International Journal of

Molecular Sciences Article

An Efficient, Versatile, and Safe Access to Supported Metallic Nanoparticles on Porous Silicon with Ionic Liquids Walid Darwich 1 , Paul-Henri Haumesser 2,3, *, Catherine C. Santini 1 and Frédéric Gaillard 2,3 1 2 3

*

CNRS-Université de Lyon-ESCPE Lyon, UMR 5265 C2P2, 43 Bd du 11 Novembre 1918, Villeurbanne 69616, France; [email protected] (W.D.); [email protected] (C.C.S.) CEA, LETI, MINATEC Campus, Grenoble 38054, France Université Grenoble Alpes, Grenoble 38000, France; [email protected] Correspondence: [email protected]; Tel.: +33-4-3878-5759; Fax: +33-4-3878-3034

Academic Editors: Andreas Taubert and Peter Hesemann Received: 30 March 2016; Accepted: 24 May 2016; Published: 3 June 2016

Abstract: The metallization of porous silicon (PSi) is generally realized through physical vapor deposition (PVD) or electrochemical processes using aqueous solutions. The former uses a strong vacuum and does not allow for a conformal deposition into the pores. In the latter, the water used as solvent causes oxidation of the silicon during the reduction of the salt precursors. Moreover, as PSi is hydrophobic, the metal penetration into the pores is restricted to the near-surface region. Using a solution of organometallic (OM) precursors in ionic liquid (IL), we have developed an easy and efficient way to fully metallize the pores throughout the several-µm-thick porous Si. This process affords supported metallic nanoparticles characterized by a narrow size distribution. This process is demonstrated for different metals (Pt, Pd, Cu, and Ru) and can probably be extended to other metals. Moreover, as no reducing agent is necessary (the decomposition in an argon atmosphere at 50 ˝ C is fostered by surface silicon hydride groups borne by PSi), the safety and the cost of the process are improved. Keywords: porous silicon; metallization; surface reactivity; metallic nanoparticles

1. Introduction Porous silicon (PSi) has been attracting a great amount of attention as a breakthrough material with exceptional characteristics for microelectronics, integrated optoelectronics, microelectromechanical systems (MEMS), layer transfer technology, solar and fuel cells, biomedicine, etc. [1,2]. Depending on the conditions for its elaboration, PSi can exhibit a wide range of morphological properties, from extremely small pores below 5 nm (microporous) to micrometric pores (macroporous). Moreover, its high surface area (microporous silicon ~1000 m2 /cm3 , mesoporous silicon ~100 m2 /cm3 , macroporous silicon ~1 m2 /cm3 ) makes it suitable to host one or more guest materials which usually results in a drastic change of its physical properties [2]. For instance, the addition of a well-chosen metal can provide specific composite structures with new electrical, optical, magnetic, plasmonic, or other properties [3]. In particular, the introduction of Cu improves the optoelectronic functions of PSi, e.g., its photoluminescence and electroluminescence [4]. The metal can be incorporated by electrochemical or electroless methods, but also by physical evaporation. However, upon evaporation or sputtering, metal atoms remain localized at the entrance of pores [3]. By contrast, immersion methods enhance the penetration of metal atoms deeper into the porous layer [5,6]. Importantly, PSi can reduce many metal ions down to their elemental state. This is due to the presence of SiHx bonds formed during the electrolytic synthesis of the material. This can in principle Int. J. Mol. Sci. 2016, 17, 876; doi:10.3390/ijms17060876

www.mdpi.com/journal/ijms

Int. J. Mol. Sci. 2016, 17, 876

2 of 9

enable the deposition of various noble and coinage metals in the porous layer by a chemical surface reaction [7–11]. In aqueous solutions, the source of electrons needed to reduce z-charged metal ions is PSi itself [12]. 4Mz` ` zSi ` 2zH2 O Ñ 4M ` zSiO2 ` 4zH`

(1)

However, the hydrophobic inner pore walls are impermeable to water contributing to concentrate metal deposits on the surface in patches rather than as a conformal layer down into the pores [7,8,12,13]. Nevertheless, this approach remains attractive and elegant, as it could provide an easy, low-cost, and versatile process to metallize PSi. Hence, the purpose of this study is to identify a liquid medium: 1. 2. 3. 4.

Capable of penetrating the diverse porous structures [14]; in which different metal precursors are soluble [15,16]; that does not prevent the reaction of these precursors with surface hydrides [17]; and in which additional reagents can be added to extend metal precursor decomposition [15,18].

We believe that ionic liquids (ILs) fulfill all these requirements. Indeed, ILs are suitable media to solubilize a wide range of metal precursors. From these solutions, zero-valent metallic nanoparticles (NPs) are easily precipitated using appropriate reducing agents such as H2 [19–22]. Moreover, ILs penetrate quite rapidly in porous materials [14]. Therefore, we investigate in this work the ability of imidazolium-based ILs containing organometallic (OM) precursors to metallize a variety of micro- to macro-porous silicon substrates. 2. Results and Discussion 2.1. Surface Chemistry of PSi Substrates Two different porous Si materials were electrochemically etched in p-doped Si wafers (resistivity of 1500 Ohm.cm): macroporous Si (MPSi) and microporous Si (µPSi). Their morphological characteristics are listed in Table 1. Note that, within the macropores of MPSi, portions of the internal surface are microporous. Samples of bulk Si (BSi) were considered as a reference throughout this study. Table 1. Samples used in this study.

Material Macroporous Si (MPSi) Microporous Si (µPSi)

Porous Layer Pore Diameter

Thickness

5 µm (+5 nm) 5 nm

20 µm 2 µm

Before use, these substrates were dried in a strong vacuum (10´5 bar, overnight) and stored in a glove box in an argon atmosphere. Diffuse reflectance infra-red Fourier transform analysis (DRIFT) was used to characterize the surface chemical groups of these various substrates (Figure 1). For all the samples, the peaks at 1100 and at 3500 cm´1 were assigned to ν(”Si-O-Si”) and ν(”Si-OH) vibrations, respectively. The peaks at 2134, 2111, 2086, and 2254 cm´1 were attributed to ν(SiSiH3 ), ν(Si2 SiH2 ), ν(Si3 SiH), and ν(O3 SiH) vibrations, respectively. The latter peaks were not detected in BSi and exhibited a much lower intensity in MPSi as compared to µPSi. This difference of surface silicon hydride concentration between the two porous materials was confirmed by temperature-programmed desorption experiments performed between 100 and 900 ˝ C. A total of 0.628 mmol¨ g´1 of H2 was released by µPSi vs. 0.055 mmol¨ g´1 of H2 for MPSi [23].

