- Email: [email protected]
Positronium lifetime measurements in micro- and mesoporous ferrisilicates
Radiation Physics and Chemistry 139 (2017) 49–54
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Positronium lifetime measurements in micro- and mesoporous ferrisilicates a,⁎
Trần Quốc Dũng , Károly Lázár a b
MARK
b
Center for Nuclear Techniques, 217 Nguyen Trai, District 1, Ho Chi Minh City, Viet Nam Centre for Energy Research HAS, H-1525 Budapest, P.O.B. 77, Hungary
A R T I C L E I N F O
A B S T R A C T
Keywords: Positron Annihilation Trapped o-Ps Porous Solids Zeolites Iron ion-exchanged zeolite
Positron annihilation lifetime, Doppler broadening measurements have been used to study microporous FER and MFI structures containing framework subtituted iron in different amounts. Mesoporous SBA-15 samples without and with iron introduced to the pore walls were also investigated by these methods. Mössbauer measurements have been accompanied to prove the extent of isomorphous incorporation of iron into the different structures. The results reveal a certain dependence of lifetimes on the various pore and channel sizes. Addition of iron does not influence significantly the observed lifetimes in FER microporous systems. However a slight decrease of lifetimes can be observed in the MFI samples. The effect of addition of iron in mesoporous SBA-15 system is more expressed. Supplement of iron also has an influence on the S parameter of Doppler broadening spectra.
1. Introduction Porous structures can advantageously be studied with positronium annihilation measurements. Annihilation of para positronium is fast (125 ps), in contrast to the lifetime of ortho positronium (o-Ps) which is larger with three orders of magnitude (140 ns). If o-Ps interacts with electrons of the environment (“pick-off” annihilation) the observed lifetime becomes shorter. If o-Ps positronium is trapped inside porous structures the observed lifetime can be correlated with the size of cages and channels (Gidley et al., 2006; Jean et al., 2003). The 10–130 ns lifetime interval corresponds to 2–100 nm pore size range by the extended Tao-Eldrup approximation ( Dull et al., 2001). Beside the commonly used lifetime measurements, energy distribution around the 511 keV annihilation peak can also be measured by Doppler broadening method. Small variations in the signal shape may provide information on the interactions of o-Ps with valency and core electrons of the host matrix via the S (shape) and W (wing) parameters, respectively (Jean et al., 2003; Liszkay et al., 2001). Usually relative tendencies are analyzed in S - W plots since selection of the considered ranges in Doppler spectra depends strongly on the experimental conditions (Kajcsos et al., 2009). Pore sizes of microporous silicates (zeolites) and mesoporous substances are in the mentioned size range. Typical microporous structures (MFI and FER) are strictly crystalline, with interpenetrating channel systems with diameters of 0.45–0.55 nm (Baerlocher, 2001). Characteristic features of mesoporous MCM-41 and SBA-15 structures are their larger channel diameter (4–6 nm) and their partly amorphous pore wall structure (Schacht et al., 1998). Both of these types of ⁎
materials have achieved broad variety of applications, thus they have been widely characterized by o-Ps measurements, too (Goworek, 2014; Kajcsos et al., 2005). Combination of micro - and mesoporous strucures, namely those with so-called hierarchical porosity can also be studied (Zubiaga et al., 2016). Studies on mesoporous substances (e.g. MCM41) with positronium lifetime measurements are also reported (Zaleski et al., 2003). Beside pure silica and alumina based porous structures, further ones containing additional transition metal ions substituted isomorphically into the framework are also widely used in the practice. Addition of the transient metal ion modifies the electron densities in the cages, thereby it may influence the lifetime of o-Ps positronium, too (Süvegh et al., 2001). Further on adsorption of gas molecules (nitrogen, oxygen and even water from humidity of air) takes easily place in cages of microand mesoporous substances, modifying thereby the actual size of cages (Cabral-Prieto et al., 2006; Kajcsos et al., 2007). Thus these measurements should usually be performed after evacuation. Iron modified MCM-41 with iron in the pore walls (Wiertel et al., 2013) as well as with Fe2O3 (Surowiec et al., 2010) or with mixed oxide Mn-ferrites (Wiertel et al., 2014) in the channels have also been characterized with the method. Iron containing zeolites and mesoporous substances can advantageously be studied by another technique, with Mössbauer spectroscopy, too. This method provides information solely on the close environment of iron. Combination of the two techniques may strenghten the interpretation of data (Zaleski et al., 2003; Wiertel et al., 2013; Kajcsos et al., 1993). In our previous publications, general aspects of studying porous
Corresponding author. E-mail addresses: [email protected] (T.Q. Dũng), [email protected] (K. Lázár).
http://dx.doi.org/10.1016/j.radphyschem.2017.05.012 Received 9 February 2017; Received in revised form 24 April 2017; Accepted 12 May 2017 Available online 17 May 2017 0969-806X/ © 2017 Elsevier Ltd. All rights reserved.
