In Situ Monitoring of Pt Nanoparticle Formation in Ethylene Glycol

Aug 14, 2012 - Telephone: ++49 (0)381 1281 319. ... For a NaOH to Pt ratio of 2, that means in acidic medium, particle formation should be dominated b...
81 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/Langmuir

In Situ Monitoring of Pt Nanoparticle Formation in Ethylene Glycol Solution by SAXSInfluence of the NaOH to Pt Ratio Norbert Steinfeldt* Leibniz-Institut für Katalyse an der Universität Rostock, Albert-Einstein-Strasse 29a, 18059 Rostock, Germany S Supporting Information *

ABSTRACT: The temporal evolution of Pt nanoparticle formation in ethylene glycol solution from H2PtCl6·6H2O at 90 °C for different molar ratios of NaOH to Pt (84, 6.5, and 2) in the presence or absence of poly(N-vinyl-2-pyrrolidone) (PVP) as protecting agent was followed in situ by small-angle X-ray scattering (SAXS). The SAXS profiles were analyzed regarding particle size and size distribution using the Guinier approximation and the indirect Fourier transform technique (IFT). The NaOH to Pt ratio has an influence on the integral nanoparticle formation rate as well as on the metal reduction rate and the ratio of nucleation to growth reactions. The fastest nanoparticle formation rate was observed for the NaOH/Pt ratio of 6.5. The obtained results indicate that the differences in the particle formation rate might be due to differences in the reduction rate of the formed Pt complexes. In alkaline reaction media (NaOH/Pt = 84 or 6.5), small nanoparticles with a relatively narrow size distribution were formed. Therefore, it is assumed that for these NaOH/Pt ratios the particle formation is dominated by nucleation reactions. Additionally, the in situ studies point out that nanoparticles prepared at the NaOH/Pt ratio of 84 do not grow further after attaining a certain particle size. For a NaOH to Pt ratio of 2, that means in acidic medium, particle formation should be dominated by growing processes and, therefore, larger particles are formed accompanied by a broader particle size distribution. The influence of PVP on the nanoparticle formation rate is relatively low. However, in acidic medium, the presence of PVP is necessary in order to protect the formed nanoparticles from irreversible aggregation reactions.

1. INTRODUCTION Nanoparticles are of considerable interest in many areas of chemistry, physics, and material science due to their extraordinary chemical and physical characteristics, which differ strongly from those of the corresponding bulk materials.1 Their electronic, optical, structural, and catalytic qualities are affected by their size, shape, and size distribution.2 A popular method of preparing catalytic active metal nanostructures in the sub-10nm size range with narrow size distribution is the polyol process.3−5 In this process, the polyol serves both as solvent and as reducing agent. Small transition metal and especially Pt nanoparticles are particularly interesting for catalytic applications, e.g. for fuel cells,6 in hydrogenation reactions,7,8 and, more generally, for synthesis of fine chemicals.9 With the polyol method, Pt nanoparticles between 1 and 7 nm are synthesized.10−13 By adding ionic species to the polyol solution, the morphology of the formed Pt nanoparticles can also be influenced.14−16 Furthermore, the polyol method can be used to synthesize Pt containing bimetallic nanoparticles.17 It was observed that synthesis parameters such as temperature, solvent, pH, or the concentration of protecting agents can affect the particle size of the formed nanoparticles. Mechanistic insight of nanoparticle formation in ethylene glycol is still in progress, and results obtained mainly derive from applying the UV/vis and TEM method or a combination of both.18−21 UV/vis was used to follow the reduction of the © 2012 American Chemical Society

metal complex, and TEM was used to study the formed nanoparticles by applying a post-mortem approach. A method which allows the characterization of nanoparticles noninvasively and directly inside the colloidal solution without any pretreatment is small-angle X-ray scattering (SAXS).22,23 It was already demonstrated that by using this method useful information about catalytically active Ru,24 Pt, and Rh25 nanoparticles prepared with the polyol method can be derived. However, SAXS not only provides information about the particle size and the size distribution of the nanoparticles in the synthesis solution. The technique can also be employed to follow, e.g., the temporal evolution of nanoparticle formation26,27 or the loss of electrochemical active surface area of a cathode electrocatalyst.28 Furthermore, it allows information to be gathered about the spatial distribution of noble metal clusters in colloidal dispersions.29 The aim of this work is to elucidate the influence of the molar NaOH to Pt ratio on Pt nanoparticle formation in ethylene glycol. For that reason, formation of nanoparticles was followed for different NaOH to Pt ratios in situ using SAXS. From the temporal evolution of the SAXS profiles, information about the influence of the chemical environment on the metal Received: June 29, 2012 Revised: August 6, 2012 Published: August 14, 2012 13072

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

reduction, nucleation, and growth reactions could be derived. The results are considered to be helpful for the development of reproducible processes for the synthesis of Pt and, more general, catalytic active noble metal nanoparticles with tunable particle size and narrow size distribution using the polyol process.