Int. J. Mol. Sci. 2016, 17, 876

3 of 9

Int. J. Mol. Sci. 2016, 17, 876

3 of 9

Int. J. Mol. Sci. 2016, 17, 876

3 of 9

Figure 1. DRIFT spectra of macroporous Si (MPSi), microporous Si (µPSi), and bulk Si (BSi) samples in

Figure DRIFT spectra of Si (MPSi), microporous Si (µPSi), and bulk Si (BSi) samples in an1.argon atmosphere at macroporous 50 °C. an argon atmosphere at 50 ˝ C. Figure 1. DRIFT spectra of macroporous Si (MPSi), microporous Si (µPSi), and bulk Si (BSi) samples in 2.2. Metallization of PSi an argon atmosphere at 50 °C.

2.2. Metallization of PSi

2.2.1. Metallization from a Suspension of Metallic NPs in IL 2.2. Metallization of PSi 2.2.1. Metallization a Suspension of Metallic NPszero-valent in IL Suspensionsfrom of well-dispersed crystalline metallic Cu-NPs of 5 nm with a narrow size 2.2.1. Metallization from a Suspension of Metallic NPs in IL distribution were routinely synthesized from the reduction of mesitylcopper 1-butyl-3Suspensions of well-dispersed crystalline metallic zero-valent Cu-NPs(CuMes) of 5 nminwith a narrow methylimidazolium bis(trifluoromethylsulphonyl)imide [C1C4Im][NTf2] [24]. These stable Cu-NP size distribution were routinely synthesized from zero-valent the reduction of ofmesitylcopper (CuMes) Suspensions of well-dispersed crystalline metallic Cu-NPs 5 nm with a narrow size in suspensions could be promising starting materials to achieve conformal copper deposition into distribution were routinely bis(trifluoromethylsulphonyl)imide synthesized from the reduction of mesitylcopper (CuMes) in 1-butyl-31-butyl-3-methylimidazolium [C1 C4 Im][NTf ] [24]. These stable 2 PSi layers. methylimidazolium bis(trifluoromethylsulphonyl)imide [C 1 C 4 Im][NTf 2 ] [24]. These stable Cu-NP Cu-NP suspensions be promising starting with materials to achieve into BSi and PSicould substrates were impregnated a suspension of 5conformal nm Cu-NPscopper in [C1Cdeposition 4Im][NTf2] suspensions could be promising starting materials to achieve conformal copper deposition into PSi layers. in an argon atmosphere for 2 h at 50 °C, an ex-situ procedure. After treatment, the samples were rinsed PSi layers. BSi were impregnated with suspension 5 nm Cu-NPs [C1 C4 Im][NTf withand CHPSi 2Cl2.substrates No Cu deposition was observed on athe BSi surfaceof(Figure 2a). Thisinindicates that 2] BSi and PSi substrates were impregnated with a suspension of 5 nm Cu-NPs in [C1C4Im][NTf2] in an surface argon atmosphere forare 2 hinefficient at 50 ˝ C, an procedure. AfterThe treatment, thewashed samplesaway werevia rinsed silanol ≡Si-OH for ex-situ the grafting of Cu-NPs. metal was in an argon atmosphere for 2 h at 50 °C, an ex-situ procedure. After treatment, the samples were rinsed rinsing2with CH2Cl 2 and subsequently dried.on By the contrast, MPSi and µPSi 2a). samples exhibited large Cu with CH waswas observed (Figure ThisThis indicates thatthat surface with2 Cl CH. 2No Cl2. Cu No deposition Cu deposition observed on BSi the surface BSi surface (Figure 2a). indicates aggregates with a broad size distribution on their surface (Figures 2b and 3a). Such a morphology silanol ”Si-OH are ≡Si-OH inefficient the grafting Cu-NPs. metalThe wasmetal washed viaaway rinsing surface silanol are for inefficient for theofgrafting of The Cu-NPs. was away washed via with could result from the agglomeration of the weakly adherent Cu-NPs during the rinsing and drying CH2 Cl and subsequently dried. By contrast, MPSi and µPSi samples exhibited large Cu aggregates rinsing with CH 2 Cl 2 and subsequently dried. By contrast, MPSi and µPSi samples exhibited large Cu 2 processes [25]. This could mean that surface hydrides are needed to graft Cu-NPs. This could also be aggregates with a broad size distribution on their surface (Figures 2b and 3a). Such a morphology with a broad size distribution on their surface (Figures 2b and 3a). Such a morphology could result related to the rougher topography of these substrates. Nevertheless, no Cu was detected in their pores result from theof agglomeration of the weakly adherent Cu-NPs during and the rinsing drying [25]. from could the agglomeration the weakly adherent Cu-NPs during the rinsing dryingand processes (Figure 3b). This can be expected in µPSi (the pores have the same size as the NPs), but is more processes [25].that Thissurface could mean that surface hydrides are needed to graft Cu-NPs. This could also beto the This could mean hydrides are needed to graft Cu-NPs. This could also be related surprising for MPSi. Finally, this procedure proves not to be efficient to metallize the internal surface related to the rougher topography of these substrates. Nevertheless, no Cu was detected in their pores rougher topography of the pores [23]. of these substrates. Nevertheless, no Cu was detected in their pores (Figure 3b). (Figure 3b). This can be expected in µPSi (the pores have the same size as the NPs), but is more This can be expected in µPSi (the pores have the same size as the NPs), but is more surprising for MPSi. surprising for MPSi. Finally, this procedure proves not to be efficient to metallize the internal surface Finally, this procedure proves not to be efficient to metallize the internal surface of the pores [23]. of the pores [23].

(a)

(b)

Figure 2. Surface of (a) BSi and (b) µPSi samples after treatment through the ex-situ procedure.

(a)

(b)

Figure 2. Surface of (a) BSi and (b) µPSi samples after treatment through the ex-situ procedure.

