Formation of silver nanoparticles in PVP matrix in supercritical CO 2 ...

ISSN 19950780, Nanotechnologies in Russia, 2009, Vol. 4, Nos. 9–10, pp. 700–710. © Pleiades Publishing, Ltd., 2009. Original Russian Text © E.V. Shtykova, K.A. Dembo, V.V. Volkov, E.E. SaidGaliev, A.I. Stakhanov, A.R. Khokhlov, 2009, published in Rossiiskie nanotekhnologii, 2009, Vol. 4, Nos. 9–10.

ARTICLES

Formation of Silver Nanoparticles in PVP Matrix in Supercritical CO2: SmallAngle Xray Scattering and Modeling E. V. Shtykovaa, K. A. Demboa, V. V. Volkova, E. E. SaidGalievb, A. I. Stakhanovb, and A. R. Khokhlovb a

b

Institute of Crystallography, Russian Academy of Sciences, Leninskii pr. 59, Moscow, 119333 Russia Institute of Elementoorganic Compounds, Russian Academy of Sciences, ul. Vavilova 28, Moscow, Russia email: [email protected] Received January 29, 2009

Abstract—The process of silverparticle formation in a PVP matrix in supercritical carbon dioxide is investi gated by means of smallangle Xray scattering (SAXS). Application of modern methods of SAXS data inter pretation, including procedure of ab initio modeling of particle structure, allowed us for the first time to reveal structural organization of both individual metal nanoparticles and of their clusters incorporated in the poly mer matrix. It is shown that pores in PVP films are not simple cavities; they have complex structures which are exhibited when these cavities are contrasted (filled) with metal compounds and/or reduced metallic nanoparticles. As a result, the shape, size, and size distribution of the obtained metallic nanoparticles and their clusters depended on the structure of the polymeric matrix used as a formation medium. DOI: 10.1134/S1995078009090122

INTRODUCTION The idea of metallic nanoparticle formation in a polymer media originated when studying the structure of selforganizing polymer complexes. It was assumed that the shape, size, and size distribution of nanopar ticles incorporated into different polymer nanocom pounds will be determined by the structural parame ters and properties of the used polymer matrices and their spatial lattices. Subsequent studies showed that this approach is promising [1–8]. This problem remains relevant even now, because modern technolo gies necessitate the production of nanocomposites with specified parameters for solving concrete indus trial tasks. It is also important that a forming medium does not only determine the shape and the size of nanoparticles; the polymer matrices with incorpo rated particles themselves should acquire new unique and useful properties [9–16]. Thus, e.g., adding silver and copper into the polymers used in medicine imparts antibacteriological and anticarcinogenic activity to the materials [17, 18]. Especial promising seems to be the incorporation of biologically active sil ver nanoparticles, which due to the size effect has a higher bacterial action when compared to that of the block metal, into the polymers used for medicine such as PVP, chitosan, collagen and polylactide [19]. Substances in a supercritical (SC) state occupy an exceptional place in the development of new polymer compositions with specified properties. In particular, supercritical carbon dioxide, which is characterized by

low viscosity (~100 times lower than that of liquids) and high diffusion coefficients (~100 times higher than those of liquids), is successfully used in polymer synthesis and modification [20–22], as well as in extraction, chromatography, etc [23]. These and other specific physicalchemical properties of supercritical CO2 have been the motivation for its application in many processes and technologies. For example, SC carbon dioxide does not oxidize and is inert in the presence of free radicals; thus, it is used as a solvent in many chemical processes, including polymerization and polycondensation [24, 25]. In addition, SC CO2 is cheap, easytoaccess, noninflammable, nonexplo sive, chemically inert, and spontaneously evaporates from the reaction system after synthesis is completed, which is of particular importance for medical and electrochemical applications. Supercritical CO2, used as a solvent, does not participate in the transfer of chains during radical polymerization and allows its solubilizing ability to be controlled using both temper ature (the conventional method) and pressure. In addition, when synthesis is over, liquid and SC CO2 can be utilized in a flow regime to remove such impu rities as residual monomers, oligomers, and catalysts from the produced polymers [23]. Thanks to its excellent transport properties, SC CO2 is a promising medium for polymer impregnation [26]. The fundamental advantage of SCimpregnation is attributed to the absolute wetting ability of the SC medium (its absolute permeability into any open

