certain time, and finally cooled down to room temperature naturally by turning off the heating units. 2.2. Characterization 2.2.1. SEM and HRTEM The morphology of gold nanoparticles was characterized using Scanning Electron Microscopy (Hitachi S-4800 & FEI NanoLab Helios 600i) with 10 kV accelerating voltage and work distance of 8 mm. The as-prepared gold nanoparticles were on SiO2 substrates. To prepare TEM sample, we scratched the gold nanoparticles off the substrates to ethanol solution using the edge of a fresh silicon wafer. With ultrasonic bathing, we made the gold nanoparticles dispersed and then dripped the solution onto a copper TEM grid coated with an ultrathin carbon film (10 nm). The high resolution TEM metrology was done by a FEI Titan G2 machine with an accelerating voltage of 300 kV. 2.2.2. EDS and XPS A HORIBA energy dispersive X-ray spectroscopy attached to HITACHI scanning electron microscopy operated at 20 kV was characterized to target sample. X-ray photoelectron spectroscopy
2.2.5. Raman scattering measurements Raman scattering spectra of two selected samples were measured on a HORIBA JOBIN YVON HR800 UV Raman Spectroscopy imaging system. The excitation source was 532 nm. Raman signals were collected using a 50x objective (numerical aperture, N.A. = 0.5) under laser power intensity of 0.5 mW and with integration time of 60 s. 2.2.6. Dark-field imaging and scattering measurements Dark Field imaging and the scattering spectroscopy of single gold nanocrystal were measured on a WITec Confocal Raman Imaging Microscope System alpha 300R. A halogen lamp (15V, 3150 K) with 100 W power intensity was used as white light source and illuminated to the sample via an dark-field objective lens (N.A. = 0.75, 100). All of the samples were measured with the integration time of 20 s. The calculation of dark-field scattering spectrum was referred to the formula as follow:
Iscat ðkÞ ¼
Iparticle Isub Ihalogen Idark
where Iscat ðkÞ presents the real scattering spectra of single nanoparticle; the spectrum of Iparticle, Isub, was separately collected from then original particle and surrounding substrate. Ihalogen, Idark is the spectrum of halogen light source and background count of measurement system, respectively. 3. Results and discussion 3.1. Preparation and morphology characterizations Fig. 1a shows the schematic of the experimental setup for the vapor-phase deposition process. Fig. 1b shows a scanning electron microscopy (SEM) image of the resultant monodispersed gold nanocrystals with well-defined shapes on a SiO2 (300 nm thick on Si substrate) surface. We observed that most of the obtained nanocrystals are octahedral, as shown by the enlarged SEM images in Fig. 1b. The trend to form octahedra is because the gold nanoparticles tend to expose their {1 1 1} crystal facets due to the energy minimization during the growth process [21]. To further investigate the morphologic uniformity of the synthesized gold nanocrystals, we studied their plasmon resonance properties using singleparticle dark-field scattering measurements [22,23]. Fig. 1c shows the corresponding dark-field optical micrograph of the nanocrys-
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
23
Fig. 1. (a) Schematic of the experimental system. (b) SEM image of gold nanocrystals (170 nm size). (c) and (d) The corresponding optical dark-field image and dark-field scattering spectra of the labelled six gold nanoparticles. Experimental parameters: In 660 mTorr pressure, substrate and source were heated from 25 °C to 600 °C during 15 min with a distance of 6 cm, and kept at 600 °C for 2 h.