I(q) =

2.1. Materials. H2PtCl6·6H2O (Sigma-Aldrich), anhydrous ethylene glycol 99.8% (Sigma-Aldrich), sodium hydroxide 98.5% (Acros Organics), and poly-N-vinylpyrrolidon (PVP, mol weight = 45 000) were used without further purification. 2.2. SAXS Measurements. The SAXS measurements were carried out with a Kratky-type instrument (SAXSess, Anton Paar, Austria) operated at 40 kV and 50 mA in slit collimation, using a twodimensional CCD detector (T = −40 °C). The sample to detector distance was 0.309 m. A Göbel mirror was used to convert the divergent polychromatic X-ray beam into a collimated line-shaped beam of Cu Kα radiation (λ = 0.154 nm). Slit collimation of the primary beam was applied in order to increase the flux and to improve the signal quality. The liquid sample cell consisted of a quartz capillary (internal diameter: 1 mm), stacked in a metal body with two windows for the X-ray beam. 2.3. Monitoring of Pt Nanoparticle Formation by SAXS. The prepared sample mixtures (see Table 1) were filled in the small,

⎛ 3[sin(qR ) − (qR ) cos(qR )] ⎞2 ⎟ P(q , R ) = ⎜ (qR )3 ⎝ ⎠

0.453 0.453 0.035 0.035 0.01 0.01 a

0.026 0.026 0.026

NaOH/Pt 84 84 6.5 6.5 2 2

(2)

and

m(R ) =

4π 3 R Δρ 3

(3)

where Δρ is the particle contrast (difference between the scattering length of a particle and that of the matrix). At slit collimation experiments, a smeared difference function is determined by subtracting the scattering curve of the matrix (solution without particles) from that of the nanoparticle containing solution. Indirect Fourier transformation (IFT, a short explanation of the method is given in the SI) was used to solve the integral equation (eq 1) with respect to the volume distribution function Dv(R) (program GIFT implemented in the PCG software package, Version 2.02.05). The program takes into account the instrumental smearing and fits the experimental (smeared) scattering curves directly. The desmeared scattering curves, obtained with GIFT, were further used to calculate the invariant Q using the following equation

Table 1. Composition of the Ethylene Glycol Solutions for Synthesis of Pt Nanoparticles (c[Pt] = 5.4 × 10−3 mol/L) c(PVP)a (mol/L)

(1)

R is the particle radius. Dn(R) is the particle size distribution function that denotes the number of particles of size R, m(R) is the integral over the excess scattering length density, and P(q,R) is the normalized scattered intensity of a single particle of size R, also called the shape factor. The scattering vector q is defined in terms of scattering angle θ and the wavelength λ of the irradiation [q = 4πλ−1·sin (θ/2)]. The volume weighted distribution function DvR (DvR = DnR·R3) was evaluated assuming a polydisperse systems of spherical particles, in which case R is the radius of a sphere. For a spherical particle of size R, the shape factor P(q,R) is given as

2. EXPERIMENTAL SECTION

c(NaOH) (mol/L)

∫ Dn(R) m2(R) P(q , R) dR

PVP/Pt 4.8 4.8

Q=

4.8

∫0

∞

I(q)q2 dq

(4)

The calculation of Porod invariante Q involves refinements via extrapolations of the intensity toward q → 0 and q → ∞. From the Porod invariante Q and the scattering intensity I in the forward direction (q = 0), the weighted mean particle volume V of the particles inside the solution can be calculated:

Related to the repeating unit vinylpyrrolidon (M = 111.14 g/mol).

cylindrical quartz capillary (internal volume: 40 μL) of the SAXSess sample cell at room temperature and stacked in the sample holder of the SAXS equipment. The solutions were prepared by adding corresponding amounts of metal salt, NaOH, and PVP to an ethylene glycol solution. After monitoring a SAXS curve at 25 °C, the sample cell was heated up to a temperature of 90 °C ± 0.1° by means of a Peltier element. The recording of the SAXS curve started immediately after attaining 90 °C. To follow the formation of Pt nanoparticles, SAXS curves were recorded in equidistant time intervals over a period of about 20 h. Generally, the time interval for monitoring of a single SAXS curve was 10 min (10 s detection time, 60 repetitions). The scattering curves of the Pt nanoparticles at different reaction times were obtained by subtracting the first SAXS curve measured at 90 °C from every following curve. Preliminary experiments have shown that for all experiments the pattern of the first SAXS curve measured at 90 °C corresponds to the pattern of the SAXS curve measured at 25 °C before heating of the solution (only a shift in intensity was observed). At 25 °C, Pt nanoparticle formation in ethylene glycol could not be observed within several days. 2.4. SAXS Data Analysis. The scattering of a polydispersed system is determined by the shape of the particles and by their size distribution. It is impossible to determine them both from the scattering experiment. Therefore, complementary techniques such as TEM have to be applied in order to obtain information regarding the particle shape (also see the Supporting Information (SI)). SAXS data analysis was performed in terms of the size distribution function. The total scattering intensity I(q) from a polydisperse system can be written as follows:30−32

V=

2π 2I(q = 0) Q

(5)

Because the intensity I at q = 0 cannot be determined experimentally, it was calculated for the corresponding nanoparticle population from the desmeared scattering curve by Guinier approximation.33

⎛ q 2R 2 ⎞ g ⎟ I(q) = I(0) exp⎜⎜ − ⎟ 3 ⎝ ⎠

(6)

The radius of gyration (Rg) has a model dependent relation to the shape of the particle. For spheres, for instance, it is Rg2 = 3/5RG2 (RG = radius of the sphere). The Guinier approximation was also applied to point out the temporal evolution of larger Pt structures in the solution. Since the experimental scattering profile only contains a small part of their scattering curve, the real size of these larger structures cannot be determined.