Figure 2. Surface of (a) BSi and (b) µPSi samples after treatment through the ex-situ procedure.

Int. J. Mol. Sci. 2016, 17, 876 Int. J. Mol. Sci. 2016, 17, 876 Int. J. Mol. Sci. 2016, 17, 876

4 of 9 4 of 9 4 of 9

(a) (a)

(b) (b)

Figure 3. sample tentatively metallized through the ex-situ Cu aggregates are Figure 3. MPSi MPSi ex-situ procedure. procedure. (a) (a) Figure 3. MPSi sample sample tentatively tentatively metallized metallized through through the the ex-situ procedure. (a) Cu Cu aggregates aggregates are are visible (b) but not in the pores. visible on on the the surface; surface; (b) (b) but but not not in in the the pores. pores. visible on the surface;

2.2.2. 2.2.2. Metallization Metallization from from aa Solution Solution of of CuMes CuMes in in IL IL 2.2.2. Metallization from a Solution of CuMes in IL In In the the first first experiment, experiment, aa MPSi MPSi substrate substrate was was placed placed in in an an autoclave autoclave and and impregnated impregnated by by aa solution solution In thein first experiment, a MPSi substrate was placed in an autoclave and impregnated by a solution of CuMes [C 1C4Im][NTf2] under argon atmosphere for 2 h at 50 °C. The autoclave was then placed of CuMes in [C1C4Im][NTf2] under argon atmosphere for 2 h at 50 °C. The autoclave was then placed of CuMes in [CH Im][NTf ] under argon atmosphere for 2 h at 50 ˝ C. The autoclave was then placed 1C under 0.9 MPa 2,4for 4 h at2 100 °C. Interestingly enough, this procedure afforded Cu islands, both at under 0.9 MPa H2, for 4 h at 100 °C. enough, this procedure afforded Cu islands, both at ˝ C.Interestingly under 0.9 MPa H2 , for 4 h at 100 Interestingly enough, this4a). procedure afforded Cu islands,analysis both at the the surface surface and and in in the the pores, pores, after after aa rinsing rinsing process process (Figure (Figure 4a). Electron Electron dispersive dispersive X-ray X-ray analysis the surface and in the pores, confirms after a rinsingCu process (Figure 4a). Electron of dispersive X-ray analysis (EDX) (EDX) (EDX) mapping mapping (Figure (Figure 4b) 4b) confirms that that Cu is is coating coating the the inner inner walls walls of the the macropores. macropores. The The highest highest mapping (Figure 4b) confirms that Cu is coating the inner walls of the macropores. The highest Cu Cu Cu concentration concentration was was detected detected at at the the pore pore walls walls near near the the cross-section cross-section plane plane (red). (red). A A closer closer concentration was detected at the pore walls near the cross-section plane (red). A closer examination examination examination of of the the morphology morphology of of the the Cu Cu deposit deposit shows shows that that it it was was formed formed of of large large (>150 (>150 nm) nm) of the morphology of the Cu deposit shows that it was formed of large (>150 nm) Cu islands on the flat Cu islands on the flat walls of the pores and smaller ones (about 12 nm) on the microporous walls Cu islands on the flat walls of the pores and smaller ones (about 12 nm) on the microporous walls walls of the pores and smaller ones (about 12 nm) on the microporous walls of macropores (Figure 4c). of of macropores macropores (Figure (Figure 4c). 4c). This This could could be be related related to to the the difference difference in in roughness roughness (rougher (rougher surfaces surfaces This could be related to the difference in roughness (rougher surfaces induce more nucleation) or to induce induce more more nucleation) nucleation) or or to to the the increased increased concentration concentration of of surface surface silicon silicon hydride hydride specifically specifically in in the micropores. increased concentration of surface silicon hydride specifically in the micropores. the the micropores.

(a) (a)

(b) (b)

(c) (c)

Figure sample metallized by decomposing CuMes in [C C Cu islands are clearly Figure 4. 4. MPSi MPSi metallized by by decomposing decomposingCuMes CuMesin in[C [C111C C444Im][NTf Im][NTf222]. ].].(a) (a) MPSi sample metallized (a)Cu Cuislands islands are are clearly clearly visible in the pores; as confirmed by (b) EDX mapping of Cu; (c) Cu is detected within the hollow visible in the pores; as confirmed by (b) EDX mapping of Cu; (c) Cu is detected within the hollow macropores. cross-section plane. macropores. The The highest highest Cu Cu concentration concentration was was detected detected at at the the pore pore walls walls near near the the cross-section cross-section plane.

In an attempt to increase metal content the MPSi layer, similar experiments were performed In an attempt attempt to to increase increasemetal metalcontent contentinin inthe theMPSi MPSilayer, layer,similar similar experiments were performed In an experiments were performed at at higher temperatures (150 and 200 °C). The back-scattered electron (BSED) scanning electron at higher temperatures (150 and 200 °C). The back-scattered electron (BSED) scanning electron ˝ higher temperatures (150 and 200 C). The back-scattered electron (BSED) scanning electron microscopy microscopy (SEM) images in 55 show that, increasing to °C, more Cu microscopy (SEM) images in Figure Figure showincreasing that, upon upontemperature increasing temperature temperature to 150 150 °C, more Cu ˝ C, more (SEM) images in Figure 5 show that, upon to 150 Cu was indeed was indeed detected throughout the porous material. At 200 °C, however, Cu was preferentially was indeed detected throughout the porous material. At 200 °C, however, Cu was preferentially ˝ detected throughout the porous material. Atto200 C, however, Cu was preferentially deposited near deposited near the blocking access pores below. The reaction was probably too fast deposited near the surface, surface, blocking access to the the pores below. The reaction was probably too fast the surface, blocking access to the pores below. The reaction was probably too fast and did not allow and did not allow for the diffusion of CuMes down to the bottom of the porous layer. and diddiffusion not allowoffor the diffusion CuMes down the bottom for the CuMes down toof the bottom of thetoporous layer.of the porous layer.