tals in Fig. 1b. The high-contrast colorful dark-field scattering image indicates that the plasmon resonance of these observed gold nanocrystals locates in the visible region. Further dark-field scattering spectra of the selected octahedra are shown in Fig. 1d. All of the measured particles show two plasmon resonance peaks locating at 570 nm and 675 nm, which well match the green and orange colors in the dark-field image and also indicate that the synthesized nanocrystals have consistent morphologies. The different scattering intensities are supposed to result from the different postures of the octahedra lying on the substrate. To prove the synthesized products are metallic gold nanocrystals with high crystallinity, we investigated the obtained nanoparticles using high-resolution transmission electron microscopy (HRTEM). Fig. 2a shows the SEM image of the octahedral nanoparticles used for the HRTEM characterization. The average edge length of the octahedron was 100 nm for this sample, and the yield of octahedral particles is over 75% (the statistics of more than 100 particles). Other shapes include decahedral, diamond-like and undefined geometry. We transferred these nanocrystals onto TEM grid with an ultrathin supporting carbon film of 10 nm for the HRTEM study. Fig. 2b presents the TEM image of a typical octahedron, and the presumable posture of this particle on the carbon film is given by the inset schematic. From the HRTEM image of a corner of this particle, as shown in Fig. 2c, we can clearly see the lattice fringes with a interplanar spacing of 2.35 angstrom ascribed to {1 1 1} plane of face-center cubic (fcc) gold. Further selected-area electron diffraction (SAED) pattern of this particle, shown in Fig. 2d, suggests the high crystallinity of the nanocrystal. The above TEM results undoubtedly confirm that the vapor-phase prepared nanocrystals from HAuCl4 are metallic gold nanoparticles. More HRTEM results of gold nanocrystals with various sizes and shapes prepared at different parameters are provided in the supplementary information (Figure S1).
The size and the space distribution of the nanocrystals can be controlled by tuning the deposition parameters, such as growth temperature, source-to-substrate distance, flow rate of the carrying gas, which are similar with chemical vapor deposition process for other nanostructures [24–26]. As a proof of concept, we demonstrated that the size of gold nanocrystals can be finely tuned with different substrate-to-source distance, as shown in Fig. 3. When the substrates were put farther from the HAuCl4 source in the downstream, the obtained gold nanocrystals became smaller. The average diameter of the nanocrystals decreased from 67 nm to 10 nm when the substrate-to-source distance increased from 3 cm to 6 cm, as indicated by the statistical diagrams. The observed result can be attributed to the different precursor concentration at the different substrate positions [27,28]. It should be noted that though the obtained nanocrystals were highly monodispersed with a uniform size distribution for each substrate, we did not observe clear trend of the particle density when varying the substrate-tosource distance, which might be caused by the instability of the gas flow in the tube furnace during the deposition process. Meanwhile, for the influence of different deposition temperature, we found that the resultant gold nanoparticles became smaller when decreasing the deposition temperature. The details can be seen in Figure S2 in the supplementary information. 3.2. Mechanism study As a common precursor, HAuCl4 is widely used for the synthesis of gold nanoparticles, but the fact that it undergoes the decomposition into evaporable products at low temperature (as low as 100 °C according to our experimental results shown in Figure S3 in the supplementary information) has never been clearly studied. A further look at the mechanism of this vapor-phase deposition process is not only of significance for the greater control of the
24
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
(a)
(b)
20 nm
200 nm
(d)
(c)
(220)
(002)
(020) 2 nm
(111)
Fig. 2. (a) SEM image of gold nanocrystals (100 nm) which were used for the TEM characterization. Inset shows an enlarged SEM image of a single gold octahedral nanocrystal. (b) TEM image of a typical gold nanocrystal. The inset gives the schematic side view of the gold octahedron. (c) High-resolution TEM micrograph at the corner of gold nanocrystal (marked in Fig. 2(b). (d) The selected-area electron diffraction pattern of the nanocrystal. Experimental parameters: In 600 mTorr pressure, substrate and source were heated from 25 °C to 500 °C during 12.5 min with a distance of 6 cm, and kept at 500 °C for 2 h.