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
media with positronium studies were discussed (Cabral-Prieto et al., 2006). Further on sorption and removal of adsorbed water in zeolites (Liszkay et al., 2001), and early stages of crystallization of zeolites (Bosnar et al., 2007) were also followed by the method. Preliminary results of positronium lifetime measurements on comparison of microporous FER, MFI with various iron contents and mesoporous SBA-15 plain silicate with its iron modified ferrisilicate counterpart have been reported recently (Dung et al., 2013). The microporous structures exhibit 2- 20 ns lifetimes for o-Ps, whereas 4 – 52 ns lifetimes were found for mesoporous SBA-15 in correspondence with the expectations of the extended Tao-Eldrup model. Addition of iron decreased the lifetimes in Fe-MFI and in Fe-SBA in comparison to the corresponding plain porous silicates. Iron species in the microporous structures were identified by Mössbauer spectroscopy as well (Dung et al., 2013). The present report accounts on further studies on various iron modified microporous (Fe)-FER and (Fe)-MFI as well as on mesoporus plain and iron modified SBA-15. In particular, lifetime measurements are completed with determinations of S and W parameters from Doppler measurements, and a broader scope in interpretation is also provided. In situ Mössbauer spectra are also shown to identify the species of iron present in the mesoporous Fe-SBA.
Table 1 List of samples. Type
Sample
Si/ Fe (at)
Si / Al (at)
Ref.
Microporous (A)
A-Fe-FER A- (Al,Fe)-FER A-Fe-MFI-50 A-Fe-MFI-100 B-Fe-SBA−15 B-SBA−15
16 64 50 100 65 –
24 200 200 – –
[20] [20] [21] [21] [23] [22]
Mesoporous (B)
acterization are presented (Lázár et al., 2004). List of samples is shown in Table 1. 2.2. Mesurements 2.2.1. Lifetime (PALS) and Doppler (DBS) positronium mesurements In all PAS measurements the radioactive positron source has been sandwiched between two disk-shaped pieces of samples. Na-22 sources were used for both DBS and PALS measurements. The zeolite powder was pressed (0.3 GPa) into disks of 8–13 mm diameter and 1–1.5 mm thickness. In the PALS studies the samples were processed and prepared sandwiching the Na-22 source wrapped in Al foil. Samples and source were placed in a sample holder in an Al vacuum chamber. Adsorbed water was removed from the pores by a combined evacuation and heat treatment before the measurements performed at room temperature. PALS mesurements with low-activity (0.5–2 MBq) Na-22 source were performed. A fast-fast coincidence PALS was equipped with XP 2020 URQ photomultipliers and BaF2 scintillators and NE111 plastic scintillator. Wide energy selection windows were set to enable high efficiency, still providing good time resolution (250 ps FWHM for 60Co). The spectra were obtained by employing time-to-pulse height converter units, in case of simultaneous use of two of them allowing also the setting of different time ranges simultaneously. The data collection chains on the PALS consisted of standard ORTEC and TENNELEC units. The data collection took place with computer-based multichannel cards - the 16k Oxford Microfast products. The time calibration values for the various settings varied from 20.1 ps/ch up to 200.3 ps/ch, respectively. In each spectrum integral counts number at 2.106 were recorded. The time base of the PALS setup was 8192 channels. A good reproducibility of the results was found by recording spectra over several days and repeating the experiments. The lifetimes and relative intensities were also evaluated using the LT v.9 fitting program (Kansy, 1996). All PALS spectra were decomposed to four lifetime components. In DBS technique the 511 keV annihilation energy, which is broadened by the Doppler shift by the total momentum of the positron– electron annihilation pair, is directly measured. The S parameter, which corresponds to positron annihilation with the valence electrons, is defined as the ratio of the central area of the 511 keV annihilation gamma ray peak to the total peak area. The W parameter, which corresponds to positron annihilation with the core electrons, is the ratio of the edge area of the DB peak to the total area electrons. In this DBS study the apparatus is typical and energies in the range of 40 keV–1.4 MeV were recorded by using a Canberra HPGe detector with 1.86 keV FWHM resolution at the 1.28 MeV line of positron source. A low-activity (0.5–2 MBq) Na-22 source were used. Each spectrum carried about 6 million counts.