Int. J. Mol. Sci. 2016, 17, 876 Int. J. Mol. Sci. 2016, 17, 876

5 of 9 5 of 9

Int. J. Mol. Sci. 2016, 17, 876

5 of 9

Figure 5. MPSi samples metallized by decomposition of a solution of CuMes in IL at 100, 150 and Figure 5. MPSi samples metallized by decomposition of a solution of CuMes in IL at 100, 150 and Figure 5. MPSi samples metallized by decomposition of a(BSED) solution 200 °C observed using the back-scattered electron detector ofof theCuMes SEM. in IL at 100, 150 and 200 ˝ C observed using the back-scattered electron detector (BSED) of the SEM. 200 °C observed using the back-scattered electron detector (BSED) of the SEM.

Finally, MPSi is a complex material which incorporates planar surface portions as well as Finally, MPSi is material which planar portions as well Finally, These MPSi different is aa complex complex material which incorporates incorporates planar surface surface portions as groups, well as as micropores. surfaces bear different amounts of surface silanol and hybride micropores. These different surfaces bear different amounts of surface silanol and hybride groups, micropores. These different surfaces bear different amounts of surface silanol and hybride groups, which are probably involved in either the NP formation or their grafting. To separate these factors, which are either the or grafting. To separate separate these factors, which are probably probably involved incan either the NP NP formation formation or their their grafting. To two reference siliconinvolved samplesin be considered: BSi, which exhibits no porosity andthese onlyfactors, silanol two reference silicon samples can be considered: BSi, which exhibits no porosity and only silanol two reference samples be containing considered:only BSi, micropores which exhibits no porosity onlyhydride silanol surface groupssilicon (Figure 1), andcan µPSi, in which surfaceand silicon surface groups 1), micropores which silicon surface groups (Figure (Figure 1), and and µPSi, containing only micropores in which surface surface silicon hydride hydride groups dominate. Moreover, theµPSi, lattercontaining material is only best suited to the in spontaneous Cu deposition upon groups dominate. Moreover, the latter material is best suited to the spontaneous Cu deposition groups dominate. Moreover, the latter material is best suited to the spontaneous deposition upon upon immersion in solutions containing cupric ions because of its reductive character Cu [5,26–28]. immersion in cupric of immersion in solutions solutions containing cupric ions because of its its reductive character [5,26–28]. [5,26–28]. In order to verify ifcontaining this reactivity ofions µPSibecause still holds inreductive IL-based character solutions, the metallization of In order to verify if this reactivity of µPSi still holds in IL-based solutions, the metallization of In order to verify if this reactivity of µPSi still holds in IL-based solutions, the metallization of µPSi and BSi was attempted by sole impregnation with a solution of CuMes in [C1C4Im][NTf 2] under µPSi and BSi was attempted by sole impregnation with a solution of CuMes in [C C Im][NTf ] under 11C 22] closed µPSi BSi was attempted by °C. soleInterestingly impregnation with a solution of CuMes in [C Im][NTf under argonand atmosphere for 2 h at 50 enough, after rinsing with CH 2Cl 2,44an almost ˝ C. Interestingly enough, after rinsing with CH Cl , an almost closed argon atmosphere hh at 50 22, anislands argon atmosphere for atµPSi 50 °C. Interestingly aftermore rinsing with CH22ClCu almost with closeda Cu film remainedfor on22the surface (Figureenough, 6b). Even interestingly, Cu film thethe µPSi surface (Figure 6b). Even more interestingly, CuHence, islands with a is diameter Cu filmremained remained µPSi surface (Figure 6b).ofEven more interestingly, CuCuMes islands with a diameter of about on 10on nm also coated the inner surface the pores (Figure 6c). indeed of about 10 nm also coated the inner surface of the pores (Figure 6c). Hence, CuMes is indeed diameter of about 10 nm the inner of theremained pores (Figure Hence,ofCuMes is readily indeed readily decomposed by also µPSi.coated In contrast, nosurface Cu deposit at the6c). surface BSi (Figure 6a). decomposed by µPSi. In contrast, no Cu deposit remained at the surface of BSi (Figure 6a). This readily decomposed by µPSi. In contrast, no Cu deposit remained at the surface of BSi (Figure 6a). This rules out the oxidation of Si as the second half-reaction responsible for metal deposition in water. rules out the of indicate Si second half-reaction responsible for deposition in water. water. This rules out oxidation the oxidation ofas Si the as the second half-reaction responsible formetal metalthe deposition in Furthermore, these results that the surface hydride groups decompose Cu precursor and Furthermore, these results indicate that the surface hydride groups decompose the Cu precursor Furthermore, these results indicate that the surface hydride decompose the Cu precursor and anchor the Cu-NPs. This would explain why different sizesgroups and surface coverage of Cu islandsand are anchor the Cu-NPs. This would explain why different sizes and surface coverage of Cu islands are anchor the Cu-NPs. This would explain why different sizes and surface coverage of Cu islands are observed in the flat vs. microporous internal surfaces of MPSi. In conclusion, this easy procedure observed the flat In conclusion, this easy observed in the flat vs. vs. microporous microporous internal surfaces ofSiMPSi. MPSi. In conclusion, this easy procedure procedure allows forin the metallization by Cu of ainternal variety surfaces of porousof samples without H2 [23]. allows allows for for the the metallization metallization by by Cu Cu of of aa variety variety of of porous porous Si Si samples samples without without H H22 [23]. [23].

(a) (a)

(b) (b)

(c) (c)

Figure 6. (a) Top-view SEM image of BSi surface; (b) top-view and (c) cross section SEM images (using Figure 6. SEM image ofofBSi surface; (b) top-view and (c) cross section SEM images (using backscattering diffusion detector) sample aftertop-view impregnation a solution of CuMes in Figure 6. (a) (a)Top-view Top-view SEM image ofµPSi BSi surface; (b) and (c)by cross section SEM images backscattering diffusion detector) of µPSi sample after impregnation by a solution of CuMes in [C1C4Im][NTf 2] under argon atmosphere for 2 h at 50 °C. (using backscattering diffusion detector) of µPSi sample after impregnation by a solution of CuMes in [C 1C4Im][NTf2] under argon atmosphere for 2 h at 50 °C. [C C Im][NTf ] under argon atmosphere for 2 h at 50 ˝ C. 1 4