700

FORMATION OF SILVER NANOPARTICLES

pores) [27]. This is why supercritical carbon dioxide is used more and more frequently to form nanosized metallic particles in polymer matrices and to impart new functional properties to polymer materials. When solving the problem of developing new nano materials, special attention is paid to the methods of investigating the structure of synthesized substances. In this case it is important not only to obtain adequate structural characteristics, but also to be able to study the samples in their natural environment and in any aggregate state, i.e. the possibility of nondestructive stagebystage control over the structure of the synthe sized composites without any pretreatment or prelim inary preparation for measurements. This allows undesirable structural changes in the samples to be avoided when manipulating them and excludes any inadequate interpretation of the experimental data that is obtained. One nondestructive testing method is the technique of smallangle Xray scattering (SAXS), which has lately become an efficient, informative, and (for a number of systems) the only accessible method of studying the superatomic structure of matter both in the solid state and in solutions with different concen trations [28–34]. One important field of SAXS application is in investigating a polymercompound structure. As an object of investigation, polymer compounds are char acterized by a number of specific peculiarities, the most significant being polydispersity. Thus, special methods for obtaining and interpreting SAXS data dif ferent from those used to analyze monodispersed compounds are utilized when the structure of polymer systems is analyzed [28, 29, 35–40]. Since these meth ods allow for simultaneously determining the struc tural characteristics of substances and the size distri bution of different inhomogeneities present in them, SAXS is an obvious choice for investigating the struc ture of polymer matrices and the processes of metallic nanoparticle formation. In the present paper, the technique of smallangle Xray scattering is applied to perform a structural analysis of polyvinyl pyrrolidone (PVP) containing sil ver nanoparticles formed in supercritical CO2. EXPERIMENT Synthesis Materials. 1Vinyl2pirrolidone (Aldrich, chemi cal purity over 99%) was distilled at a pressure of 11 mm of mercury, and the basic fraction with Tboil = 93–95°C was collected. AIBN (Fluka, chemical purity over 98.0%) was used as an initiator without preliminary purification. For silver impregnation, we used the complex (1,5 –cyclooctadien)1,1,1,5,5,5 hexafluoride acetylacetonate of silver CODAg[hfacac] NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

701

(Aldrich, chemical purity 99%) with gross formula C13H13AgF6O2 and structural formula CF3 C

O Ag

CH C

.

O

CF3

Polymerization and impregnation were performed in a CO2 atmosphere (GOST 805085, purity 99.997%, moisture content 0.001%). Hydrogen (TU 62000209585372004, purity 99.97%, moisture content 0.02%) was used to reduce the silver complex. Technique. PVP powder in the quantity required for producing a film of specified thickness was solved in a mixture of methanol and methylene chloride (1 : 1 in volume). The film was formed by pouring on a teflon substrate. After the solvent evaporated, the film was placed in a vacuum cabinet and dried for 12 hours at 20°C, and then for 8 hours at 60°C. The produced films were 10–13 μm, 78–90 μm, and 190–199 μm thick. CODAg[hfacac] was impregnated into PVP films in SC CO2 using the following technique. PVP film was placed in the upper part of a metallic reactor divided into two parts by a metallic grid, whereas the impregnated complex and a mixer in a teflon coating were put into the lower part. The reactor was sealed, blown through with CO2, and heated to 100°C; then a pressure of 9 MPa was set with a manual press. The impregnation duration was 5 h. CODAg[hfacac] was reduced in the polymer matrix in the hydrogen current. A film impregnated with a silver complex was placed on the bottom of the metallic reactor. The reactor was sealed, blown through with a small amount of CO2, and heated to 65°C. Hydrogen from a cylinder was fed through the needle input tap until the pressure reached 20 MPa. The reduction time was 4 h. Investigation Technique Smallangle Xray experiment and data analysis. The measurements were performed using conven tional SAXS with a Hecus laboratory diffractometer (Austria) at the fixed radiation wavelength λ = 0.1542 nm. The experimental scattering curves were measured within the range of wave vectors 0.07 < s < 0.6 nm–1 (s = 4πsinθ/λ, 2θ is the scattering angle). The obtained experimental data were normalized to the incident beam intensity; after that they were cor rected for collimation distortions according to the standard procedure [29]. The obtained experimental data on smallangle scattering were preprocessed with PRIMUS software [38]. The GNOM interactive program [35] was used to 2009

702

SHTYKOVA et al.

analyze the size distributions of structural inhomoge neities in the analyzed samples. In this case, volume distribution functions DV(R) were found by solving the integral equation R max

I ( s ) = ( Δρ )

2

∫ D ( R )m ( R )i ( sR ) dR 2

V

0

(1)

R min

under the assumption of the sphericity of scattering objects. Here I(s) is the intensity of smallangle scat tering; R is the sphere radius; Rmin and Rmax are the minimal and the maximal size; and i0(x) = {[sin(x) – xcos(x)]/x3}2 and m(R) = (4p/3)R3 are the sphere formfactor and volume, respectively. When calculat ing DV(R), the value of Rmin was taken as zero, whereas Rmax was chosen individually for each concrete case by the best fit. The radii of gyration of the nanoparticles Rg were calculated based on SAXS data using Guinier approx 2 2 imation Iexp(s) = I(0)exp( – s R g /3 ), which is valid within the range (sRg) < 1.3 [29]; they were also deter mined from the distance distribution function p(r) = r2γ(r), which was calculated from the experimental data using GNOM software of the indirect Fourier inversion [35], with the equation S max