3. RESULTS 3.1. Pt Nanoparticle Formation Followed in Situ by SAXS (Experimental Results). Figure 1a and b shows the temporal evolution of the scattering profile for Pt nanoparticles synthesized with and without PVP at the NaOH/Pt ratio of 84. Scattering intensity increases asymptotically with reaction time, independent of whether the nanoparticles are protected by PVP 13073

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

particle formation, an acquisition time of 10 min turned out to be too long in order to enable time-resolved scattering curves recording. Subsequently, the first seven stored data files (one single file contained 60 successive single SAXS profiles with an exposition time of 10 s) were divided to obtain SAXS curves with acquisition times of 5, 2, or 1 min, using only 30, 12, or 6 repetitions. By applying this method, the time-dependent SAXS intensity can be followed with a time-resolution comparable to the NaOH/Pt ratio of 84 (Figure 1a and b). For reaction times later than 70 min, the scattering intensity of two subsequent measurements was constant, and mean values of 60 repetitions (acquisition time: 10 min) were used again. Figure 1e and f shows the temporal evolution of scattering profiles for the molar NaOH/Pt ratio of 2. Scattering from Pt nanoparticles was not observed for the first 300 min, independent of whether the nanoparticles were protected by PVP or not. Later on, the shape of the scattering curves for the solutions with and without PVP differs considerably. The steep decrease in scattering intensity for the solution without PVP reveals a formation of larger Pt structures which was completely suppressed when PVP was present in the solution. Figure 2 compares the temporal evolution of the Guinier radius (Rg) derived from the innermost part of the experimental

Figure 1. Temporal evolution of SAXS profiles (smeared) of Pt nanoparticles for different NaOH/Pt ratios in the absence (a, c, e) and presence (b, d, f) of PVP; NaOH/Pt = 84 (a, b); 6.5 (c, d); and 2 (e, f); (SAXSess sample cell, c[Pt] = 5.4 × 10−3 mol/L, T = 90 °C, PVP/ Pt = 5).

Figure 2. Temporal evolution of the Guinier radius (Rg) at different NaOH to Pt ratios without (a) and with (b) PVP (SAXSess sample cell, c[Pt] = 5.4 × 10−3 mol/L, T = 90 °C, PVP/Pt = 5; Rg was derived from the innermost part of the scattering curve, q < 0.5 nm−1).

or not. The increase in scattering intensity between two successive measurements is generally low, and therefore, it can be assumed that during the recording of a single SAXS curve (10 min) the size distribution is approximately constant. Scattering of Pt nanoparticles was observed for the first time after 35 min. While the slope in scattering intensity is moderate for the central part of the curve between 0.5 < q < 6 nm−1, the profiles show a steep decrease in scattering intensity at low qvalues. The moderate slope in the central part reveals the formation of small dispersed Pt nanoparticles, and the steep slope at the low q-regime (q < 0.5 nm−1) indicates the formation of larger Pt structures. However, it should be noted that the first scattering curves in the PVP free solution did not show the steep slope in the scattering intensity at low q values. Moreover, the transition in scattering intensity from the first slope for 0.5 < q < 6 to the second slope (q < 0.5) is distinctly sharper for the PVP-free sample than for the PVP containing solution. When decreasing the NaOH/Pt ratio from 84 to 6.5, the steep slope of scattering intensity at low q-regime (q < 0.5 nm−1) disappears, as shown in Figure 1c and d. Almost identical scattering profiles were obtained for solutions with and without PVP. Moreover, nanoparticle formation at this NaOH/Pt ratio proceeds significantly faster. The first nanoparticle scattering was obtained after 8 min (15 min without PVP), and identical scattering profiles were already monitored after a reaction time of about 70−80 min. Because of the fast

scattering curve for the different NaOH/Pt ratios. As already mentioned in section 2.4, this value does not reflect the true size of the larger Pt structures, but Rg can be used to follow its temporal evolution. Rg was generally lower when PVP was present. For the NaOH/Pt ratio of 84, in the absence of PVP, Rg increased continuously with time until attaining an approximately constant value after 3.5 h. The PVP containing solution formed larger Pt structures at early reaction times (40 min) already. Here, Rg is nearly constant over the whole time interval studied. For the NaOH to Pt ratio of 6.5, the differences in Rg between the particles without and with PVP protection were small. The largest differences in Rg between the solutions with and without PVP were observed for the NaOH/ Pt ratio of 2. For the solution without PVP, Rg increased strongly with time after the Pt nanoparticles were detected. With PVP, it stays nearly constant at about 2.5 nm over the whole time interval studied. 3.2. Results of SAXS Data Evaluation. In the above section, the experimental scattering profiles of the Pt nanoparticles were described qualitatively. In this section, the SAXS curves were evaluated using the IFT technique in order to determine the particle size and the size distribution of the nanoparticles at different stages of their formation. The IFT method (for a short description, see the SI) requires no 13074

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

Figure 3. Influence of time on the volume weighted size distribution of Pt nanoparticles obtained from the fit to the experimental scattering curves by the IFT-method assuming a polydisperse system of spherical particles: a, NaOH/Pt = 84; b, NaOH/Pt = 6.5; c, NaOH/Pt = 2; (c[Pt] = 5.4 × 10−3 mol/L; T = 90 °C; PVP/Pt = 5).

mathematical assumptions about the particle size distribution. However, for determination of the size distribution, the shape of the particle has to be known (see Figures S9−S12 in the SI). The effects of the NaOH to Pt ratio on the temporal evolution of the particle size distribution, obtained from the fitting procedure described above, are shown in Figure 3 for solutions containing PVP. Similar size distributions for the small Pt nanoparticle population (NaOH/Pt = 84 and 6.5) were obtained if the nanoparticles were formed in the absence of PVP (see Figures S2 and S4 in the SI). For the NaOH/Pt of 84, the maximum of the size distribution function (Rmax) of the small nanoparticle population only slightly increases with time from 0.35 nm at 47 min to 0.4 nm after 1033 min. Simultaneously, the width of the distribution function slightly increases. For the NaOH to Pt ratio of 6.5, Rmax increases from 0.3 nm at 18 min to 0.70 nm after 67 min. The width of the size distribution is, again, relatively narrow. The size distribution found after 98 min resembles that after 721 min. For the NaOH/Pt ratio of 2, the particle size distribution is significantly broader in comparison to a NaOH/Pt ratio of 84 and 6.5 Rmax shifts from about 1.25 nm at 618 min to 1.69 nm after 1033 min. In the absence of PVP, it was impossible to fit the scattering curves for this low NaOH/Pt ratio, probably because of the high polydispersity in size and shape. For a single particle, the scattering intensity I at q = 0 is independent of particle size and shape, and depends solely on the scattering length density contrast (Δρ) and the particle volume V. I(0) = (Δρ)2 V 2