2. Experimental 2.1. Samples 2.1.1. Microporous (type A) samples Fe-FER samples with Si/T atomic ratio of 16 have been synthesized. T stands for the modifying ion substituted isomorphously to tetrahedral sites into the siliceous framework. In the first case, for Fe-FER sample, solely Fe was used, resulting in an average composition of FeSi16O34. In the second case, for (Al,Fe)-FER sample, both Al and Fe (in Al:Fe =3 ratio) were introduced into the tetrahedral framework sites (FeAl3Si64O136). Thus, the amount of iron was only ¼th in the latter sample in comparison to the first Fe-FER. Further details of the synthesis and characterization can be found (Lázár et al., 1998). The extent of isomorphous substitution with Fe and Al ions located in tetrahedral sites was significanly smaller for Fe-MFI samples than for the previous Fe-FER ones, Si/T exceeded 40 in each cases. Si/Fe ratios are 50 and 100 for samples Fe-MFI-50 and 100, respectively. In other words, iron content was doubled in Fe-MFI-50 as compared to Fe-MFI100. The dominant substituting ion was Fe, with addition of Al in 1/3 portion. Thus the average nominal compositons were AlFe2Si100O206 and AlFe2Si200O406 respectively. To provide the possibility for obtaining Mössbauer spectra at these low iron contents 57Fe isotope was used for sample preparation. Futher details of preparation and characterization are reported in (Fejes et al., 2003). 2.1.2. Mesoporous SBA-15 (type B) samples SBA-15 was prepared by using mixture of P123 block copolymer, water, acidified with HCl. The siliceous pore structure was formed from gelation with tetra-ethyl-orthosilicate with subsequent hydrothermal treatment in an autoclave at 100 °C for 24 h, then filtered and dried at room temperature. The template was removed by calcination at 450 °C in oxygen for 5 h. The specific surface area of the product was 580 m2g−1, the average channel diameter was 5.7 nm as determined from BET measurements. Further details of preparation and characterization are described in the publications (Beck et al., 2008). Fe-SBA-15 was prepared with Si/Fe ~65 ratio (approximate nominal composition FeSi65O132). The preparation was similar to that applied for the previous plain SBA-15 with addition of iron into the starting synthesis mixture. Namely, P123 block copolymer was dissolved in 1.9 NHCl, then FeCl3·6H2O was added to the solution. Further stages were similar to that applied for the synthesis of plain SBA-15. The specific surface area of the Fe-SBA-15 was 715 m2 g−1, the average channel diameter was 8.6 nm. Further details of synthesis and char-
2.2.2. Mössbauer measurements Mössbauer measurements were performed by using a spectrometer using 1 GBq 57Co source. Fe-FER, Fe-MFI and Fe-SBA samples were treated at different conditions (evacuation at 650 K, reduction in hydrogen at 620 K) in an in situ Mössbauer cell (Lázár et al., 1984). Most of spectra were collected at room temperature. Parameters were extracted by decomposing the spectra to components of Lorentzian line 50
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Table 2 O-Ps lifetimes obtained from decomposition of PALS spectra (numbers in parentheses show errors).
Microporous
Sample
τ3 ns
τ4 ns
I3 %
I4 %
A-Fe-FER
1.927 (0.024) 1.736 (0.041) 1.851 (0.020) 2.794 (0.046)
13.460 (0.350) 13.520 (0.500) 15.420 (0.300) 20.530 (0.730)
1.66 (0.04) 1.73 (0.05) 1.31 (0.02) 0.93 (0.04)
0.46 (0.01) 0.47 (0.01) 0.47 (0.01) 0.46 (0.01)
3.900 (0.130) 3.434 (0.074)
35.320 (0.190) 51.186 (0.098)
0.77 (0.05) 1.37 (0.02)
3.93 (0.24) 21.79 (0.29)
A-(Al,Fe)FER A-FeMFI−50 A-FeMFI−100 Mesoporous
B-FeSBA−15 B-SBA−15
Fig. 1. PALS spectra recorded for mesoporous SBA-15 and Fe-SBA-15. All spectrum are collected in 8 K channels, here illustrated in 3000 channels only.
shape. Data are referred to metallic alpha-iron. Accuracy of positional parameters (isomer shift and quadrupole splitting) is ± 0.03 mm/s.
syntheses treatments probably facilitated the removal of most of extraframework species. Considering the high Si/Fe ratio (i.e. the low iron content) most of pores and channels were probably open and accessible for o-Ps.