2 1 sin ( sr) ds. γ ( r ) = 2 s I ( s ) sr 2π s = 0

∫

(2)

The function p(r) was also used to determine the maximum sizes of scattering objects Dmax. This param eter is necessary for ab initio structural modeling, i.e. for modeling which does not require any a priori infor mation and is performed only based on the data on smallangle scattering. The ab initio method was described in detail in [36] and is successfully being used to reconstruct the spatial structure of different samples, including polymer ones [41–45]. This method is based on annealing modeling and is imple mented as DAMMIN software [36]. This program shows the object under study as a totality of virtual atoms positioned in a sphere whose diameter is equal to the maximum object size. Starting from an arbitrary configuration of virtual atoms, the software, based on the annealing technique, constructs a model so as to minimize the deviations of χ from the experimental data Iexp(s) and the scattering curve calculated for the model: 2 1 χ = N–1

∑ j

I exp ( s j ) – cI calc ( s j ) 2 , σ ( sj )

(3)

where N is the number of experimental points; c is the scaling coefficient; and Icalc(sj) and σ(sj) are the calcu lated intensity for the model and the experimental error for the angular vector sj, respectively.

Because the analyzed samples are characterized by some polydispersity, the ab initio technique was repeatedly applied to each sample (no less than 10 independent passages of DAMMIN) and the obtained models were averaged. Electronmicroscopic investigation. The measure ments were performed with an EM301 transmission electron microscope (Philips) at the accelerating volt age of 80 kW. The samples to be investigated were crushed in an agathic mortar. The obtained particles were deposited onto an object grid precoated with a carbon film. Spetrophotometry. UVspectra of the initial, impregnated and reduced polymer samples were taken with a Helios Alpha&beta device in the form of thin films deposited onto quartz substrates from the solu tion in the mixture of methanol and methylene chlo ride. RESULTS AND DISCUSSION To successively study the processes of silver nano particle formation in PVP matrices using smallangle Xray scattering, we studied the structure of the poly mer matrix itself, then PVP impregnated with (COD)Ag[acac], and then the sample with reduced metal nanoparticles. The obtained experimental scat tering curves are shown in Fig. 1. SAXS profiles for all the samples are characterized by central scattering, i.e., scattering in the range of the smallest angles, which points to the inhomogeneous electron density of the samples at which scattering takes place. For a polymer matrix, this is scattering by the pores of inter nal matrix cavities; for the other two samples this cen tral scattering is scattering from PVP and impregnated silver component (curve 2) and scattering from the matrix with reduced metallic nanoparticles (curve 3). The first stage of structural analysis for these sam ples lies in determining the size distribution of scatter ing inhomogeneities, i.e., clusters of impregnated sil ver compounds and reduced metallic nanoparticle. Scattering of the initial PVP matrix is negligible; therefore, the size distribution of initial polymer pores in this case is impossible. However, indirectly, the pore sizes can be judged by the sizes of silver complex clus ters, which contrast the pores when filling due to the higher electron density. To determine the size distribu tions of both reduced silver nanoparticles and (COD)Ag[acac] impregnated into PVP, it is necessary to subtract scattering from the polymer matrix from the total scattering from the samples. An insertion in Fig. 2 gives the obtained difference curves: curve 1 is spattering at nanoclusters of the metal component incorporated into the PVP matrix and curve 2 corre sponds to scattering when silver nanoparticles are reduced in the matrix. These difference curves were used to define radii of gyration Rg, to calculate size dis tribution functions DV(R), and to determine shapes of

NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

2009


703

IgI, relative units

5 1 2 3

4

3

0

1

2

3

4

5 s, nm−1

Fig. 1. Experimental curves of smallangle scattering for PVP. (1) Initial PVP matrix; (2) PVP matrix with impregnated 1,5 cyclooctadien) 1,1,1,5,5,5 hexafluoride acetylacetonate of silver; and (3) PVP with reduced silver nanoparticles.