2

2.3. Generalization to Other Metals 2.3. Generalization to Other Metals 2.3. Generalization to Other As highlighted in theMetals introduction, this procedure should be suited to deposit other metals as As highlighted in the introduction, this procedure should be suited to deposit other metals as well.As Tohighlighted explore this in possibility, this process appliedshould to solutions of other OM precursors in IL. the introduction, this was procedure be suited to deposit other metals as well. To explore this possibility, this process was applied to solutions of other OM precursors in IL. well. To explore this possibility, this process was applied to solutions of other OM precursors in IL. 2.3.1. Palladium 2.3.1. Palladium BSi and µPSi samples were treated with a solution of Pd(dba)2 in [C1C4Im][NTf2] for 2 h at 50 °C BSi and µPSi samplesAs were treatedno with solution of observed Pd(dba)2 in 1CBSi 4Im][NTf 2] for 2 h at7a). 50 °C under argon atmosphere. expected, Pd adeposit was on[C the surface (Figure By under argon atmosphere. As expected, no Pd deposit was observed on the BSi surface (Figure 7a). By

Int. J. Mol. Sci. 2016, 17, 876

6 of 9

2.3.1. Palladium BSi and µPSi samples were treated with a solution of Pd(dba)2 in [C1 C4 Im][NTf2 ] for 2 h at 50 ˝ C 6 of 9 As expected, no Pd deposit was observed on the BSi surface (Figure67a). of 9 By contrast, a population of large islands was present thesurface surfaceofofµPSi µPSi(Figure (Figure7b) 7b)and andaadense dense contrast, a population of large PdPd islands was present atat the contrast, a population of large Pd islands in was present at the(Figure surface7c). of µPSi (Figure 7b) and asimilar dense population of Pd-NPs (12 nm in diameter) the micropores These results are thus population of Pd-NPs (12 nm in diameter) in the micropores (Figure 7c). These results are thus similar population of Pd-NPs (12 nm inindicating diameter) that in the micropores (Figure 7c). These results are thus similar to those obtained with CuMes, the same mechanisms are at play. to those obtained with CuMes, indicating that the same mechanisms are at play. to those obtained with CuMes, indicating that the same mechanisms are at play. Int. J. Mol. Sci. 2016, 17, 876 under argon atmosphere. Int. J. Mol. Sci. 2016, 17, 876

(a) (a)

(b) (b)

(c) (c)

Figure 7. SEM images after reaction with Pd(dba)2 in [C1C 4Im][NTf2] at 50 °C ˝ Cunder Figure Im][NTf22]] at at 50 50 °C underargon argonatmosphere atmosphere Figure 7. 7. SEM SEM images images after after reaction reactionwith withPd(dba) Pd(dba)22in in[C [C11C C44Im][NTf under argon atmosphere for 2 h. (a) BSi surface; µPSi (b) surface and (c) micropores (cross section). for 2 h. (a) BSi surface; µPSi (b) surface and (c) micropores (cross section). for 2 h. (a) BSi surface; µPSi (b) surface and (c) micropores (cross section).

2.3.2. Platinum 2.3.2. 2.3.2. Platinum Platinum With Pt(dba)2 as an OM precursor, no Pt deposit formed on the (BSi) surface (Figure 8a), whereas With an OM OM precursor, precursor, no (BSi) surface surface (Figure (Figure 8a), With Pt(dba) Pt(dba)22 as as an no Pt Pt deposit deposit formed formed on on the the (BSi) 8a), whereas whereas populations of large Pt islands and small particles were observed on the µPSi surface and in the populations of large Pt islands and small particles were observed on the µPSi surface and populations of large Pt islands and small particles were observed on the µPSi surface and in in the the micropores, respectively (Figures 8b,c). The EDX analysis confirms the presence of Pt with some micropores, respectively (Figure 8b,c). The EDX analysis confirms the presence of Pt with some traces micropores, respectively (Figures 8b,c). The EDX analysis confirms the presence of Pt with some traces of ILs. Note that the overall amount of deposited Pt is weaker than for Pd and Cu. This could of ILs. of Note overall amount of deposited Pt is Pt weaker thanthan for Pd Cu. Cu. ThisThis could be traces ILs. that Notethe that the overall amount of deposited is weaker forand Pd and could be ascribed to (i) the larger size of Pt particles on the surface (≈400 nm vs. ≈40 nm for Cu and Pd) and ascribed to (i) the larger size of Pt particles on the surface («400 nm vs. «40 nm for Cu and Pd) and be ascribed to (i) the larger size of Pt particles on the surface (≈400 nm vs. ≈40 nm for Cu and Pd) and (ii) a partial leaching of the Pt-NPs during the rinsing process (the anhydrous CH2Cl2 went from (ii) (ii) aa partial partial leaching leaching of of the the Pt-NPs Pt-NPs during during the the rinsing rinsing process process (the (theanhydrous anhydrous CH CH22Cl Cl22 went went from from colorless to red amid rinsing). colorless to red amid rinsing). colorless to red amid rinsing).

(a) (a)

(b) (b)

(c) (c)

Figure 8. SEM images after reaction with Pt(dba)2 in [C1C4Im][NTf2] at 50 °C under argon atmosphere Figure 8. SEM images after reaction with Pt(dba)2 in [C1C4Im][NTf2] at 50 °C ˝ under argon atmosphere for 2 h. (a) BSi surface; µPSireaction (b) surface and (c) micropores (cross section). Figure 8. SEM images after with Pt(dba) 2 in [C1 C4 Im][NTf 2 ] at 50 C under argon atmosphere for 2 h. (a) BSi surface; µPSi (b) surface and (c) micropores (cross section). for 2 h. (a) BSi surface; µPSi (b) surface and (c) micropores (cross section).

To increase Pt content in µPSi, the more reactive (COD)Pt(Me)2 was tested. With this precursor, To increase Pt content in µPSi, the more reactive (COD)Pt(Me)2 was tested. With this precursor, the density and size of Pt-NPs werethe similar those of Cu and Pd (Figure 9). However, metal To increase content in µPSi, more to reactive tested. With this the precursor, the density and Pt size of Pt-NPs were similar to those(COD)Pt(Me) of Cu and Pd2 was (Figure 9). However, the metal concentration decreased with depth. As explained with high-temperature experiments with CuMes the density and size of Pt-NPs were similar to those of high-temperature Cu and Pd (Figureexperiments 9). However, theCuMes metal concentration decreased with depth. As explained with with (see Figure 5), this can be with ascribed to an overly fast decomposition of (COD)Pt(Me) 2, as compared to concentration decreased depth. As explained with high-temperature experiments with CuMes (see Figure 5), this can be ascribed to an overly fast decomposition of (COD)Pt(Me)2, as compared to the rate of penetration the precursor the fast pores. (see Figure 5), this can of be to aninto overly decomposition of (COD)Pt(Me)2 , as compared to the rate of penetration of ascribed the precursor into the pores. the rate of penetration of the precursor into the pores.