3. Results and discussion 3.1. Results
3.1.3. SBA-15 and Fe-SBA-15 samples PALS spectra recorded for mesoporous SBA-15 and Fe-SBA-15 are shown in Fig. 1. Spectra are distinctly different from those of the microporous samples. The deconvolution reveals that longer lifetimes of o-Ps are present. The increase is noticeable in τ3 and expressed in τ4, with 35 and 51 ns τ4 lifetimes for Fe-SBA-15 and SBA-15, respectively (see Table 2). Results for DBS measurements are shown in Fig. 2. It is clearly seen that the shapes of both DBS curves (in particular that recorded on ironfree SBA-15) are narrow, they are distinctly different from those recorded on microporous FER and MFI. Decomposition of in situ Mössbauer spectra recorded on Fe-SBA-15 (with 1.2 wt% Fe) is shown in Fig. 3. Mössbauer spectra of the sample recorded after evacuation demonstrates the incorporation of iron in ionic dispersion to the pore walls. Only a small part of iron is present as Fe2+, most of it retains the Fe3+ oxidation state in octahedral and tetrahedral coordinations. This assignments are in correspondence with results of other observations. For the interpretation the modest crystallinity of the pore walls in mesoporous silicates should also be considered (Lázár et al., 2004).
3.1.1. Fe-FER and (Fe,Al)-FER samples Data extracted for the o-Ps components are shown in Table 2 (neglecting the components characteristic for p-Ps with lifetimes less than 1 ns). o-Ps lifetimes in Fe-FER and (Fe,Al)-FER are rather short, both τ3 are less than 2 ns, and τ4 parameters are the smallest ones among the data in the table. DBS spectra for Fe-FER and (Fe,Al)-FER samples are shown in Fig. 2. The DB energy distributions obtained for both Fe-FER and (Fe,Al)-FER adhere closely to the data obtained for the other (MFI) microporous samples. Mössbauer spectra for Fe-FER and (Fe,Al)-FER sample are presented in our previous publication (Dung et al., 2013). Their interpretation is shortly amended here as follow. Spectra of Fe-FER display dominant presence of framework substituted iron on tetrahedral positions, i.e. after evacuation at 620 K most of the channels and pores is empty. However, later treatments (e.g. reduction in hydrogen at 620 K) may result in relocation of iron from framework to extra-framework positions in a certain extent. Only evacuation was applied before the PALS measurements, thus significant presence of extra-framework iron should not be expected for this sample. Mössbauer spectra for the (Fe,Al)-FER sample exhibit slightly different features. Evacuation results in slightly increased extent from framework to extra-framework relocation connected to auto-reduction, as can be seen in the modest appearance of Fe2+. Thus, the (Fe,Al)-FER sample is less stable, the channels and pores can be contaminated in a certain extent with various FeO+, AlO+ etc. extra-framework fragments, too (Lázár et al., 1998).
3.2. Discussion 3.2.1. Comparison of samples The lifetime and intensity data are compared in Fig. 4.
3.1.2. Fe-MFI-50 and Fe-MFI-100 samples Data extracted for the o-Ps components (τ3, τ4) are shown in Table 2. τ3 lifetimes are silmilar or slightly exceed those obtained for the previous Fe-FER samples, whereas τ4 data are significantly larger. DBS spectra for Fe-MFI-50 and Fe-MFI-100 are shown in Fig. 2. The DB energy distributions are similar for all the four measured microporous samples. Preparation and Mössbauer analysis of MFI samples are described in detail (Fejes et al., 2003). Samples were exposed to deferration in order to remove extra-framework iron. The removal was successful only in part, both framework substituted and iron located in extra-framework sites could be distinguished in the corresponding Mössbauer spectra collected after evacuation (Fejes et al., 2003). In general, the post
Fig. 2. Doppler broadening spectra six samples by the conventional method.