Figure 4. Influence of the NaOH/Pt ratio on the temporal evolution of the mean particle radius (Rmean) (a, b) and the mean particle number (c, d) for Pt nanoparticles formed in ethylene glycol with (b, d) and without (a, c) the presence of PVP (c[Pt] = 5.4 × 10−3 mol/L; T = 90 °C; PVP/Pt = 5; the lines are draw in for clarity).

smallest particles were formed at the NaOH to Pt ratio of 84, independent of whether PVP was present or not. However, for this ratio Rmean slightly increases only during the first 400 min from 0.4 nm (0.5 nm with PVP) to about 0.6 nm. After this time, it stays constant. Contrary to Rmean, the mean number of particles [I(0)/V2] increases continuously over the whole time period. For the NaOH/Pt ratio of 6.5, the mean particle radius increases from 0.5 nm after 25 min (PVP: 0.58 nm after 15 min) to about 0.67 nm within 70 min (PVP: 0.79 nm). During this time interval, the relative number of particles strongly increases as well. After approximately 70 min, the particle size and number of particles stay constant. For the NaOH to Pt ratio of 2, the mean particle radius obtained from the first scattering curve (after 300 min) is clearly larger (1.5 nm) than those obtained for the other two NaOH to Pt ratios at the same reaction time. With increasing reaction time, both Rmean and the mean particle number [I(0)/ V2] increase over the whole time interval studied, leading to a Rmean of 2.1 nm after 1033 min. Compared to both other NaOH to Pt ratios, the number of formed particles was clearly lower.

(7)

Assuming that the solution contains N identical particles with the same volume (e.g., the mean volume), the scattering intensity I at q = 0 of this ensemble can be calculated by

I(0) = N (Δρ2 )V 2 mean

(8)

Assuming furthermore a constant scattering density inside of the formed particles (constant Δρ2), the quotient I(0)/V2 is an appropriate quantity to monitor evolution of the mean particle number vs time. Figure 4 shows the change in mean particle radius (Rmean), derived from the mean volume (Vmean), and assuming a spherical particle shape (Rmean = (3Vmean/4π)1/3; see also eq 5) together with the evolution of relative particle number for the different NaOH/Pt ratios. Both the temporal evolution of Rmean and the number of particle within the solution are strongly influenced by the NaOH to Pt ratio. Rmean derived from the particle volume (Figure 4a and b) with the assumptions described above is in good accordance with Rmax of the particle size distribution obtained from fitting the SAXS profiles (see Figure 3). The 13075

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

Table 2. Comparison of SAXS (Rmax) and TEM (dmean) Results for Pt Nanoparticles Formed in the SAXSess Sample Cell (in Situ Monitoring) and in a Small Batch Reactor (c[Pt] = 5.4 × 10−3 mol/L, PVP/Pt = 4.8) and the pH in the Batch Reactor at Two Times (See Also Figures S9−S14 in the SI) in situ

batch

1h

17 h

SAXS

SAXS

SAXS

1h TEM

SAXS

TEM

20 h

NaOH/Pt (M)

Rmax (nm)

Rmax (nm)

Rmax (nm)

dmean (nm)

Rmax (nm)

dmean (nm)

pH 0 h

pH 20 h

84 6.5 2

0.4 0.7

0.4 0.8 1.7

0.3 0.9

n.d. 2.0 ± 0.47

0.6 0.9 2.4

1.4 ± 0.43 2.1 ± 0.41 4.9 ± 0.93

12.7 11.6 1.3

12.3 4.4 0.5

nucleation and growth reactions on particle formation for the respective NaOH to Pt ratio. For the nucleation process in Pt nanoparticle formation, different mechanism were discussed. It is generally assumed that Pt(II) or Pt(IV) complexes first react with the reducing agent, forming isolated Pt(0) atoms. When a critical concentration of Pt(0) atoms is achieved, the formation of metal clusters begins by aggregation of these atoms.35 A further mechanism was suggested from theoretical studies which assumed, as the key step in nucleation, the formation of Pt− Pt bonds between Pt complexes in higher oxidation states.36 For [PtCl6]2− solved in strong alkaline ethylene glycol solution, it was observed that after heating up for 5 min at 80 °C the Pt compound undergoes reduction to [PtCl4]2− species, followed by an exchange of the ligand sphere leading to formation of [Pt(OH)4]2− species inside of 10 min.37 Assuming a similar time scale and reactions, here at 90 °C for the NaOH/Pt ratio of 84 and a low solubility of the reduced Pt species (Pt atoms) in the solution, the reduction of the Pt(II) species to Pt atoms might be considered as the rate determining step of the Pt nanoparticle formation in strongly alkaline medium at this temperature. The slow reduction rate of the [Pt(OH)4]2− was indicated by the time delay between the point of attaining 90 °C and the detection of the first scattering curve from the Pt nanoparticles (about 35 min). A decrease of the reduction rate at addition of NaOH was already observed using PtCl42− and H2 for nanoparticle formation in aqueous solution.38 The formed Pt atoms in the strongly alkaline solution mainly undergo nucleation reactions, which is indicated by the small increase of the nanoparticle size with time (0.4−0.6 nm), connected with an increase in the particle number (Figure 4). It has to be noted that during the studied time interval only a certain part of the available Pt ions are involved in nanoparticle formation. SAXS data analysis showed that the relative number of nanoparticles in the solution continuously increased over the whole time interval studied while the particle size increased only within the first 400 min (Figure 4). That means processes have to be present which hinder already formed nanoparticles from further growing by reaction with reduced Pt species (e.g., Pt atoms) once they have reached a certain particle size. This assumption is further supported by the narrow particle size distribution, which does not change with time. It is believed that either OH− ions are involved in the process which are available in high concentration in the solution39 or oxidation products of ethylene glycol21 such as glycolates are involved. Another reason could be the formation of carbonate like species on the nanoparticle surface which were observed during electrooxidation of ethylene glycol on Pt in alkaline solution.40 At a NaOH to Pt ratio of 6.5, the integral rate of nanoparticle formation rises drastically compared to the case of the ratio of