60 50 40 30 20
σ=3.7 nm
70 60 50 40 30 20 10
10 0
80
0 10 20 30 40 50 60 70 80 90 100
Diameter (nm)
0
100
90
Avg=34.7 nm
80 70
Avg=22.5 nm
90
Avg=10.9 nm
σ=1.1 nm
80
σ=1.1 nm
60 50 40 30 20 10
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
0
Frequency (%)
70
(d)
100
90
Frequency (%)
Frequency (%)
80
100
Avg=67.3 nm σ=9.3 nm
Frequency (%)
100 90
(c)
(b)
(a)
70 60 50 40 30 20 10
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
0
0
10 20 30 40 50 60 70 80 90
Diameter (nm)
Fig. 3. SEM images of gold nanocrystals prepared with different source-to-substrate distances. (a) 3 cm, (b) 4 cm, (c) 5 cm, and (d)p6ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cm. Statistic histograms of the size of gold ffi nanocrystals on the SiO2 substrates are shown in the bottom panel. Over 100 nanoparticles were counted for statistics (d ¼ 2 0:385A). A is the area of the nanoparticles, which is obtained by ImageJ software [29]. All of the samples were prepared at 500 °C and 600 mTorr. All scale bars are 200 nm.
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
preparation of gold nanocrystals using this method, but also provides insights to understand other HAuCl4-related reaction process. To study the mechanism, we did both ex-situ SEM and insitu TEM experiments to understand the mass-transport and nucleation-growth process in this vapor-phase deposition process. The first thing we tried to do was to determine the intermediate vapor products during the deposition process. Metallic gold can be first ruled out as an intermediate vapor product due to its melting temperature is higher than 1000 °C while our deposition temperature can be as low as 100 °C in supplementary information (see Figure S3). It is known that HAuCl4 can be easily decomposed into AuCl3 and HCl under thermal treatment with a temperature as low as 110 °C [19]. To explain the vapor-phase process, we further assume that the decomposed AuCl3 is able to vaporize and transport to distant regions, so the reactive equation can be written as: D
HAuCl4 ðsÞ ! AuCl3 ðgÞ þ HClðgÞ:
ð1Þ
To verify our assumption, we deposited some product onto a SiO2 substrate at room temperature by putting the substrate at the outside of the heating zone. The source temperature was set to be 500 °C. A typical SEM image of the product was provided in Figure S4a in supplementary information, which shows the film-like morphology on the substrate. With the help of energy dispersive X-ray spectroscopy (EDX or EDS), we found both gold and chlorine elements exist in the film (see Figure S4b in supplementary information). We further studied the chemical structure of as-deposited film by X-ray photoelectron spectroscopy (XPS). The peak of 91.1 eV near in Au (III) 4f5/2 and 88.6 eV near in Au (I) 4f5/2 was clearly shown in Figure S4c in supplementary information, respectively [30,31]. Au (I) existed in the as-deposited film owing to the source temperature of 500 °C. Under high-temperature environment, vaporized AuCl3 as stated in Eq. (1) simultaneously might occur the reaction as follow: D
AuCl3 ðgÞ ! AuClðgÞ þ Cl2 ðgÞ:
ð2Þ
Hence, the existence of Au (I) in film was originated from further decomposition of vaporized AuCl3. Vapor-phase AuCl3 and AuCl were together transferred with carried gas flow and deposited onto distant substrate at room temperature. Therefore, as-deposited film was a mixture of AuCl3 and AuCl. With the above assumption, we further investigated the nucleation growth process of the gold
25