51
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Fe-SBA-15
Before comparison of particular data it should be noticed in general that lifetime data shown in Fig. 1 and in Table 2 are extracted from 2γ annihilation data, originated mostly from annihilations proceeding after of o-Ps = > p-Ps conversions. However annihilation of o-Ps may proceed directly, too, in 3γ annihilations with arbitrary selection of energies and angles among the 3 γ quanta. Thus, the intensity data shown in Fig. 4 and Table 2 are inherently not correct in a strict sense since 3γ events are omitted (Kajcsos et al., 2005). As for the interpretation of lifetime data obtained on microporous Fe-FER and Fe-MFI substances, the particular pore and channel structure of the FER and MFI frameworks should be considered. The two structures are similar, basically composed from two intercrossing channel systems. The FER frameworks are dominated by an interpenetrating 10 member (-Si-O-) ring channel (with ca. 0.48 nm diameter), and another 8 member ring channel system. The MFI framework is slightly more spacious, both the intercrossing channels are composed from ten member (-Si-O-) rings – with one straight channel and with another zig-zag one (Baerlocher et al., 2001). Smaller cages and pores are also present in these structures, they are constructed from combinations of 5, 6 (-Si-O-) member frames. Results of earlier measurements can also be considered for evaluation of our recent PALS data. On pure MFI silicalites (without iron) similar data were obtained for τ3, e.g. 2.1 ns in average. A certain mean composed from τ4 (6.5 ns) and τ5 (34 ns) values given (Kajcsos et al., 2005), where PALS spectra were decomposed to five components, may correspond to our recent various 13 < τ4 < 21 ns values shown in Table 2. τ3 can probably be attributed to the lifetimes observed in the small 5, 6 (-Si-O-) member cages and voids, whereas τ4 is probably originated from o-Ps bouncing in the elongated channel systems. Shortening of τ4 lifetimes in FER (13.5 ns) in comparison to MFI (with 15.4 and 20.5 ns values) can probably be attributed to the difference in one of the channel diameters. FER is composed from an 8 and a 10 member ring channel system, whereas the MFI framework is slightly more spacious composed from two 10 member ring channels. Data of more recent measurements on MFI (Zubiaga et al., 2016) are also in correspondence with our presented interpretation. Mesoporous SBA-15 samples provide distinctly different PALS spectra as compared the microporous FER and MFI samples. The average channel diameters for SBA-15 and Fe-SBA-15 are 5.7 and 8.6 nm-s respectively as mentioned in the Experimental part. These values are larger with one order of magnitude than characteristic for the microporous systems. The observed longest τ4 lifetimes (35–50 ns) may correspond to 2–6 nm pore sizes by using the modified Tao-Eldrup approximation (Dull et al., 2001). The short τ3 lifetimes (3.4 – 3.9 ns) may indicate certain presence of smaller voids. However, the existence of mesoporous channels is dominant, the relative intensities for τ4 exceed more than one order of magnitude the values related to τ3. Data published on MCM-41 silicates (Surowiec et al., 2010) can be considered to compare our o-Ps lifetime values with data published on similar mesoporous silicates. Channel diameters in these MCM-41 systems were slightly smaller (3 – 4 nm), the corresponding τ3 and τ4 lifetimes were similar to our ones, 4 and 41 ns, respectively, (Surowiec et al., 2010). S and W parameters extracted from the Doppler measurements are shown in Fig. 5. The pore structure difference is clearly reflected in the S-W plot, the increase of channel diameters is connected to increase of the S parameter with simultaneous decrease of related W parameters.
as rec.
vac. 650 K
H2, 620 K
-4
-2
0 2 velocity, mm/s
4
Fig. 3. In situ Mössbauer spectra of Fe-SBA-15 after various treatments (as rec: as received, vac. 650: evacuation at 650 K for 2 h, H2 620 K: treatment in hydrogen at 620 K, for 2 h). Green: Fe2+, blue: tetrahedrally coordinated Fe3+, brown: octahedrally coordinated Fe3+.
3.2.2. Effects of insertion of iron into the framework In advance it can be noticed that most of the iron in our samples was isomorphically substituted into tetrahedral framework sites, as Mössbauer studies revealed for all the five iron containing samples. Thus most of ions was evenly distributed in the walls of pores and channels. A minor portion of iron was probably removed from these framework positions to extra-framework sites as slight appearance of Fe2+ component upon autoreduction taking place at evacuation at
Fig. 4. Change of lifetime and intensity of the 3rd and 4th components of the samples.