Int. J. Mol. Sci. 2016, 17, 876

7 of 9

J. Mol. 2016, 876 Int.Int. J. Mol. Sci.Sci. 2016, 17,17, 876

7 of79of 9

(a) (a)

(b)

(c)

(b)

(c)

Figure 9. SEM images of (a) the surface (b) the cross-section of µPSi after reaction with Pt(Me)2(COD) Figure 9. SEM SEM images images of of (a) (a) the the surface surface (b) (b) the the cross-section cross-section of of µPSi µPSi after (COD) Figure after reaction reaction with with Pt(Me) Pt(Me)22(COD) C4Im][NTf 2] at 50˝°C under argon atmosphere for 2 h; (c) EDX mapping of Pt in the cross section. in [C19. in [C C Im][NTf ] at 50 C under argon atmosphere for 2 h; (c) EDX mapping of Pt in the cross section. in [C11C44Im][NTf22] at 50 °C under argon atmosphere for 2 h; (c) EDX mapping of Pt in the cross section.

2.3.4. Ruthenium 2.3.3. Ruthenium 2.3.4. Ruthenium With (COD)Ru(2-methylallyl)2, it was found that rinsing with CH2Cl2 resulted in a dark solution. With rinsing with CH Cl22 resulted resulted in aa dark dark solution. 22,, it With (COD)Ru(2-methylallyl) (COD)Ru(2-methylallyl) it was was found that rinsing withand CH22properly Cl in solution. Several successive rinses were needed to found obtainthat a clear solution remove residual IL Several successive rinses were needed to obtain a clear solution and properly remove residual IL from Several successive wereinneeded to obtain a clear solution and(RuNPs properly residual IL from the surface. rinses Therefore, this case, significant leaching of Ru or remove starting material) the surface. Therefore, in this case, significant leaching of (RuNPs starting material) occurred. occurred. As a result, only sparse particles remained onleaching theRu µPSi surface (Figure 10a). Nevertheless, from the surface. Therefore, in this case, significant of Ru or (RuNPs or starting material) As a result,were only sparse particles onremained the µPSi surface (Figure 10a). Nevertheless, RuNPs were RuNPs present in theremained pores, as confirmed byon SEM EDX mapping (Figure occurred. As a still result, only sparse particles theand µPSi surface (Figure 10a).10b). Nevertheless, still present thepresent pores, as by confirmed SEM and EDX mapping (Figure 10b). (Figure 10b). RuNPs wereinstill in confirmed the pores, as by SEM and EDX mapping

(a)

(b)

Figure 10. SEM images 2 in (a) of the surface of µPSi after reaction with (COD)Ru(2-methylallyl) (b) [C1C4Im][NTf2] at 50 °C under argon atmosphere for 2 h and after washing. (a) Left twice or right five Figure 10. images of ofand µPSi after reaction reaction with (COD)Ru(2-methylallyl) 2 in 2; (b) cross section of µPSi corresponding EDX mapping of Ru. times10. withSEM CH2Cl Figure SEM images of the the surface surface of µPSi after with (COD)Ru(2-methylallyl) 2 in

4Im][NTf2] at 50 °C [C1C ˝ under argon atmosphere for 2 h and after washing. (a) Left twice or right five [C 1 C4 Im][NTf2 ] at 50 C under argon atmosphere for 2 h and after washing. (a) Left twice or right Cl 2; (b) cross section of µPSi and corresponding EDX mapping of Ru. times withwith CH2CH 3. Materials and Methods five times 2 Cl2 ; (b) cross section of µPSi and corresponding EDX mapping of Ru.