nanocrystals. A possible growth mechanism is shown in Fig. 4a and can be briefly described as follows. Firstly, a homogeneous AuCl3/AuCl film is rapidly deposited onto the substrate during the pyrolysis process at a low pressure. Under continuous thermal treatment, the AuCl3/AuCl decompose into Au to form ultrasmall monodispersed gold seeds embedded in AuCl3/AuCl film. These gold seeds further grow to be gold nanoparticles due to the continuous decomposition of the remaining AuCl3/AuCl. During the above processes, small gold nanoparticles may merge each other due to Ostwald ripening [32] or Rayleigh instability [33]. To further confirm the above hypothesis on growth mechanism, we carried out a series of thermal treatments on the AuCl3/AuCl precursor film obtained on a substrate put at the room temperature region. The results are shown by the SEM images in Fig. 4b– d. Fig. 4b shows the as-deposited film, and no obvious gold nanoparticles can be seen. After a thermal treatment at 100 °C for 30 min (Fig. 4c), the continuous AuCl3/AuCl film transferred to isolated droplets in which gold seeds were wrapped by surrounding AuCl3/AuCl product. Note that the gold nanoparticles are much brighter in the SEM images due to their higher secondary electron yields compared to AuCl3/AuCl. Further increasing the reaction time to 60 min resulted in larger gold nanocrystal seeds and less remaining AuCl3/AuCl, as shown in Fig. 4d. By elevating the thermal-treatment temperature to 400 °C, shape-defined gold nanocrystals with larger size and clean surface formed, as shown in Fig. 4e. These series of results clearly demonstrate the scenarios of deposition of AuCl3/AuCl precursor film, formation of tiny gold seeds, and growth of the gold nanocrystals from the AuCl3/AuCl precursor, which provide sufficient and convincible evidences to verify the crystallization and growth mechanism as indicated by the schematics. From above results, we also surprisingly see that AuCl3/AuCl product can be slowly reduced to Au (0) at a temperature as low as 100 °C. It is known that disproportionation reaction could occur in AuCl and transfer into AuCl3 and Au (0) at relatively low temperature [34]. The possible reaction equation is given below: D
3AuClðsÞ ! AuCl3 ðsÞ þ 2AuðsÞ:
ð3Þ
Though the existing literature only report a lowest reaction temperature of 260 °C [19], lower reaction temperature could be possible at the nanometer scale [35]. Once tiny gold nanoparticles were
Fig. 4. (a) A series of intuitive schematics showing different stages of the gold nanocrystals in the vapor-phase deposition process. (b)–(e) Ex-situ SEM images of experimental results to show the different stages of the gold nanocrystals under a mild thermal treatment. All SEM images were done on the same sample. The AuCl3/AuCl film was deposited at 22 Torr pressure with source temperature of 500 °C and substrate temperature of 25 °C (i.e. room temperature). The scale bars are 200 nm.
26
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
(b)
(a)
(c)
(d)
Thin film Thick film Thick film Thin film
100 nm
(e)
25e C (222) (220) (200) (111)
50 nm
(f)
200e C
(220)
50 nm (220)
(g)
(111)
400e C
50 nm
C 600e
(h) (220) (111)
(111)
Fig. 5. In-situ TEM observation of the dynamic evolution of gold nanoparticles from the AuCl3/AuCl precursor film. (a)–(d) TEM micrographs show the morphologies of the formed gold nanoparticles after a series of thermal annealing under different temperatures. (e)–(h) The electron diffraction patterns of the samples corresponding to TEM images in (a)–(d), respectively.