52
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
The results obtained on microporous FER and MFI samples reveal a certain dependence of lifetimes on the various pore and channel sizes. These results are in correspondence with those obtained in earlier studies performed on samples without iron. Thus, in the first approximation addition of iron does not influence significantly the observed lifetimes in microporous systems. However, a slight decrease of lifetimes can be observed in the MFI samples. The effect of addition of iron in mesoporous SBA-15 system is more expressed, more significant decrease of the corresponding life time can be detected. Simultaneously, addition of iron has an influence on the S parameter of Doppler broadening spectra, too. Namely the S parameter decreases considerably upon addition of iron both in mesoporous SBA-15 and in microporous MFI. Acknowledgements
Fig. 5. W parameters vs. S parameters for six samples of two type zeolite.
elevated indicated. This amount was small (less than c.a 20% of total iron), and this removal from framework positions did not resulted in appearance of separate iron oxide (hematite, magnetite, etc.) phases. Thus, noticeable filling of pores with agglomerated extra framework moieties probably does not take place in a significant extent in our samples of low Fe contents. Comparison of PALS and DBS data obtained on the first set of samples, Fe-FER and (Fe,Al)-FER reveals that the increase of iron content by four times (Table 1), does not result in any significant difference neither in the values of lifetimes, nor in values of S and W parameters. Thus the expected quenching effect of the iron cannot be demonstrated on our Fe-FER and (Al,Fe)-FER samples. In contrast, the expectation is fulfilled in a certain extent on our next set of MFI samples. The expected increase of the lifetimes with the decrease of the iron content is clearly manifested in comparison of both τ3 and τ4 lifetimes. Namely, the former one increases from 1.8 to 2.8, the latter one from 15 to 20 ns, respectively upon cutting the iron content in half. The effect of decrease of iron content in one of the DBS parameters, namely in the noticeable increase of S parameter is also reflected. Considering our mesoporous SBA-15 samples the effect of addition of iron is significant. Substitution of 1.2 w% iron into the SBA-15 structure resulted in significant decrease of lifetime of o-Ps experienced in the 6 – 8 nm channels, τ4 dropped from 51 ns to 35 ns. Simultaneously, value of S parameter depends also noticeably on the iron content as Fig. 5 attests. The response of S parameters to the increase of iron content is in accordance with the expectation. Namely S parameter is more closely related to the density of electrons close to the valency level, where the densities can be influenced by addition of iron. W parameter is related to levels of electrons occupying deeper levels with more limited accessibility, thus, the latter parameter is less sensitive to addition of iron. For comparison with literature data, it can be mentioned that expressed decrease of lifetimes was reported upon addition of iron into mesoporous MCM-41 structures. However the amounts of iron in these Fe-MCM-41 samples exceeded significantly the iron content present in our Fe-SBA-15 sample (Wiertel et al., 2013). Before concluding, it should be noted here that, for FER and MFI samples, the intensities (I4) of the longest lifetime components are very small. This should be further researched.