All operations were performed in the strict absence of oxygen and water under purified argon 3. Materials and Methods 3. atmosphere Materials and Methods using glove-box (MBraun) or vacuum-line techniques. Bis(dibenzylideneacetone) All operations performed in the strict absence of oxygen 2and water under purified argon platinum, Pt(dba) 2were , Bis(dibenzylideneacetone)palladium(0) ) (Strem), (1,5-cyclooctadiene) All operations were performed in the strict absence(Pd(dba) of oxygen and water under purified dimethylplatine(II) (Pt(Me)2(COD)) (Nanomeps), bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium(II) atmosphere using glove-box (MBraun) or vacuum-line techniques. Bis(dibenzylideneacetone) argon atmosphere using glove-box (MBraun) or vacuum-line techniques. Bis(dibenzylideneacetone) (Ru(2-methylallyl) (COD)) (Sigma Aldrich, St. Louis, MO,(Pd(dba) USA) 2and mesitylcopper (CuMes) platinum, Pt(dba)2, 2Bis(dibenzylideneacetone)palladium(0) ) (Strem), (1,5-cyclooctadiene) platinum, Pt(dba)2 , Bis(dibenzylideneacetone)palladium(0) (Pd(dba)2 ) (Strem), (1,5-cyclooctadiene) (Nanomeps) were (Pt(Me) kept in 2a(COD)) refrigerator in the glove box and used as received. dimethylplatine(II) (Nanomeps), bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium(II) dimethylplatine(II) (Pt(Me)2 (COD)) (Nanomeps), bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium(II) 1-Butyl-3-methylimidazoliumbis(trifluoromethylsulphonyl)imide, 1C4Im][NTf 2], was prepared (Ru(2-methylallyl) 2(COD)) (Sigma Aldrich, St. Louis, MO, USA) [C and mesitylcopper (CuMes) (Ru(2-methylallyl)2 (COD)) (Sigma Aldrich, St. Louis, MO, USA) and mesitylcopper (CuMes) as already reported [29]. Its purity was checked via NMR spectra recorded on a Bruker Advance (Nanomeps) were kept in a refrigerator in the glove box and used as received. (Nanomeps) were kept in a for refrigerator the MHz glovefor box and used as received. 1H and at in 13C. spectrometer at 300 MHz 75.43 After purification, the halide content was 1-Butyl-3-methylimidazoliumbis(trifluoromethylsulphonyl)imide, [C1C4Im][NTf 2], was prepared 1-Butyl-3-methylimidazoliumbis(trifluoromethylsulphonyl)imide, [C1 Cof prepared 4 Im][NTf 2 ], was found to be below 100 ppm (HR-SM), and water found to be ~12 ppm (limit Karl Fischer titration). as already reported [29]. Its purity was checked via NMR spectra recorded on a Bruker Advance as already purity was checked via NMR spectra on a2] Bruker Advance The reported solutions [29]. were Its freshly prepared dissolving in recorded [C1C4Im][NTf to the desired 1H and 13C.CuMes spectrometer at 300 MHz for at 75.43byMHz for 13 After purification, the halide content was 1 spectrometer at (5 300× 10 MHz for H 75.43 tube MHzunder for stirring C. Afteratpurification, the halide content was −2 mole·L −1) and concentration in a at Schlenk room temperature. found to be below 100 ppm (HR-SM), and water found to be ~12 ppm (limit of Karl Fischer titration). found to be below 100 ppm (HR-SM), and water foundreported to be ~12 ppm of Karl Fischerby titration). Suspensions of Cu-NPs were prepared as already [24]. The(limit NPs were observed TEM The solutions were freshly prepared by dissolving CuMes in [C1C4Im][NTf2] to the desired Thea solutions wereatfreshly prepared by dissolving CuMes inthe [Csuspensions the desired using Philips CM120 120 kV. For this purpose, in the glove box, 1 C4 Im][NTfwere 2 ] to deposited −1 concentration (5 × 10−2´mole·L tube under stirring at room temperature. 2 mole¨ L)´in 1 )ainSchlenk concentration (5 and ˆ 10transferred Schlenk tube underfurther stirringpreparation. at room temperature. on a TEM grid into thea microscope without For each suspension, Suspensions of Cu-NPs were prepared as already reported [24]. The NPs were observed by TEM theSuspensions size of at least NPs was measured. size distribution was The thenNPs fittedwere by aobserved lognormalby law. of200 Cu-NPs were preparedTheir as already reported [24]. TEM using a Philips CM120 at 120 kV. For this purpose, in the glove box, the suspensions were deposited using a Philips CM120 at 120 kV. For this purpose, in the glove box, the suspensions were deposited on a TEM grid and transferred into the microscope without further preparation. For each suspension, on a TEM grid and transferred into the microscope without further preparation. For each suspension, the size of at least 200 NPs was measured. Their size distribution was then fitted by a lognormal law. the size of at least 200 NPs was measured. Their size distribution was then fitted by a lognormal law.

Int. J. Mol. Sci. 2016, 17, 876

8 of 9

Different types of Si surfaces were used—two porous and one non-porous (bulk) surface as a reference. The porous layers were electrochemically generated in a hydrofluoric solution on 200-mm Si wafers using an automated piece of equipment composed of a cleaning and an anodizing chamber. 4. Conclusions An original process for the metallization of porous Si (PSi) materials was investigated here. Ionic liquids were shown to be efficient in impregnating PSi layers and to convey organometallic (OM) precursors at the bottom of the pores. Interestingly enough, microporous Si (µPSi), which is known to reduce metal ions in aqueous solutions, is able to decompose OM precursors in ILs as well. However, this reaction most probably involves the surface hydride groups, in contrast with aqueous media (in which Si is oxidized). In addition, these groups contribute to the grafting of the metal. The reduction of CuMes by these surface hydride groups is still under investigation. However, it has been successfully generalized to other metals such as Pt, Pd, and Ru. Finally, this approach does not require additional reducing agents; hence, the safety and the cost of the process are improved. Acknowledgments: Walid Darwich thanks the French Region Rhône-Alpes through the ARC6 Program for the Ph.D. grant. Author Contributions: Catherine C. Santini and Paul-Henri Haumesser conceived and designed the experiments; Walid Darwich performed the experiments and analyzed the data; Frédéric Gaillard contributed silicon material and silicon chemistry expertise; Walid Darwich, Paul-Henri Haumesser and Catherine C. Santini wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations IL NP [C1 C4 Im][NTf2 ] SEM EDX

ionic liquids nanoparticle 1-butyl-3-methylimidazolium bis(trifluoromethylsulphonyl)imide scanning electron microscope energy-dispersive X-ray spectroscopy

References 1. 2. 3. 4.

5.

6.

7. 8.

9.

Buriak, J.M. Organometallic chemistry on silicon surfaces: Formation of functional monolayers bound through Si-C bonds. Chem. Commun. 1999, 12, 1051–1060. [CrossRef] Granitzer, P.; Rumpf, K. Porous silicon—A versatile host material. Materials 2010, 3, 943–998. [CrossRef] Herino, R. Nanocomposite materials from porous silicon. Mater. Sci. Eng. B Solid State Mater. Adv. Technol. 2000, 69, 70–76. [CrossRef] De la Mora, M.B.; del Río, J.A.; Nava, R.; Tagüeña-Martínez, J.; Reyes-Esqueda, J.A.; Kavokin, A.; Faubert, J.; Lugo, J.E. Anomalous patterned scattering spectra of one-dimensional porous silicon photonic crystals. Opt. Express 2000, 18, 22808–22816. [CrossRef] [PubMed] Bandarenka, H.; Prischepa, S.; Fittipaldi, R.; Vecchione, A.; Nenzi, P.; Balucani, M.; Bondarenko, V. Comparative study of initial stages of copper immersion deposition on bulk and porous silicon. Nanoscale Res. Lett. 2013, 8, 1–8. [CrossRef] [PubMed] Bandarenka, H.; Redko, S.; Smirnov, A.; Panarin, A.; Terekhov, S.; Nenzi, P.; Balucani, M.; Bondarenko, V. Nanostructures formed by displacement of porous silicon with copper: From nanoparticles to porous membranes. Nanoscale Res. Lett. 2012, 7, 1–10. [CrossRef] [PubMed] Tsuboi, T.; Sakka, T.; Ogata, Y.H. Metal deposition into a porous silicon layer by immersion plating: Influence of halogen ions. J. Appl. Phys. 1998, 83, 4501–4506. [CrossRef] Parbukov, A.N.; Beklemyshev, V.I.; Gontar, V.M.; Makhonin, I.I.; Gavrilov, S.A.; Bayliss, S.C. The production of a novel stain-etched porous silicon, metallization of the porous surface and application in hydrocarbon sensors. Mater. Sci. Eng. C Biomim. Supramol. Syst. 2001, 15, 121–123. [CrossRef] Harraz, F.A.; Tsuboi, T.; Sasano, J.; Sakka, T.; Ogata, Y.H. Metal, deposition onto a porous silicon layer by immersion plating from aqueous and nonaqueous solutions. J. Electrochem. Soc. 2002, 149, C456–C463. [CrossRef]

Int. J. Mol. Sci. 2016, 17, 876

10. 11. 12. 13. 14. 15. 16.

