Supplementary Figure 1 In situ QCM measurements of Pd ALD on different oxide

Transcription

1 Supplementary Figure 1 In situ QCM measurements of Pd ALD on different oxide surfaces using either HCHO or H 2 as reducing agent at 150 ºC. (a) On Al 2 O 3 ; (b) on ZrO 2 ; (c), on TiO 2. See Supplementary Notes 1 and 3. 1

2 Supplementary Figure 2 In situ QCM measurements of Pt ALD on alumina surface using O 2 as oxidizing agent at different temperatures. Oxygen partial pressure during exposure was 0.1 Torr. See Supplementary Note 2. 2

3 Supplementary Figure 3 In situ QCM measurements of Pt ALD on different oxide surfaces using either O 3 or O 2 as oxidizing agent at 150 ºC. (a) On Al 2 O 3 ; (b) on ZrO 2 ; (c) on TiO 2. On Al 2 O 3 surface, Pt ALD was also carried out using different O 2 partial pressures at 0.1 and 0.6 Torr to study the pressure effect. On ZrO 2 and TiO 2 surfaces, O 2 partial pressure was kept at 0.1 Torr. 3

4 Supplementary Figure 4 In situ QCM measurements of Ru ALD at low temperatures. a, comparison of Ru ALD using the sequences of Ru(EtCp) 2 -O 2, Ru(EtCp) 2 -H 2, and Ru(EtCp) 2 -O 2 -H 2 at 200 ºC. b, ABC-type Ru ALD using a sequence of Ru(EtCp) 2 -O 2 -H 2 at 150 and 200 ºC, respectively. See Supplementary Note 4. 4

5 Supplementary Figure 5 Growth behaviors of PtPd alloy ALD at 150 ºC. (a) In situ QCM measurements of PtPd alloy ALD using the sequence of MeCpPtMe 3 -O 2 -Pd(hfac) 2 -H 2, and MeCpPtMe 3 -O 2 -H 2 -Pd(hfac) 2 -H 2 -O 2, respectively. (b) In situ QMS measurements of CO 2 formation (m/e= 44) during these two PtPd alloy ALD processes. See Supplementary Note 5. 5

6 Supplementary Figure 6 In situ QCM measurements of PtPd ALD at 150 ºC. (a) PtPd ALD using the sequence of MeCpPtMe 3 -O 2 -H 2 -Pd(hfac) 2 -H 2, and MeCpPtMe 3 -O 2 -H 2 -Pd(hfac) 2 - H 2 -O 2, respectively. (b) Detailed mass changes during the individual precursor exposures for the PtPd ALD. The sequential exposure steps of each reactant are shown as columns in color at the bottom of each figure, nitrogen purge step is in between the columns. See Supplementary Note 5. 6

7 Supplementary Figure 7 A representative STEM image of Pd/Al 2 O 3 sample synthesized by 4c Pd ALD using Pd(hfac) 2 and HCHO at 200 ºC. Scale bar, 10 nm. 7

8 a b Supplementary Figure 8 Representative STEM images of Pt/Al 2 O 3 sample synthesized by 12c Pt ALD using MeCpPtMe 3 and O 3 at 150 ºC. (a) A STEM image at low magnification. Scale bar 20 nm. (b) A STEM image at high magnification. Scale bar 5 nm. 8

9 Supplementary Figure 9 In situ CO chemisorption FTIR difference spectra recorded during well-mixed PdPt alloy synthesis by ALD at 150 ºC. Here, difference spectra are presented to highlight the spectral changes produced by each of the metal components. The difference spectrum for each ALD step is obtained by subtracting the spectrum recorded after the previous step. 9

10 a b Supplementary Figure 10 A representative aberration-corrected HAADF-STEM image and corresponding EDS line profiles of ALD 5Pd-core 15Pt-shell nanoparticles on spherical alumina support. Scale bar, 20 nm. 10

11 a b Supplementary Figure 11 A representative aberration-corrected HAADF-STEM image and corresponding EDS line profiles of ALD 12Pt-core 20Pd-shell nanoparticles on spherical alumina support. Scale bar, 20 nm. 11

12 a b Supplementary Figure 12 A representative aberration-corrected HAADF-STEM image and corresponding EDS line profiles of ALD 12Pt10Pd alloy nanoparticles on spherical alumina support. Scale bar, 10 nm. 12