30.0k
(a) Intensity (Counts)
25.0k
(b) 614 1362
20.0k 776 15.0k
1185 1311
1509 1650 1570
10.0k 5.0k
100 nm
0.0
600
800
1000
1200
1400
1600
Raman Shift (cm-1) Fig. 6. (a) A SEM image of gold nanocrystals (100 nm) on SiO2 substrate with a high particle density, and amount of sub-10-nm gaps (highlighted by red arrows) can be formed. (b) The Raman spectra of R6G molecules on high-density gold nanocrystals (blue curve) and a gold film (red curve). Both the source and substrate temperature for this sample were 500 °C. All of the other experimental parameters were the same as described in Fig. 3. The sample was away from the source with a distance of 3 cm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
formed in the AuCl3/AuCl film, they may serve as catalytic sites for particle growth via a self-catalysis growth model [36–38], resulting in the formation of larger nanocrystals at a relatively low temperature. To further demonstrate the formation of gold nanocrystals from the vapor-deposited AuCl3/AuCl film under thermal treatment, insitu TEM study was done. To do this, we first deposited an AuCl3/ AuCl film on the TEM grid with an ultrathin continuous carbon film of 10 nm, and the experimental condition for AuCl3/AuCl deposition was the same as that in Fig. 4. In order to minimize the electron-beam influence, the electron beam was highly dispersed for imaging and was shut off during heating. A representative TEM image of the as-deposited AuCl3/AuCl film is shown in Fig. 5a, from which we can see that the deposited film is almost homogeneous on the substrate and no obvious gold nanoparticles can be observed, similar with the result obtained from SEM image Fig. 4b. The corresponding electron diffraction pattern of the asdeposited film is given by Fig. 5e. Interestingly, from the diffraction pattern, we can see the diffraction rings of the gold, indicating the
existent of ultrasmall sub-nanometer gold nanocrystals in the deposited film. Fig. 5b–d show the evolution of the gold nanoparticles obtained at different annealing temperature and time. Even with a low temperature at 200 °C, amount of AuCl3/AuCl precursor transferred into gold nanoparticles with an average diameter less than 5 nm, as shown in Fig. 5b. It is worth noting that relatively larger gold nanoparticles formed at the original darker area (marked by ‘‘thick film’’ in Fig. 5a) than those in brighter area (marked by ‘‘thin film’’) due to the different thickness of initial AuCl3/AuCl film, indicating that the thickness of AuCl3/AuCl film was a key role to the size of gold nanoparticles. While elevating the annealing temperature, the size of the gold nanoparticles gradually increased and the particle density decreased, as shown in Fig. 5c and d. Further electron diffraction results, shown in Fig. 5f–h, prove the gradual crystallization of the gold nanoparticles. It should be noted that due to the different wetting properties between gold and carbon film, the final formed gold nanoparticles did not have well-defined shapes, which is different from what we observed on the SiO2 substrates. To investigate how the wetting properties of the substrates
Y. Chen et al. / Journal of Colloid and Interface Science 439 (2015) 21–27
affect the synthesis results, we further did the control experiments on carbon fiber and silicon substrate (see Figure S5 in supplementary information). All of the resultant gold nanoparticles on the two different substrates do not show well-defined shape. Therefore, in the in-situ TEM study, the unwell-defined shape of synthesized gold nanoparticles was attributed to the different wetting property of carbon film compared to SiO2 substrate. 3.3. SERS application A promising application area of gold nanostructures is to utilize their plasmonic properties for sensing applications. One of the advantages of this vapor-phase method is its potential to obtain gold nanocrystals in large scale with clean surface and tunable particle density, which may be applied in various applications such as surface-enhanced Raman spectroscopy (SERS) for molecule-scale detections. Fig. 6a shows a SEM image of high-density gold nanocrystals in a large area and the particle density is 116 counts/ lm2. The sample was prepared with an appropriate source-to-substrate distance. However, due to the high-density distribution of gold nanoparticles, adjacent particles might coalesce together, resulting in the formation of irregular gold nanoparticles. When using the samples as SERS substrates to detect R6G molecules with an excitation wavelength at 532 nm, the Raman scattering intensities were significantly enhanced with a factor of 103 compared to common gold film (50 nm thick) substrate. Such a significant enhancement factor can be attributed to the local surface plasmon resonance of these monodispersed high-quality gold nanocrystals. Furthermore, due to the high density of the nanocrystals, amount of small gaps less than 10 nm can be formed between the nanocrystals. These tiny gaps could significantly enhance the electric field of the incident light [39,40] to form so-called ‘‘hot spots’’ for Raman scattering. Further improvement of the enhancement factor may be achieved in the future by optimizing the particle density to obtain more sub-10-nm gaps. 4. Conclusion In summary, we reported a vapor-phase method to prepare gold nanocrystals by pyrolyzing a common HAuCl4 precursor at relatively low temperature without using any reducing agents and catalysts. With this method, we demonstrated that large-scale singlecrystalline gold nanocrystals with well-defined shapes can be synthesized on solid substrates. The systemic study of mechanism make us understood the process of crystallization. It is helpful to tune shape and size of crystals by feasible experiment optimum. Meanwhile, we can analogize that this vapor-phase method could also be used to prepare gold nanorods and nanowires by introducing appropriate assistant reactants, and similar process may be applied to HPtCl4 and HPdCl4 to prepare Pt and Pd nanostructures. This method provides a good candidate to obtain super-clean sample for improving the accuracy of molecular probing or to study reactive dynamics on metal nanocrystals surface in catalysis. Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant Nos. 11274107, 61204109), Program for New Century Excellent Talents in University (Grant No. NCET-130185), Foundation for the Author of National Excellent Doctoral Dissertation of China (Grant No. 201318), Interdisciplinary Program of Hunan University, and Open Project of State Key Laboratory of Solid State Microstructures in Nanjing University.