17.

18. 19.

20.

21.

22.

23. 24. 25. 26.

27. 28. 29.

9 of 9

Ogata, Y.H.; Kobayashi, K.; Motoyama, M. Electrochemical metal deposition on silicon. Curr. Opin. Solid State Mater. Sci. 2006, 10, 163–172. [CrossRef] Fukami, K.; Kobayashi, K.; Matsumoto, T.; Kawamura, Y.L.; Sakka, T.; Ogata, Y.H. Electrodeposition of noble metals into ordered macropores in p-type silicon. J. Electrochem. Soc. 2008, 155, 443–448. [CrossRef] Morinaga, H.; Suyama, M.; Ohmi, T. Mechanism of metallic particle growth and metal-induced pitting on Si wafer surface in wet chemical-processing. J. Electrochem. Soc. 1994, 141, 2834–2841. [CrossRef] Aylett, B.J.; Harding, I.S.; Earwaker, L.G.; Forcey, K.; Giaddui, T. Metallisation of porous silicon by chemical vapour infiltration and deposition. Thin Solid Films 1996, 276, 253–256. [CrossRef] Liu, H.; Jiang, L. Wettability by ionic liquids. Small 2016, 12, 9–15. [CrossRef] [PubMed] Dupont, J.; Scholten, J.D. On the structural and surface properties of transition-metal nanoparticles in ionic liquids. Chem. Soc. Rev. 2010, 39, 1780–1804. [CrossRef] [PubMed] Campbell, P.S.; Podgorsek, A.; Gutel, T.; Santini, C.C.; Padua, A.A.H.; Costa, G.M.F.; Bayard, F.; Fenet, B.; Chauvin, Y. How do physical-chemical parameters influence the catalytic hydrogenation of 1,3-cyclohexadiene in ionic liquids? J. Phys. Chem. B 2010, 114, 8156–8165. [CrossRef] [PubMed] Helgadottir, I.; Freychet, G.; Arquilliere, P.; Maret, M.; Gergaud, P.; Haumesser, P.H.; Santini, C.C. Ru-core/Cu-shell bimetallic nanoparticles with controlled size formed in one-pot synthesis. Nanoscale 2014, 6, 14856–14862. [CrossRef] [PubMed] Campbell, P.S.; Prechtl, M.H.G.; Santini, C.C.; Haumesser, P.H. Ruthenium nanoparticles in ionic liquids—A saga. Curr. Org. Chem. 2013, 17, 414–429. [CrossRef] Fonseca, G.S.; Umpierre, A.P.; Fichtner, P.F.P.; Teixeira, S.R.; Dupont, J. The use of imidazolium ionic liquids for the formation and stabilization of Ir0 and Rh0 nanoparticles: Efficient catalysts for the hydrogenation of arenes. Chem. Eur. J. 2003, 9, 3263–3269. [CrossRef] [PubMed] Prechtl, M.H.G.; Campbell, P.S.; Scholten, J.D.; Fraser, G.B.; Machado, G.; Santini, C.C.; Dupont, J.; Chauvin, Y. Imidazolium ionic liquids as promoters and stabilising agents for the preparation of metal(0) nanoparticles by reduction and decomposition of organometallic complexes. Nanoscale 2010, 2, 2601–2606. [CrossRef] [PubMed] Dupont, J.; Meneghetti, M.R. On the stabilization and surface properties of soluble transition-metal nanoparticles in non-functionalized imidazolium-based ionic liquids. Curr. Opin. Colloid Interface Sci. 2013, 18, 54–60. [CrossRef] Podgorsek, A.; Pensado, A.S.; Santini, C.C.; Gomes, M.F.C.; Padua, A.A.H. Interaction energies of ionic liquids with metallic nanoparticles: Solvation and stabilization effects. J. Phys. Chem. C 2013, 117, 3537–3547. [CrossRef] Darwich, W. Thin Metallic Films Deposited from a Suspension of Nanoparticles in Ionic Liquids; Univeristé Claude Bernard Lyon 1: Lyon, France, 2005. Arquillière, P.P.; Haumesser, P.H.; Santini, C.C. Copper nanoparticles generated in situ in imidazolium based ionic liquids. Microelectron. Eng. 2012, 92, 149–151. [CrossRef] Kolasinski, K.W.; Barclay, W.B.; Sun, Y.; Aindow, M. The stoichiometry of metal assisted etching (MAE) of Si in V2 O5 + HF and HOOH + HF solutions. Electrochim. Acta 2015, 158, 219–228. [CrossRef] Sham, T.K.; Coulthard, I.; Lorimer, J.W.; Hiraya, A.; Watanabe, M. Reductive deposition of Cu on porous silicon from aqueous solutions: An X-ray absorion study at the Cu L3,2 Edge. Chem. Mater. 1994, 6, 2085–2091. [CrossRef] Coulthard, I.; Sham, T.K. Morphology of porous silicon layers: Image of active sites from reductive deposition of copper onto the surface. Appl. Surf. Sci. 1998, 126, 287–291. [CrossRef] Kumar, P.; Huber, P. Nucleation and growth of copper on mesoporous silicon by immersion plating. J. Phys. D Appl. Phys. 2007, 40, 2864–2869. [CrossRef] Crowhurst, L.; Lancaster, N.L.; Perez-Arlandis, J.M.; Welton, T. Manipulating solute nucleophilicity with room temperature ionic liquids. J. Am. Chem. Soc. 2004, 126, 11549–11555. [CrossRef] [PubMed] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).