nanoparticles of the metal (both reduced and in the form of (COD)Ag[acac] clusters). The volumetric size distribution functions Dv(R) for these samples calculated with GNOM software are shown in Fig. 2. The most widespread size of nanopar ticles for both samples is on the order of 2R = 1 nm, and there is a small but detectable number of large par ticles with sizes up to 2R = 15–16 nm. The radii of gyration obtained when calculating size distribution functions are 5.7 nm ± 0.2 nm for silver complex clus ters and 5.0 nm ± 0.05 nm for reduced metal nanopar ticles; i.e. the latter are slightly more compact than the initial clusters of the metal component used. It should be taken into account that the radii of gyration are z average magnitudes and they carry the weight fraction of the larger particles present in both samples; i.e. the calculated Rg are actually larger than the radii of the major fraction of the smallest nanoparticles. It should be noted that the amplitude of the func tion Dv(R) for reduced nanoparticles of metals is suf ficiently higher than that of the volumetric distribution function for (COD)Ag[acac] clusters. Because the electron density of a pure metal is higher than the elec tron densities of their compounds and the amplitude of the function Dv(R) is proportional to the electron NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

density of scattering objects, we can conclude that the process of silver recovery has been completed success fully. The second important conclusion from the analysis of the obtained volumetric size distribution function lies in the fact that their shapes for both samples are similar: the major fraction of the smallest nanoparti cles actually has one and the same size (it is apparent from Fig. 2 that the average size of the major fraction of silver complex nanoclusters only slightly differs toward an increase from the average sizes of reduced particles), and larger nanoparticles are present in both samples. This means that they were formed in one and the same limited space specified by the internal struc ture of the polymer matrix: its pores. As it is apparent from Fig. 2, size distribution in the major nanoparticle fraction is very narrow and the number of large nanoparticles is small; i.e., these dis tributions are actually monodispersed. Thus an attempt was made to determine the shape of the (COD)Ag[acac] cluster particles and silver nanoparti cles using the DAMMIN software intended for mod eling the spatial structure of particles based on scatter ing data. The main advantage of this technique is attributed to the fact the shape is calculated ab initio, 2009

704

SHTYKOVA et al. DV(R) 2.5 IgI, relative units 5

2.0

1 2 4 1.5

3 1.0 2 0

1

2

0.5

3

4 5 s, nm−1

0 0

2

4

6

8 R, nm

Fig. 2. Size distributions of (1) nanoclusters of a metallic component incorporated into PVP matrix and (2) silver nanoparticles reduced in the matrix. The insert shows the corresponding difference curves: (1) the curve of scattering from nanoclusters of the metal component incorporated into PVP matrix and (2) the curve of scattering from silver nanoparticles reduced in the matrix.

i.e., no additional data on the object under study (except those on the maximum size and the function p(r)) are required and, thus, no preliminary hypothesis on the structure to introduce subjective factors into the calculations has to be advanced. The peculiarity of applying the ab initio method and DAMMIN software in the present study is as fol lows: different regions of the experimental curve of smallangle scattering are used for the first time to reconstruct multidimensional shapes which are present in the samples under analysis. The smallangle scattering curve contains data on both the shape and internal structure of particles [28, 29]. Each region of the curve corresponds to scattering of a certain size: the data on the largest sizes and, correspondingly, on the shape of particles are contained in the initial region of the curve, whereas the structure of internal homo geneities of a particles can be determined based on its middle region. At large scattering angles, the intensity curve corresponds to the structure at the atomic reso lution level and is unfit for the investigations utilizing

the smallangle scattering technique due to the low intensity and poor selfdescriptiveness of the data. Therefore, in the present study we used corresponding regions of the scattering curve to obtain data on both the shape of the smallest nanoparticles and a configu ration of their aggregates or clusters. To determine the structure as a whole and the shapes of nanoparticles and their clusters or aggregates, the calculations were performed in the region of modules of scattering vec tors within the range 0.97 < s < 5.0–5.2 nm–1, i.e. actu ally on the total experimental scattering curve without taking into account its region, which corresponded to larger angles and was the region characterized by increased statistic noise and affecting the stability of problem solutions. To determine the shape of individ ual nanoparticles, the initial part of experimental scat tering curves (the region of polydispersity and large formations) was cut off and final calculations were performed within the range of 1.75 < s < 5.2 nm–1. The main criterion for the possibility of applying the ab initio method at different regions of the scatter


Vol. 4

Nos. 9–10

2009


705

lgI, relative units p(r) 12 10 8 6 4 2

5

4

0

5

10

15 r, nm

3 1 2 3

0

1

2

3

4

5 s, nm–1

Fig. 3. Reconstructed shapes of silver chelator clusters in PVP matrices: (1) SAXS experimental curve, (2) the curve of extrapo lation into zero angle when calculating the p(r) function with GNOM software, and (3) scattering curve from the ab initio model. The top left insert shows the p(r) function of the given sample. The top right insert shows a spatial model of silver chelator clusters.

ing curve is that the Guinier region extends out enough (sRg < 1.3, where the Guinier law Iexp(s) = 2