In order to prove that the assumption of the spherical particle shape was acceptable, Pt nanoparticles were additionally synthesized in a small batch reactor at identical conditions (temperature, NaOH/Pt ratio, PVP/Pt ratio). TEM images and results of the SAXS data evaluation for Pt nanoparticles formed in the small batch reactor at a reaction time of 20 h (for the NaOH/Pt ratio of 6.5 also after 1 h) are presented in the SI (Figure S9−S14). The TEM images show that the shape of the nanoparticles formed at the applied conditions is, independent of the NaOH/Pt ratio, almost spherical. The results of TEM and SAXS analysis of the Pt nanoparticles formed in the batch reactor are summarized in Table 2, which also compares them to results of SAXS data evaluation from the in situ experiment at comparable time intervals. Particle sizes determined by TEM are in good accordance with results of SAXS data analysis. Pt nanoparticles formed in the SAXSess sample cell are slightly smaller than particles formed in the batch reactor. The small differences in particle size between the in situ and the small batch experiments are attributed to the lower reaction volume (in situ sample cell, 0.04 mL; small batch, 2 mL) and the high surface to volume ratio in the SAXS sample cell. However, in both types of experiments, the smallest particles were formed at the NaOH/ Pt ratio of 84 and the largest particles were obtained for the NaOH/Pt of 2. Moreover, the same temporal behavior of the integral rate of nanoparticle formation in dependence on the NaOH/Pt ratios was observed. Therefore, the results obtained from the in situ SAXS data analysis are considered as representative for studying processes of nanoparticle formation at different NaOH ratios.

4. DISCUSSION 4.1. Influence of the NaOH to Pt Ratio on Nucleation and Growth Reactions. At 90 °C, Pt nanoparticle formation in ethylene glycol proceeds in time scales that enable the monitoring by SAXS with reasonable time resolution for all applied NaOH to Pt ratios. The presented results demonstrate that the processes leading to nanoparticle formation are strongly affected by the initial NaOH to Pt ratio. Noble metal nanoparticle formation is generally studied in terms of metal precursor reduction, nucleation, and particle growth reactions.34 Nucleation was defined as the process whereby a discrete particle of a new phase forms in a previously singlephase system. Growth was defined as a process in which additional material deposits on this particle, causing it to increase in size. Because of the detection limits of the SAXS, the nucleation process cannot be observed directly. However, the temporal evolution of the particle size and size distribution can be used to gain a deeper insight into the contribution of 13076

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

84. As shown in Table 2, nanoparticle formation starts in alkaline medium, but the medium shifted to weakly acidic during the reaction, probably because of consumption of OH− ions. The faster nanoparticle formation rate at this NaOH to Pt ratio is connected with a faster reduction rate of the metal precursor as well as faster nucleation and growth rates. Acceleration of the reduction rate is deduced from the shorter time delay between the time of achieving 90 °C and the first appearance of the particle scattering (8 min). The relatively short time interval between the appearance of the first scattering curve and the time where the scattering profile becomes constant (70 min), together with the relatively small particle size and the strong increase in particle number, indicate that nucleation must have taken place much faster compared to the case of the NaOH to Pt ratio of 84. However, the concomitant increase in particle size shows that also particle growth reactions were accelerated. The slightly larger particle size in comparison to the case of the NaOH to Pt ratio of 84 implies that the influence of growth reaction on the particle formation at this NaOH to Pt ratio is not significantly larger (Figure 4) than that for the ratio of 84. Therefore, also for this NaOH/Pt ratio, nucleation is considered as the dominant process in nanoparticle formation. The constant scattering profile after about 70 min indicates that all available Pt atoms were consumed for particle formation and that reactions between the already formed nanoparticles could only proceed to a minor degree. At a NaOH to Pt ratio of 2, nanoparticle formation completely occurred in acidic medium. For a reaction temperature of 110 °C, it was reported that in acidic medium the [PtCl6]2‑‑ precursor will be reduced slowly from Pt(IV) to Pt(II) in a first stage, followed by a second reduction of the Pt(II) species.20 Therefore, the long time delay between attaining 90 °C and the time where the first scattering curve of Pt nanoparticles was observed (300 min) should be caused by a very slow reduction of the Pt species at 90 °C. The slow reduction rate might also be the reason for the low number of particles in the reaction mixture (Figure 4). Otherwise, the mean particle radius and the width of the size distribution are clearly larger in comparison to the nanoparticles formed in alkaline medium at the same stage of reaction. An explanation for this behavior is that for the NaOH to Pt ratio of 2, or more general for an acidic medium, the influence of growth reactions on Pt nanoparticle formation is much higher than in alkaline medium. Because of a slow nucleation rate, the number of nuclei formed at a certain time interval will be low and, therefore, a larger number of reduced Pt species is available for growth reactions. Another explanation might be that the formed nuclei catalyze the reaction between the nanoparticle surface atoms and Pt species. Nuclei formed at earlier stages of the reaction have more time to grow than nuclei formed later on, and as a result, larger particles with a broader size distribution are obtained, as shown in Figures 3 and 4. The parallel increase of particle number as well as particle size with time indicates that the observed particle growth proceeds by reaction of the reduced Pt species with the nanoparticle’s surface and not by coalescence of the formed nanoparticles. This assumption is further supported by the approximately linear increase of the particle radius with time, which is an indication of particle growth by surface reaction.34 A formal scheme for Pt nanoparticle formation in ethylene glycol at different NaOH/Pt ratios is shown in Figure 5.