13 a b Supplementary Figure 13 A representative aberration-corrected HAADF-STEM image and corresponding EDS line profiles of ALD 1c-Pt core 35c-Ru rich shell nanoparticles on spherical alumina support. Scale bar, 10 nm. 13

14 Supplementary Figure 14 XRF spectra of graphene supported ALD Ru, RuPd, Pt, and PtPd samples. See Supplementary Note 6. 14

15 Supplementary Note 1 Pd ALD using HCHO vs H 2 Using HCHO as reducing agent, Pd ALD can be achieved above 110 ºC, even though the initial nucleation is slower at lower temperature 1. However, using H 2 as reducing agent, the Pd ALD on oxide surfaces was largely inhibited, which is consistent with previous literature 2 (Supplementary Figure 1). It is noteworthy that the mass uptake in the first cycle is mainly caused by the heavy hfac ligands, which was observed using FTIR 3. Therefore, depositing Pd nanoparticles on oxide surfaces with controlled particle size can be obtained using HCHO as the reducing agent. The formed Pd nanoparticles can be used as either monometallic catalysts or further as the nucleation seeds for selective deposition of other metals to grow bimetallic or multi-components nanoparticles. On the other hand, we could use H 2 as the reducing agent for Pd ALD to achieve selective deposition on other pre-existing metal nanoparticle surfaces to grow Pd shell or alloy nanoparticles (Fig. 1a and 2). Supplementary Note 2 Pt ALD at different temperatures Pt ALD showed a nucleation period of slower growth for the first 15 cycles, then quickly reached its steady-state growth rate of ~ 110 ng/cm 2 per cycle (0.51 Å/cycle) at 300 ºC, as shown in Supplementary Figure 2. Pt ALD was largely hampered at 200 ºC and completely inhibited at 150 ºC with the same oxygen exposure. This observation is consistent with previous literature, where Aaltonen et al. reported that Pt ALD can be performed using MeCpPtMe 3 and pure oxygen at deposition temperature as low as 200 C 4. Supplementary Note 3 Pt ALD using O 3 vs O 2 Similar to the observation for Pd ALD in Supplementary Figure 1, the nucleation and growth of Pt ALD on oxide surfaces can be controlled by using either O 3 or O 2 as 15

16 the oxidizing agent at 150 ºC. Depositing Pt nanoparticles on oxide surfaces with controlled particle size can be achieved using O 3 as the oxidizing agent at 150 ºC. The formed Pt nanoparticles can be used as either monometallic catalysts or further as the nucleation seeds for selective deposition of other metals to grow bimetallic or multi-components nanoparticles. On the other hand, Pt ALD using O 2 as the oxidizing agent at 150 ºC can provide selective deposition on other pre-existing metal nanoparticle surfaces to grow alloy or Pt shells, since there is no nucleation delay to grow on metal surfaces (Fig.1b). Supplementary Note 4 Low temperature ABC-type Ru ALD Ru ALD using Ru(EtCp) 2 and O 2 is typically performed at 300 C 5. Here we further explored the lower temperature limit for the Ru ALD. We found no growth at 200 C using either O 2 or H 2 as shown in Supplementary Figure 4a. However, a linear growth rate of 22.4 ng/cm 2 per cycle was achieved using an ABC-type method consisting of an exposure sequence of Ru(EtCp) 2, O 2, and H 2 ( ). Moreover, with this ABC-type method, Ru ALD could be achieved as low as 150 C with a growth rate of 10.5 ng/cm 2 per cycle (Supplementary Figure 4b). This new ABC-type Ru ALD is similar to a recent study where Hämäläinen et al. reported that Ir ALD can be achieved using Ir(acac) 3, O 3, and H 2 at low temperatures between 165 and 200 C 6. The success of Ru ALD at low temperatures using this new method allows growing supported bimetallic nanoparticles by combining with other ALD metals, such as Pd and Pt at 150 C. Supplementary Note 5 Surface chemistry compatibility between Pt and Pd ALD The surface chemistry for the Pt and Pd ALD are very different. For instance, surface O species are the active sites for Pt ALD during the MeCpPtMe 3 exposures, while surface H species are the active sites for Pd ALD 16