2

I(0)exp( – s R g /3 holds) [29]. We obtained the follow ing values of radii of gyration Rg calculated using Guinier approximation: (i) For a sample with impregnated (COD)Ag[acac], 5.9 ± 0.04 nm; sRg = 0.793–1.32 for (COD)Ag[acac] clusters; 0.8 ± 0.07 nm; sRg = 1.47–1.63 for individual nanoparticles. (ii) For a sample with reduced metal, 5.0 ± 0.05 nm; sRg = 0.679–1.41 for nanoparticle clusters; 0.7 ± 0.01 nm; sRg = 0.995–1.32 for individual nanoparticles. Taking into account the polydispersity of the sam ple, the values of sRg were quite admissible in order to determine the shapes of scattering objects, except for individual particles of silver complex clusters. As far as the latter were concerned, their range of sRg values were beyond admissible values; thus the shape of indi NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

vidual nanoparticles in these clusters was not calcu lated. It should be noted that, like in the case of calculat ing radii of gyration with GNOM software (see above), Rg for the aggregates and individual nanoparticles of the reduced metal turned out to be a bit smaller than those for (COD)Ag[acac] clusters, which proved once again that the compound that was used becomes more dense in the process of metal reduction. The distance distribution function p(r) for silver complex clusters, the curve of scattering from the spa tial structural model obtained using DAMMIN soft ware, and the ab initio model itself are shown in Fig. 3. The clusters of the silver complex in the PVP matrix are elongated loose spiralshaped bodies with a mini mal size (length) of 16 nm and a crossregion on the order of 5–6 nm. For the obtained model, χ = 1.41. Based on the structural model that was obtained, we can conclude that silver complex clusters do not con sist of pronounced individual particles. In addition, no reliable Guinier range for calculating the shapes of these individual particles was found. Though in Fig. 3 we can seen weakly pronounced dense regions in (COD)Ag[acac] clusters (the regions where the pro 2009

706

SHTYKOVA et al. IgI, relative units

(a) p(r) 5 4

5

3 2 1 0

4

0

2

4

6

8

10 12 14 r, nm

4

5 s, nm−1

1 2

3

3

0

1

2

3

lgI, relative units (b)

4

1 2 3

p(r) 10 8 6 4

3

2 0

0

0.2 0.4 0.66 0.8 1.0 1.2 1.4 r, nm 1

2

3

4

5 s, nm−1

Fig. 4. (a) Experimental and model scattering curves for the p(r) function for the sample with reduced silver nanoparticles calcu lated for the range 0.07 < s < 5.2 nm–1. (1) SAXS experimental curve, (2) the curve of extrapolation into zero angle when calcu lating p(r) function with GNOM software, and (3) the model curve for the given function p(r). The insert shows p(r) functions for the given range. (b) Experimental and model scattering curves for the p(r) function for the sample with reduced silver nanoparti cles calculated for the range 1.75 < s < 5.2 nm–1. (1) SAXS experimental curve, (2) the curve of extrapolation into zero angle when calculating the p(r) function with GNOM software, and (3) the model curve for the given function p(r). The insert shows the p(r) functions for the given range.

cess of silver reduction is likely to start at the stage of impregnation), the considered clusters are, on the whole, formations comparatively homogeneous in density that fill the pores of a polymer matrix and contrast it.

When silver was reduced using the ab initio method, spatial structural models were obtained both for individual nanoparticles of metal as for its aggre gates. First of all, the functions p(r), which correspond


Vol. 4

Nos. 9–10

2009


707

lgI, relative units 2 1 5

4

1 2 3

3

0

1

2

3

4

5 s, nm–1

Fig. 5. Reconstructed shapes of individual nanoparticles of silver and their aggregates in PVP matrices: (1) SAXS experimental curve, (2) the curve of scattering from ab initio model for silver nanoparticle aggregates, and (3) the curve of scattering from the ab initio model for individual silver nanoparticles. The top insert shows a spatial model of individual silver nanoparticles; (2) is the spatial model of nanoparticle aggregates.

200 nm

(а)

(b)

50 nm

Fig. 6. Microphotograph of PVP sample particle (a) and a microphotograph of a particle of PVP with unreduced metal (b). NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

2009

708

SHTYKOVA et al.

Optical density, D

3

4

2

1

300

600

900 λ, nm

Fig. 7. UV spectra of (1) initial PVP film, (2) CODAg[hfacac] complex (solution in methanol), (3) PVP film impregnated with CODAg[hfacac] complex, and (4) the same film after complex reduction with hydrogen.