27
We are grateful to Yufang Ma for Raman scattering measurement and to Zhengmei Yang for helpful discussion. We also thank Wei Chen in National University of Defense Technology for SEM measurements. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.10.017. References [1] Y. Zhao, J. Chu, S.H. Li, W.W. Li, G. Liu, Y.C. Tian, H.Q. Yu, Electroanalysis 26 (2014) 656–663. [2] H. Hu, H. Duan, J.K.W. Yang, Z.X. Shen, ACS Nano 6 (2012) 10147–10155. [3] A. Gopalakrishnan, M. Chirumamilla, F. De Angelis, A. Toma, R.P. Zaccaria, R. Krahn, ACS Nano 8 (2014) 7986–7994. [4] W.L. Chen, F.C. Lin, Y.Y. Lee, F.C. Li, Y.M. Chang, J.S. Huang, ACS Nano 8 (2014) 9053–9062. [5] A. Corma, H. Garcia, Chem. Soc. Rev. 37 (2008) 2096–2126. [6] M. Valden, X. Lai, D.W. Goodman, Science 281 (1998) 1647–1650. [7] H. Duan, H. Hu, K. Karthik, Z.X. Shen, J.K.W. Yang, ACS Nano 5 (2011) 7593– 7600. [8] Y. Wang, M. Abb, S.A. Boden, J. Aizpurua, C.H. Groot, O.L. Muskens, Nano Lett. 13 (2013) 5647–5653. [9] P. Nagpal, N.C. Lindquist, S.H. Oh, D.J. Norris, Science 325 (2009) 594–597. [10] K.D. Schierbaum, U. Weimar, W. Göpel, Sens. Actuators B 3 (1991) 205–214. [11] W.E.A. Moerner, J. Phys. Chem. B 106 (2002) 910–927. [12] G. Chen, Y. Tan, B. Wu, G. Fu, N. Zheng, Chem. Commun. 48 (2012) 2758–2760. [13] P. Mohanty, I. Yoon, T. Kang, K. Seo, K.S.K. Varadwaj, W. Choi, Q.H. Park, J.P. Ahn, Y.D. Suh, H. Ihee, B.S. Kim, J. Am. Chem. Soc. 129 (2007) 9576–9577. [14] Y. Yoo, I. Yoon, H. Lee, J. Ahn, J.P. Ahn, B. Kim, ACS Nano 4 (2010) 2919–2927. [15] H. Choi, S.H. Park, J. Am. Chem. Soc. 126 (2004) 6248–6249. [16] Y. Yoo, K. Seo, S. Han, K.S.K. Varadwaj, H.Y. Kim, J.H. Ryu, H.M. Lee, J.P. Ahn, H. Ihee, B. Kim, Nano Lett. 10 (2010) 432–438. [17] I. Yoon, T. Kang, W. Choi, J. Kim, Y. Yoo, S.W. Joo, Q.H. Park, H. Ihee, B. Kim, J. Am. Chem. Soc. 131 (2009) 758–762. [18] R.G. Parkhomenko, N.B. Morozova, G.I. Zharkova, Y.V. Shubin, S.V. Trubin, V.V. Kriventsov, B.M. Kuchumov, T.P. Koretskaya, Chem. Vap. Deposition 18 (2012) 336–342. [19] R.G. Palgrave, I.P. Parkin, Chem. Mater. 