This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under Grant number 103.04–2013.11 and was conducted on the Memorandum of Understanding for scientific co-operation between the Centre for Nuclear Techniques, Vietnam and the Wigner Research Centre for Physics, Hungarian Academy of Sciences. The authors are indebted to late Zsolt Kajcsos, who provided substantial help in the early stages. Technical assistance of Csilla Bogdán in preparation of samples is also appreciated. References Baerlocher, Ch, Meier, W.M., Olson, D.H., 2001. Atlas of Zeolite Framework Types. Elsevier. Beck, A., Horváth, A., Stefler, Gy, Katona, R., Geszti, O., Tolnai, Gy, Liotta, L.F., Guczi, L., 2008. Formation and structure of Au/TiO2 and Au/CeO2 nanostructures in mesoporous SBA-15. Catal. Today 139, 180–187. Bosnar, S., Kosanovic, C., Subotic, B., Bosnar, D., Kajcsos, Zs, Liszkay, L., Lohonyai, L., Molnár, B., Lázár, K., 2007. On the potential of positron lifetime spectroscopy for the study of early stages of zeolites formation from their amorphous precursors. Radiat. Phys. Chem. 76, 252–256. Dull, T.L., Frieze, W.E., Gidley, D.W., Sun, J.N., Yee, A.F., 2001. Determination of pore size in mesoporous thin films from the annihilation lifetime of positronium. J. Phys. Chem. B 105, 4657–4662. Dung, T.Q., Lázár, K., Havancsák, K., Kajcsos, Zs, 2013. o-Ps lifetimes in iron containing micro- and mesoporous media. Mater. Sci. Forum 733, 197–202. Fejes, P., Kiricsi, I., Lázár, K., Marsi, I., Rockenbauer, A., Korecz, L., Nagy, J.B., Aiello, R., Testa, F., 2003. Attempts to produce uniform Fe(III) siting in Fe-content ZSM-5 zeolites Determination of framework/extra-framework ratio of Fe(III) in zeolites by EPR and Mössbauer spectroscopy. Appl. Catal., A: General. 242, 247–266. Gidley, D.W., Peng, H.-G., Vallery, R.S., 2006. Positron annihilation as a method to characterize porous materials. Annu. Rev. Mater. Res. 36, 49–79. Goworek, T., 2014. Positronium as a probe of small free volumes in crystals, polymers and porous media. Ann. Univ. Mariae Curie-Skłodowska 69, 1–110. Jean, Y.C., Mallon, P.E., Schrader, D.M. (Eds.), 2003. Principles and Applications of Positron & Positronium Chemistry. World Scientific. Kajcsos, Zs, Lázár, K., Brauer, G., 1993. Ion-exchange with Pd/Pt and Fe and their reduction to metallic state in zeolites: positron annihilation and Mössbauer studies. J. De. Phys. IV 3, 197–200. Kajcsos, Zs, Liszkay, L., Duplatre, G., Varga, L., Lohonyai, L., Pászti, F., Szilágyi, E., Lázár, K., Kótai, E., Pál-Borbély, G., Beyer, H.K., Caullet, P., Patarin, J., Azenha, M.E., Gordo, P.M., Lopes Gil, C., de Lima, A.P., Ferreira Margues, M.F., 2005. Positronium trapping in porous solids: means and limitations for structural studies. Acta Phys. Pol. A 107, 729–737. Kajcsos, Zs, Liszkay, L., Duplátre, G., Lázár, K., Lohonyai, L., Varga, L., Gordo, P.M., de Lima, A.P., Lopes de Gil, C., Ferreira Marques, M.F., Bosnar, D., Bosnar, S., Kosanovic, C., Subotic, B., 2007. Competitive positron and positronium trapping in porous media. Radiat. Phys. Chem. 76, 231–236. Kajcsos, Zs, Liszkay, L., Duplatre, G., Varga, L., Lohonyai, L., Kosanovic, C., Bosnar, S., Subotic, B., Lázár, K., Bosnar, D., Havancsák, K., Gordo, P.M., 2009. Critical parameters of positron and positronium annihilation in grainy and porous solids: ionic crystals and zeolites. Phys. Status Solidi C 6, 2540–2545. Kansy, J., 1996. Microcomputer program for analysis of positron annihilation lifetime spectra. Nucl. Instrum. Methods, Phys. Res. A374, 235–244. Lázár, K., Matusek, K., Mink, J., Dobos, S., Guczi, L., Vizi-Orosz, A., Markó, L., Reiff, W.M., 1984. Spectroscopic and catalytic study on metal carbonyl clusters supported on Cab-O-Sil. I. Impregnation and decomposition of Fe3(CO)12. J. Catal. 87, 163–178. Lázár, K., Lejeune, G., Ahedi, R.K., Shevade, S.S., Kotasthane, A.N., 1998. Interpreting the oxidative catalytic activity in iron-substituted ferrierites using in situ Mössbauer spectroscopy. J. Phys. Chem. B 102, 4865–4870.