to different regions of the scattering curve, were calcu lated. They are shown in Figs. 4a, 4b. It is of interest to compare the p(r) functions calculated for the total curve of the samples containing unreduced and reduced silver. It can be noted that the p(r) profile for the sample containing reduced silver (Fig. 4a) is char acterized by two maxima and the maximal size Dmax = 14.5 nm. The presence of two maxima means that there are two different major sizes of nanoparticles, namely 1 and 5–6 nm. On the other hand, an asym metric curve shape is characteristic for elongated bod ies [29]. The length of the particle (Dmax = 14. 5 nm) is its largest size, and its crosssectional dimension can be 5–6 nm. Then inhomogeneities inside the particle itself are 1 nm in size. The size of these inhomogene ities coincides with the size of the major fraction of reduced silver nanoparticles (Fig. 2); i.e., it can be assumed that metal nanoparticles with an average size of 1 nm form elongated aggregates on the order of 14 nm long and crosssectional dimensions on the order of 5–6 nm. At the same time the distance distribution function p(r) for the sample with an impregnated silver compo nent calculated in the same angle range (Fig. 3) has only one pronounced maximum and a weakly expressed shoulder in the region with small sizes on the order of 1–2 nm; i.e., the individual particles which form the clusters of unreduced silver do not actually form, which corresponds to the analysis of

Guinier regions on the curve performed earlier, to the shape of clusters obtained ab initio. The reconstructed shapes of individual silver nano particles and their aggregates obtained using DAM MIN software are shown in Fig. 5 together with the calculated scattering curves from ab initio models. The deviation of model curves from the experimentally obtained χ in this case was 1.5 and 1.6 for individual silver nanoparticles and their aggregates, respectively. As is apparent from Fig. 5, the reduced silver in the PVP matrix takes the form of spiralshaped aggregates 14.5 nm long and 5–6 nm in the crossregion, which consist of regions with different densities. The denser regions are practically spherical nanoparticles about 1 nm in size. Their shape coincides with that found for individual nanoparticles of silver. On the whole, the structure of aggregates is similar to that of (COD)Ag[acac] clusters, which points to the fact that they were formed in the same internal space of the polymer matrix in cavities of the PVP film, which are not simply hollows, but have a complex structure which becomes apparent when these cavities are con trasted (filled) either with silver compounds or reduced metallic nanoparticles. Thus it is the structure of the polymer matrix that determined the shape and the size of the metallic nanoparticles formed in it. The cluster sizes of (COD)Ag[acac] or reduced metal that we obtained using the technique of small angle scattering agree well with the sizes of pores con


Vol. 4

Nos. 9–10

2009


trasted with unreduced silver according to the data of electron microscopy (Fig. 6) [46]. In addition, using smallangle Xray scattering we showed that the reduction of silver in a PVP matrix in the medium of SC CO2 was successful. This conclusion is in agree ment with the spectrophotometric data. It is apparent from Fig. 7 that the region of PVP film absorption (curve 1) is 300 nm lower and does not hinder any observation of the changes which take place when a silver complex is incorporated into the matrix. After the PVP film is impregnated with (COD)Ag[hfacac] complex, a new broad band with a maximum at ~418 nm (the band of the Plasmon resonance of silver particles [47]) (curve 3) appears along with the intrin sic absorption band of the complex (λmax = 324 nm); after reduction, this new band remains the only one (curve 4). Hence, the reduction of the (COD)Ag[hfa cac] complex with hydrogen was successful, though such a reduction starts by the stage of impregnation and Ag nanoparticles of different sizes are formed; this stage manifests itself in the broadening of correspond ing absorption bands (curves 3 and 4) [48]. This con clusion is confirmed by the aboveconsidered ab initio analysis of (COD)Ag[hfacac] cluster shapes. Thus, the technique of smallangle scattering allowed to determine not only the character of inner structure of PVP film pores filled with reduced or unreduced silver, but also to reveal the shape of indi vidual silver nanoparticles and their size distribution in the analyzed composite polymer systems. CONCLUSIONS Two main problems were stated in the present paper: first, to investigate the structure of complex polymer systems and processes of metallic nanoparti cle formation in them and, second, to show the mod ern potentialities of the technique of smallangle scat tering as applied to an analysis of such systems and processes. Making use of the latest developments, we managed to study the structural organization of a PVP matrix impregnated with a silver complex and reduced metallic particles with a 0.5–1 nm resolution. Differ ent regions of experimental smallangle scattering curves were used for the first time to reconstruct mul tidimensional forms contained in the studied samples. The corresponding size distribution functions were calculated and the shape, size, and size distribution of silver nanoparticles formed in a PVP film was shown to be controlled by the corresponding parameters of its porous system. The obtained results show that a super critical technique makes it possible to impregnate organic metallic complexes (precursors of metallic nanoparticles) into polymer molecular solutions and ensures the formation of a set of metallic nanoclusters with small sizes and narrow size distributions in a polymer. Thus the supercritical technique is undoubt edly valuable from the practical point of view, because it makes it possible to produce nanometallic polymer NANOTECHNOLOGIES IN RUSSIA