17 during the Pd(hfac) 2 exposures. In order to investigate the compatibility of these two processes for growing PtPd alloys, a direct combination of the two processes was attempted first. As shown in Supplementary Figure 5a, the PtPd alloy prepared using AB-type ALD processes for both metals showed a linear growth rate of 90.9 ng/cm 2 per alloy cycle at 150 ºC. Considering the active sites required by the two processes, we next attempted to improve the compatibility by adding a H 2 reduction step after each Pt ALD cycle to prepare an H-terminated metal surface for the next Pd ALD, and an O 2 oxidation step after each Pd ALD cycle to generate a metal oxide surface for the next Pt ALD. Indeed, a growth rate of ng/cm 2 per alloy cycle was observed at 150 ºC, considerably higher than the one using the AB method. The higher growth rate of the alloy growth is more desirable for growing supported PdPt alloy nanoparticles, because it enhances the selective growth on the metal particle surface but not on the oxide support. To investigate the ABC-type PdPt alloy ALD, we performed in situ quadrupole mass spectrometry (QMS) measurements as described previously 7. As shown in Supplementary Figure 5b, in situ QMS revealed that CO 2 was formed during the Pd(hfac) 2 exposures when the Pt and Pd ALD were performed using the conventional AB-type processes. This finding suggests that hfac ligands were partially decomposed on the Pt oxide surface during the previous Pt ALD cycle. In contrast, no CO 2 was observed during the Pd(hfac) 2 exposures when Pt and Pd ALD were performed using the ABC-type processes. Therefore, the decomposition of Pd(hfac) 2 by surface oxygen is likely to be the origin for the reduced PtPd alloy growth rate using the AB-type processes. On the other hand, combining the ABC-type Pt with the AB-type Pd ALD showed almost no change compared to using both ABC-type process as seen in Supplementary Figure 6a. 17

18 However the mass gains for the Pt and Pd in each cycle changed to 76.6 ng/cm 2 and 54.9 ng/cm 2, respectively, giving a mole ratio of Pd:Pt = 4:3 (Supplementary Figure 6b). Supplementary Note 6 Selective metal ALD on carbon support Bimetallic nanoparticles supported on carbon have broad interest in both heterogeneous catalysis and in electrocatalysis, such as fuel cells. In order to compare the nucleation behavior on metal and carbon surfaces, the second metal, either 10-cycle Pt or10-cycle Ru ALD, was performed on carbon support with and without the presence of Pd nanoparticles at 150 ºC. Here the carbon powder support was stacked graphene platelet nanochips (Strem Chemicals). We used the graphene platelet nanochips after grinding without other treatments. Pd nanoparticles supported on graphene were prepared by 10-cycle Pd ALD using Pd(hfac) 2 and HCHO at 200 ºC (10cPd-graphene). As shown in Supplementary Figure 14, the amounts of Pt and Ru were significantly smaller on the bare graphene compared with the 10cPd-graphene, which provides evidence for selective metal ALD on pre-existing metal particles surface but not on the carbon support. Supplementary References 1 Lu, J. L. & Stair, P. C. Nano/Subnanometer Pd nanoparticles on oxide supports synthesized by AB-type and low-temperature ABC-type atomic layer deposition: Growth and morphology. Langmuir 26, (2010). 2 Senkevich, J. J. et al. Substrate-independent palladium atomic layer deposition. Chem Vapor Depos 9, (2003). 3 Goldstein, D. N. & George, S. M. Surface poisoning in the nucleation and growth of palladium atomic layer deposition with Pd(hfac)(2) and formalin. Thin Solid Films 519, (2011). 4 Aaltonen, T. et al. Atomic layer deposition of noble metals: Exploration of the low limit of the deposition temperature. Journal of Materials Research 19, (2004). 5 Kim, W. H., Park, S. J., Kim, D. Y. & Kim, H. Atomic layer deposition of ruthenium and ruthenium-oxide thin films by using a Ru(EtCp)(2) precursor and oxygen gas. J Korean Phys Soc 55, (2009). 18

19 6 Hamalainen, J. et al. Iridium metal and iridium oxide thin films grown by atomic layer deposition at low temperatures. J. Mater. Chem. 21, (2011). 7 Christensen, S. T. & Elam, J. W. Atomic Layer Deposition of Ir-Pt Alloy Films. Chem Mater 22, (2010). 19