Figure 5. Formal scheme of Pt nanoparticle formation in ethylene glycol at 90 °C for different NaOH/Pt ratios.

The observed differences of the integral rate of nanoparticle formation are attributed to differences in the oxidation potential of ethylene glycol at different NaOH/Pt ratios and/ or to differences in the structure of the Pt complexes which has to be reduced to yield the Pt atoms. 4.2. Influence of PVP. The influence of PVP on the integral nanoparticle formation rate is relatively low compared to the effect of the NaOH/Pt ratio, as can be seen in Figures 1 and 4. For the three respective NaOH to Pt ratios studied, the size and the size distribution of the primary formed nanoparticles were comparable for synthesis solutions with and without PVP (Figures 3, 4 and Figures S2, S4 in the SI). This means that the processes which are involved in formation of this population are only slightly influenced by interactions between the reduced Pt species and PVP. It was supposed that covalent bonds between Pt precursor ions and PVP might act as preferential sites for nucleation of Pt clusters.41 The main contribution of PVP consists in protecting the primary formed Pt nanoparticles from irreversible aggregation reactions between them. The inhibition in formation of larger Pt structures by PVP might be caused by interactions between nanoparticle surfaces and the carbonyl group of the pyrrol ring,42 and/or by physically occupying the space around the nanocluster,43 leading to a steric stabilization. The steep slope in scattering intensity at low q-values for the PVP containing solution at the NaOH to Pt ratio of 84 is attributed to formation of agglomerates which probably contain small nanoparticles in a close neighborhood connected by PVP.29 Such structures can be formed, e.g., if one PVP chain contains more than one nanoparticle or if different PVP chains which contain nanoparticles are entangled. When no PVP was present, a certain part of the primary formed nanoparticles undergoes an aggregation reaction, as indicated by the temporal evolution of the Guinier radius with time (Figure 2). However, compared to the NaOH/Pt ratio of 2, the number of such larger Pt structures will be relatively low. In acidic medium, the formed nanoparticles undergo fast aggregation reactions if the nanoparticles are not protected by PVP.

5. CONCLUSIONS SAXS is a technique which allows the in situ monitoring of Pt nanoparticle formation and, more generally, noble metal formation in ethylene glycol. Evaluation of the SAXS profiles not only offers information about the influence of NaOH/Pt ratio on the integral nanoparticle formation rate, but it also allows deduction of information about the influence of this ratio on the metal reduction, nucleation, and growth reaction rates. It is believed that the differences in the integral nanoparticle formation rate between the different NaOH/Pt ratios are 13077