Vol. 4

Nos. 9–10

709

composites with the specified functional characteris tics required for their application in medicine, elec tronics, tribology, and many other fields of science and engineering. ACKNOWLEDGMENTS This study was supported by State Program for Leading Scientific Schools (project NSh 1955.2008.2), the Russian Foundation for Basic Research (project no.080312152), Russian Acad emy of Sciences (program OKhNM–6) and European grant EU RIDS SAXIER Contract 011934 (www.sax ier.org). REFERENCES 1. M. Moffitt, H. Vali, and A. Eisenberg, Chem. Mater. 10, 1021 (1998). 2. T. Hashimoto, M. Harada, and N. Sakamoto, Macro molecules, 32, 6867 (1999). 3. R. S. Underhill and G. Liu, Chem. Mater. 12, 2082 (2000). 4. Z. Lu, G. Liu, H. Phillips, J. M. Hill, J. Chang, and R. A. Kydd, Nano Lett, 1 683 (2001). 5. L. M. Bronstein, D. I. Svergun, and A. R. Khokhlov, in Polymer Gels and Networks, Ed. by Y. Osada and A. R. Khokhlov (Marcel Dekker, New York, 2001), pp. 103–130. 6. Metal–Polymer Nanocomposites, Ed. by L. Nicolais and G. Carotenuto (John Wiley and Sons, Hoboken, NJ, United States, 2004), p. 320. 7. R. S. Underhill and G. Liu, Chem. Mater. 12, 2082 (2000). 8. D. I. Svergun, M. B. Kozin, P. V. Konarev, E. V. Shtyk ova, V. V. Volkov, D. M. Chernyshov, P. M. Valetsky, and L. M. Bronstein, Chem. Mater. 12, 3552 (2000). 9. Nanotechnology Research Directions: Vision for Nano technology in the Next Decade, Ed. by M. C. Roco, R. S. Williams, and P. Alivisatos (Kluwer, New York, 1999; Mir, Moscow, 2002). 10. G. Yu. Yurkov, D. A. Astaf’ev, L. N. Nikitin, Yu. A. Koksharov, N. A. Kataeva, E. V. Shtykova, K. A. Dembo, V. V. Volkov, A. R. Khokhlov, and S. P. Gubin, Neorg. Mater. 42 (5), 556–562 (2006) [Inorg. Mater. 42 (5), 496–502 (2006)]. 11. V. I. Salyanov, A. I. Evseev, V. I. Popenko, A. A. Gasanov, K. A. Dembo, O. V. Kondrashina, Yu. V. Shtykova, and Yu. M. Yevdokimov, Biofizika, 52 (3), 452–459 (2007) [Biophysics 52 (3), 288–292 (2007)]. 12. T. Liu, C. Burger, and B. Chu, Prog. Polym. Sci. 28, 5– 26 (2003). 13. V. V. Volkov, V. V. Klechkovskaya, E. V. Shtykova, K. A. Dembo, N. A. Arkharova, G. I. Ivakin, and R. Yu. Smyslov, Kristallografiya, 54 (2), 197–201 (2009) [Crystallogr. Rep. 54 (2), 169–173 (2009)]. 14. S. Xu, J. Zhang, C. Paquet, Y. Lin, and E. Kumacheva, Adv. Funct. Mater. 13, 468–472 (2003). 2009

710

SHTYKOVA et al.

15. H. H. Pham and E. Kumacheva, Macromol. Symp. 192 191–205 (2003). 16. A. Alteheld, I. Gourevich, L. M. Field, C. Paquet, and E. Kumacheva, Macromolecules 38, 3301–3306 (2005). 17. Lifeng Qi, Zirong Xu, Xia Jiang, Van Li, and Mingi Wang, Bioorg. Med. Chem. Lett. 15, 1397–1399 (2005). 18. Yong Zheng, Ying Yi, Yipeng Qi, Yuting Wang, Weian Zhiang, and Ming Du, Bioorg. Med. Chem. Lett. 16, 4127–4129 (2006). 19. L. S. Villata, E. Wolcan, M. R. Felis, and A. L. Caparelli, J. Photochem. Photobiol., A 115, 185 (1998). 20. A. I. Cooper, J. Mater. Chem. 10, 207–234 (2000). 21. S. G. Kazarian, Polym. Sci., Ser. C 42 (1), 78 (2000). 22. E. J. Beckman, J. Supercrit. Fluids 28, 121–191 (2004). 23. M. A. McHugh and V. J. Krukonis, Supercritical Fluid Extraction: Principles and Practice (Butterworth–Hei nemann, Boston, MA, United States, 1994). 24. J. L. Kendall, D. A. Canelas, J. L. Young, and J. M. DeSimone, Chem. Rev. 99, 543–563 (1999). 25. E. E. SaidGaliyev, I. V. Pototskaya, and Ya. S. Vygodskii, Vysokomol. Soedin., Ser. C 46 (12), 2124–2139 (2004) [Polym. Sci., Ser. C 46 (1), 1–13 (2004)]. 26. E. N. Sobol’, V. N. Bagratashvili, A. E. Sobol’, and S. M. Howdle, Dokl. Akad. Nauk 356 (6), 777–780 (1997) [Dokl. Phys. Chem. 356 (4–6), 346–348 (1997)]. 27. E. J. Beckman, J. Supercrit. Fluids 28, 121–191 (2004). 28. SmallAngle XRay Scattering, Ed. by O. Glatter and O. Kratky (Academic, London, 1982), p. 515. 29. L. A. Feigin and D. I. Svergun, Structure Analysis by SmallAngle XRay and Neutron Scattering (Plenum, New York, 1987). 30. D. I. Svergun, J. Appl. Crystallogr. 33, 530–534 (2000). 31. D. I. Svergun and M. H. Koch, Curr. Opin. Struct. Biol. 12 (5), 654 (2002). 32. D. I. Svergun, J. Appl. Crystallogr. 40, S10–S17 (2007). 33. M. H. Koch, P. Vachette, and D. I. Svergun, Q. Rev. Biophys. 36 (2), 147–227 (2003). 34. M. V. Petoukhov and D. I. Svergun, Curr. Opin. Struct. Biol. 17 (5), 562–571 (2007).