19 (2007) 4639–4647. [20] H. Shi, L. Zhang, W. Cai, Mater. Res. Bull. 35 (2000) 1689–1695. [21] M. Grzelczak, J. Pérez-Juste, P. Mulvaney, Chem. Soc. Rev. 37 (2008) 1783– 1791. [22] J.J. Mock, M. Barbic, D.R. Smith, D.A. Schultz, S. Schultz, J. Chem. Phys. 116 (2002) 6755–6759. [23] K.H. Su, Q.H. Wei, X. Zhang, Nano Lett. 3 (2003) 1087–1090. [24] X. Li, W.C. Magnuson, A. Venugopal, J. An, J.W. Suk, B. Han, M. Borysiak, W. Cai, A. Velamakann, Y. Zhu, L. Fu, M.E. Vogel, E. Voelkl, L. Colombo, S.R. Ruoff, Nano Lett. 10 (2010) 4328–4334. [25] Z. Li, P. Wu, C. Wang, X. Fan, W. Zhang, X. Zhai, C. Zeng, Z. Li, J. Yang, J. Hou, ACS Nano 5 (2011) 3385–3390. [26] H. Peng, Z. Pan, L. Xu, X. Fan, N. Wang, C.S. Lee, S.T. Lee, Adv. Mater. 13 (2001) 317–320. [27] J. Xu, X. Zhuang, P. Guo, Q. Zhang, W. Huang, Q. Wan, W. Hu, X. Wang, X. Zhu, C. Fan, Z. Yang, L. Tong, X. Duan, A. Pan, Nano Lett. 12 (2012) 5003–5007. [28] F. Gu, Z. Yang, H. Yu, J. Xu, P. Wang, L. Tong, A. Pan, J. Am. Chem. Soc. 133 (2011) 2037–2039. [29] S.M. Kim, K.K. Kim, Y.W. Jo, M.H. Park, S.J. Chae, D.L. Duong, C.W. Yang, J. Kong, Y.H. Lee, ACS Nano 5 (2011) 1236–1242. [30] H. Kitagawa, N. Kojima, T. Nakajima, J. Chem. Soc. Dalton Trans. (1991) 3121– 3125. [31] A.B. Smetana, K.J. Klabunde, C.M. Sorensen, A.A. Ponce, B. Mwale, J. Phys. Chem. B 110 (2006) 2155–2158. [32] T.W. Chang, M.R. Gartia, S. Seo, A. Hsiao, G.L. Liu, Nanotechnology 25 (2014) 145304–145312. [33] A.S.K. Hashmi, Chem. Rev. 107 (2007) 3180–3211. [34] M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chem. Lett. 2 (1987) 405–406. [35] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chem. Rev. 105 (2005) 1102– 1125. [36] I. Lisiecki, J. Phys. Chem. B 109 (2005) 12231–12244. [37] P.D. Cozzoli, M.L. Curri, C. Giannini, A. Agostiano, Small 2 (2006) 413–421. [38] L. Scarabelli, M. Coronado-Puchau, J.J. Giner-Casares, J. Langer, L.M. LizMarzán, ACS Nano 8 (2014) 5833–5842. [39] H. Duan, A.I. Fernández-Domínguez, M. Bosman, S.A. Maier, J.K.W. Yang, Nano Lett. 12 (2012) 1683–1689. [40] T.J. Collins, Biotechniques 43 (2007) S25–S30.