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

(8) Toshima, N.; Shiraishi, Y.; Teranishi, T.; Miyake, M.; Tominaga, T.; Watanabe, H.; Brijoux, W; Bönnemann, H.; Schmid, G. Various Ligand-Stabilized Metal Nanoclusters as Homogeneous and Heterogeneous Catalysts in the Liquid Phase. Appl. Organomet. Chem. 2001, 15, 178−196. (9) Campelo, J. M.; Luna, D.; Luque, R.; Marinas, J. M.; Romero, A. A. Sustainable Preparation of Supported Metal Nanoparticles and Their Applications in Catalysis. ChemSusChem 2009, 2, 18−45. (10) Koebel, M. M.; Jones, L. C.; Somorjai, G. A. Preparation of SizeTunable, Highly Monodisperse PVP-Protected Pt-Nanoparticles by Seed-Mediated Growth. J. Nanopart. Res. 2008, 10, 1063−1069. (11) Wang, Y.; Ren, J.; Deng, K.; Gui, L.; Tang, Y. Preparation of Tractable Platinum, Rhodium, and Ruthenium Nanoclusters with Small Particle Size in Organic Media. Chem. Mater. 2000, 12, 1622− 1627. (12) Park, J.; Joo, J.; Kwon, S. G; Jang, Y.; Hyeon, T. Synthesis of Monodisperse Spherical Nanocrystals. Angew. Chem., Int. Ed. 2007, 46, 4630−4660. (13) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A. High-Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction Studies of Platinum Nanoparticles in Mesoporous SBA15 Silica. J. Phys. Chem. B 2005, 109, 2192−2202. (14) Teranishi, T.; Kurita, R.; Miyake, M. J. Shape Control of Pt Nanoparticles. J. Inorg. Organomet. Polym. 2000, 10, 145−156. (15) Xia, Y.; Xiong, Y.; Lim, B.; Skrabalak, S. E. Shape-Controlled Synthesis of Metal Nanocrystals: Simple Chemistry Meets Complex Physics. Angew. Chem., Int. Ed. 2008, 48, 60−103. (16) Song, H.; Kim, F.; Connor, S.; Somorjai, G. A; Yang, P. Pt Nanocrystals: Shape Control and Langmuir Blodgett Monolayer Formation. J. Phys. Chem. B 2005, 109, 188−193. (17) Garcia-Gutierrez, D. I.; Gutierrez-Wing, C. E.; Giovanetti, L.; Ramallo-Lopez, J. M.; Requejo, F. G.; Jose-Yacaman, M. Temperature Effect on the Synthesis of Au-Pt Bimetallic Nanoparticles. J. Phys. Chem. B 2005, 109, 3813−3821. (18) Rioux, R. M.; Song, H.; Yang, P.; Somorjai, G. A. Platinum Nanoclusters' Size and Surface Structure Sensitivity of Catalytic Reactions. In Metal Nanoclusters in Catalysis and Material Sciences: The Issue of Size Control, 1st ed.; Corain, B., Schmid, G., Toshima, N., Eds.; Elsevier B.V.: Amsterdam, 2008; Chapter 7. (19) Nandanwar, S. U.; Chakraborty, M.; Mukhodadhyay, S.; Shenov, K. T. Stability of Ruthenium Nanoparticles Synthesized by Solvothermal Method. Cryst. Res. Technol. 2011, 46, 393−399. (20) Chen, J.; Herricks, T.; Geissler, M.; Xia, Y. J. Single Crystal Nanowires of Platinum can be Synthesized by Controlling the Reaction Rate of a Polyol Process. J. Am. Chem. Soc. 2004, 126, 10854−10855. (21) Bock, C.; Paquet, C.; Couillard, M.; Botton, G. A.; MacDougall, B. R. Size-Selected Synthesis of PtRu Nano-Catalysts: Reaction and Size Control Mechanism. J. Am. Chem. Soc. 2004, 126, 8028−8037. (22) Hatakeyama, Y.; Okamoto, M.; Torimoto, T.; Kuwabata, S.; Nishikawa, K. Small-Angle X-ray Scattering Study of Au Nanoparticles Dispersed in the Ionic Liquids 1-Alkyl-3-methylimidazolium Tetrafluoroborate. J. Phys. Chem. C 2009, 113, 3917−3922. (23) Svergun, D. I.; Stykova, E. V.; Dembo, A. T.; Bronstein, L. M.; Platonova, O. A.; Yakunin, A. N.; Valetsky, P. M; Khokhlov, A. R. Size Distribution of Metal Nanoparticles in Polyelectrolyte Gels. J. Chem. Phys. 1998, 109, 11109−11116. (24) Steinfeldt, N.; Sebek, M.; Jähnisch, K. Liquid Phase Hydrogenation of Methyl-N-Boc-pyrrole-2-carboxylate over Tailored Ru Nanoparticles. J. Catal. 2012, 289, 249−258. (25) Berger-Karin, C.; Sebek, M.; Pohl, M. M.; Bentrup, U.; Kondratenko, V. A.; Steinfeldt, N.; Kondratenko, E. V. Tailored Noble Metal Nanoparticles on γ-Al 2 O3 for High Temperature CH 4 Conversion to Syngas. ChemCatChem 2012, 4, 1368−1375. (26) Abecassis, B.; Testard, F.; Spalla, O.; Barboux, P. Probing in Situ the Nucleation and Growth of Gold Nanoparticles by Small-Angle Xray Scattering. Nano Lett. 2007, 7, 1723−1727. (27) Tanaka, H.; Koizumi, S.; Hashimoto, T.; Itoh, H.; Satoh, M.; Naka, K.; Chujo, Y. Combined in Situ and Time-Resolved SANS and

mainly caused by differences in the reduction rate of the Pt precursor, which at 90 °C is much faster at the NaOH/Pt of 6.5 than at a ratio of 84 and 2. The NaOH/Pt ratio also has a strong influence on the ratio of nucleation to growth reactions, leading to different sized nanoparticles depending on the pH of the synthesis solution. In alkaline reaction medium, the primary nanoparticle formation is dominated by nucleation reactions. Under these conditions, small nanoparticles with a relatively narrow size distribution are formed. In acidic medium, the influence of growth reaction on nanoparticle formation is considerably higher. This led to the formation of larger nanoparticles with a relatively broad size distribution. Here, particle growth is assumed by reaction of formed nanoparticles with reduced Pt species.

■

ASSOCIATED CONTENT

S Supporting Information *

Selected scattering profiles of Pt nanoparticles for different NaOH/Pt ratios formed in the SAXSess sample cell, results of fitting using the IFT method and the corresponding particle size distribution, description of experimental details of Pt nanoparticle formation in a small batch reactor, TEM images and SAXS profiles of Pt nanoparticles formed in the batch reactor, and results of data evaluation from TEM and SAXS. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: ++49 (0) 381 1281 319. Fax: ++49 (0)381 1281 5319. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The author would like to thank the Leibniz-Gemeinschaft and the BMBF for financial support, and Mrs. Dipl.-Ing. A.-M. Vogt for her assistance by carrying out the SAXS experiments. Sincere thanks is extended to Dr. Marga-Martina Pohl for carrying out the TEM measurements.