35. D. I. Svergun, J. Appl. Crystallogr. 25 495–503 (1992). 36. D. I. Svergun, Biophys. J. 76 (6), 2879 (1999). 37. V. V. Volkov and D. I. Svergun, J. Appl. Crystallogr. 36 (31), 860–864 (2003). 38. P. V. Konarev, V. V. Volkov, A. V. Sokolova, M. H. J. Koch, and D. I. Svergun, J. Appl. Crystallogr. 36 (5), 1277–1282 (2003). 39. P. V. Konarev, V. V. Volkov, M. V. Petoukhov, and D. I. Svergun, J. Appl. Crystallogr. 39 (2), 277–286 (2006). 40. M. V. Petoukhov, P. V. Konarev, A. G. Kikhney, and D. I. Svergun, J. Appl. Crystallogr. 40 (s1), s223–s228 (2007). 41. V. V. Volkov, V. A. Lapuk, R. L. Kayushina, E. V. Shtyk ova, E. Yu. Varlamova, M. Malfois, and D. I. Svergun, J. Appl. Crystallogr. 36, 503–508 (2003). 42. E. V. Shtykova, E. V. Shtykova, Jr., V. V. Volkov, P. V. Konarev, A. T. Dembo, E. E. Makhaeva, I. A. Ronova, A. R. Khokhlov, H. Reynaers, and D. I. Svergun, J. Appl. Crystallogr. 36, 669–673 (2003). 43. L. M. Bronstein, S. N. Sidorov, V. Zhirov, D. Zhirov, Y. A. Kabachii, S. Y. Kochev, P. M. Valetsky, B. Stein, O. I. Kiseleva, S. N. Polyakov, E. V. Shtykova, E. V. Nikulina, D. I. Svergun, and A. R. Khokhlov, J. Phys. Chem. B 109, 18 786–18 798 (2005). 44. E. V. Shtykova, X. Huang, N. Remmes, D. Baxter, B. Stein, B. Dragnea, D. I. Svergun, and L. M. Bron stein, J. Phys. Chem. C 111 (49), 18 078–18 086 (2007). 45. E. V. Shtykova, X. Gao, X. Huang, J. C. Dyke, A. L. Schmucker, N. Remmes, D. V. Baxter, B. Stein, B. Dragnea, P. V. Konarev, D. I. Svergun, and L. M. Bronstein, J. Phys. Chem. C 112 (43), 16 809–16 817 (2008). 46. E. E. SaidGaliev, A. I. Stakhanov, I. V. Blagodatskikh, E. M. Ivanova, A. R. Khokhlov, A. V. Naumkin, I. O. Volkov, V. V. Volkov, E. V. Shtykova, K. A. Dembo, and S. A. Pisarev, Vysokomol. Soedin., Ser. A 51 (2009) (in press) [Polym. Sci. 51 (2009) (in press)]. 47. A. O. Rybaltovskii, A. A. Aksenov, V. I. Gerasimova, Yu. S. Zavorotskii, V. K. Popov, V. V. Zosimov, and V. N. Bagratashvili, Sverkhkrit. Flyuidy: Teor. Prakt. 3 (2), 74–81 (2008). 48. A. O. Rybaltovskii, A. A. Aksenov, V. I. Gerasimova, V. V. Zosimov, V. K. Popov, A. B. Solov’eva, P. S. Tima shev, and V. N. Bagratashvili, Sverkhkrit. Flyuidy: Teor. Prakt. 3 (1), 50–57 (2008).


Vol. 4

Nos. 9–10

2009