■

REFERENCES

(1) Corain, B.; Schmid, G.; Toshima, N. Metal Nanoclusters in Catalysis and Material SciencesThe Issue of Size Control, 1st ed.; Elsevier B.V.: Amsterdam, 2008. (2) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. The Chemistry and Properties of Nanocrystals of Different Shapes. Chem. Rev. 2005, 105, 1025−1102. (3) Bonet, F.; Delmas, V.; Grugeon, S.; Urbina, R. H.; Silvert, P.-Y.; Tekaia-Elhsissen, K. Synthesis of Monodisperse Au, Pt, Pd, Ru and Ir Nanoparticles in Ethylene Glycol. Nanostruct. Mater. 1999, 11, 1277− 1284. (4) Skrabalak, S. E.; Wiley, B. J.; Kim, M.; Formo, E. V.; Xia, Y. On the Polyol Synthesis of Silver Nanostructures: Glycolaldehyde as a Reducing Agent. Nano Lett. 2008, 8, 2077−2081. (5) Chen, Y.; Liew, K. Y.; Li, J. Size-Controlled Synthesis of Ru Nanoparticles by Ethylene Glycol Reduction. Mater. Lett. 2008, 62, 1018−1021. (6) Liu, Z.; Shamsuzzoha, M.; Ada, E. T.; Reichert, W. M.; Nikles, D. E. Synthesis and Activation of Pt Nanoparticles with Controlled Size for Fuel Cell Electrocatalysts. J. Power Sources 2007, 164, 472−480. (7) Roucoux, A.; Schulz, J.; Patin, H. Reduced Transition Metal Colloids: A Novel Family of Reusable Catalysts. Chem. Rev. 2002, 102, 3757−3778. 13078

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079

Langmuir

Article

SAXS Studies of Chemical Reactions at Specific Sites and SelfAssembling Processes of Reaction Products: Reduction of Palladium Ions in Self-Assembled Polyamidoamine Dendrimers as a Template. Macromolecules 2007, 40, 4327−4337. (28) Smith, M. C.; Gilbert, J. A.; Mawdsley, J. R.; Seifert, S.; Myers, D. J. In Situ Small-Angle X-ray Scattering Observation of Pt Catalyst Particle Growth During Potential Cycling. J. Am. Chem. Soc. 2008, 130, 8112−8113. (29) Hashimoto, T.; Saijo, K.; Harada, M.; Toshima, N. Small-Angle X-ray Scattering Analysis of Polymer-Protected Platinum, Rhodium, and Platinum/Rhodium Colloidal Dispersions. J. Chem. Phys. 1998, 109, 5627−5638. (30) Glatter, O. A New Method for the Evaluation of Small-Angle Scattering Data. J. Appl. Crystallogr. 1977, 10, 415−421. (31) Glatter, O. Evaluation of Small-Angle Scattering Data from Lamellar and Cylindrical Particles by the Indirect Transformation Method. J. Appl. Crystallogr. 1980, 13, 577−584. (32) Glatter, O.; Kratky, O. Small-Angle X-ray Scattering; Academic Press: London, 1982. (33) Guinier, A.; Fournet, G. Small-Angle Scattering of X-rays; Wiley: New York, 1955. (34) Turkevich, J.; Stevenson, P. C.; Hillier, J. A Study of the Nucleation and Growth Processes in the Synthesis of Colloidal Gold. Discuss. Faraday Soc. 1951, 11, 55−75. (35) Duff, D. G.; Edwards, P. P.; Johnson, B. F. G. Formation of a Polymer-Protected Platinum Sol: A New Understanding of the Parameters Controlling Morphology. J. Phys. Chem. 1995, 99, 15934−15944. (36) Ciacchi, L. C.; Pompe, W.; De Vita, A. Initial Nucleation of Platinum Clusters after Reduction of K2PtCl4 in Aqueous Solution: A First Principles Study. J. Am. Chem. Soc. 2001, 123, 7371−7380. (37) Hwang, B.-J; Sarma, L. S.; Chen, C.-H.; Bock, C.; Lai, F.-J.; Chang, S.-H.; Yen, S.-C.; Liu, D.-G.; Sheu, H.-S.; Lee, J.-F. Controlled Synthesis and Characterization of Rucore−Ptshell Bimetallic Nanoparticles. J. Phys. Chem. C 2008, 112, 19922−19929. (38) Henglein, A.; Giersig, M. Reduction of Pt(II) by H2: Effects of Citrate and NaOH and Reaction Mechanism. J. Phys. Chem. B 2000, 104, 6767−6772. (39) Yang, J.; Deivaraj, T. C.; Too, H.-P.; Lee, J. Y. Acetate Stabilization of Metal Nanoparticles and its Role in the Preparation of Metal Nanoparticles in Ethylene Glycol. Langmuir 2004, 20, 4241− 4245. (40) Chang, S. C.; Ho, Y.; Weaver, M. J. Applications of Real-Time FTIR Spectroscopy to the Elucidation of Complex Electroorganic Pathways: Electrooxidation of Ethylene Glycol on Gold, Platinum, and Nickel in Alkaline Solution. J. Am. Chem. Soc. 1991, 113, 9506−9513. (41) Ciacchi, L. C.; Mertig, M.; Pompe, W.; Meriani, S.; De Vita, A. Nucleation and Growth of Platinum Clusters in Solution and on Biopolymers. Platinum Met. Rev. 2003, 47, 98−107. (42) Tao, A. R.; Habas, S.; Yang, P. Shape Control of Colloidal Metal Nanocrystals. Small 2008, 4, 310−325. (43) Ott, L. S.; Finke, R. G. Transition-Metal Nanocluster Stabilization for Catalysis: A Critical Review of Ranking Methods and Putative Stabilizers. Coord. Chem. Rev. 2007, 251, 1075−1100.

13079

dx.doi.org/10.1021/la3026232 | Langmuir 2012, 28, 13072−13079