4. Conclusions Positronium lifetime and Doppler broadening measurements have been performed on microporous FER and MFI structures containing framework subtituted iron in different amounts and on mesoporous SBA-15 without and with iron introduced to the pore walls. The measurements have been complemented with Mössbuer studies to prove the extent of isomorphous incorporation of iron into the different structures. The amount of iron added to the structures was modest or less (Si/Fe ratio varied from 16 to 100). 53
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Süvegh, K., Vankó, G., Domján, A., Vértes, A., 2001. Temperature dependent electron acceptor properties of zeolite Y: a combined Mössbauer and positron lifetime spectroscopy study. Mater. Sci. Forum Vols. 363–365, 266–268. Wiertel, M., Surowiec, Z., Budzyński, M., Gac, W., 2013. Positron annihilation studies of mesoporous iron modified MCM-41 silica. Nukleonika 58, 245–250. Wiertel, M., Surowiec, Z., Gac, W., Budzynski, M., 2014. Positron annihilation in MnFe2O4/MCM-41 nanocomposite. Acta Phys. Pol. A 125, 793–797. Zaleski, R., Wawryszczuk, J., Borowka, A., Goworek, J., Goworek, T., 2003. Temperature changes of the template structure in MCM-41 type materials; positron annihilation studies. Microporous Mesoporous Mater. 62, 47–60. Zubiaga, A., Warringham, R., Boltz, M., Cooke, D., Crivelli, P., Gidley, D., Perez-Ramirez, J., Mitchell, S., 2016. The assessment of pore connectivity in hierarchical zeolites using positron annihilation lifetime spectroscopy: instrumental and morphological aspects. Phys. Chem. Chem. Phys. 18, 9211.
Lázár, K., Calleja, G., Melero, J.A., Martinez, F., Molina, R., 2004. Influence of synthesis routes on the state of Fe-species in SBA-15 mesoporous materials. Stud. Surf. Sci. Catal. 154, 805–812. Liszkay, L., Kajcsos, Zs, Duplátre, G., Lázár, K., Pál-Borbély, G., Beyer, H.K., 2001. Positronium interactions in zeolites: effect of adsorbed water. Mater. Sci. Forum 363–365, 377–379. Cabral-Prieto, A., Garcia-Sosa, I., Jimenez-Becerril, J., Lopez-Castanares, R., OleaCardoso, O., 2006. Positron annihilation in Co2+-exchanged zeolite LTA. Microporous Mesoporous Mater. 93, 199–204. Schacht, S., Janicke, M., Schüth, F., 1998. Modeling X-ray patterns and TEM images of MCM-41. Microporous Mesoporous Mater. 22, 485–493. Surowiec, Z., Wiertel, M., Zaleski, R., Budzyński, M., Goworek, J., 2010. Positron annihilation study of iron oxide nanoparticles in mesoporous silica, MCM-41 template. Nukleonika 55, 91–96.
54
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Positronium lifetime measurements in micro- and mesoporous ferrisilicates a,⁎
Trần Quốc Dũng , Károly Lázár a b
MARK
b
Center for Nuclear Techniques, 217 Nguyen Trai, District 1, Ho Chi Minh City, Viet Nam Centre for Energy Research HAS, H-1525 Budapest, P.O.B. 77, Hungary
A R T I C L E I N F O
A B S T R A C T
Keywords: Positron Annihilation Trapped o-Ps Porous Solids Zeolites Iron ion-exchanged zeolite
Positron annihilation lifetime, Doppler broadening measurements have been used to study microporous FER and MFI structures containing framework subtituted iron in different amounts. Mesoporous SBA-15 samples without and with iron introduced to the pore walls were also investigated by these methods. Mössbauer measurements have been accompanied to prove the extent of isomorphous incorporation of iron into the different structures. The results reveal a certain dependence of lifetimes on the various pore and channel sizes. Addition of iron does not influence significantly the observed lifetimes in FER microporous systems. However a slight decrease of lifetimes can be observed in the MFI samples. The effect of addition of iron in mesoporous SBA-15 system is more expressed. Supplement of iron also has an influence on the S parameter of Doppler broadening spectra.
1. Introduction Porous structures can advantageously be studied with positronium annihilation measurements. Annihilation of para positronium is fast (125 ps), in contrast to the lifetime of ortho positronium (o-Ps) which is larger with three orders of magnitude (140 ns). If o-Ps interacts with electrons of the environment (“pick-off” annihilation) the observed lifetime becomes shorter. If o-Ps positronium is trapped inside porous structures the observed lifetime can be correlated with the size of cages and channels (Gidley et al., 2006; Jean et al., 2003). The 10–130 ns lifetime interval corresponds to 2–100 nm pore size range by the extended Tao-Eldrup approximation ( Dull et al., 2001). Beside the commonly used lifetime measurements, energy distribution around the 511 keV annihilation peak can also be measured by Doppler broadening method. Small variations in the signal shape may provide information on the interactions of o-Ps with valency and core electrons of the host matrix via the S (shape) and W (wing) parameters, respectively (Jean et al., 2003; Liszkay et al., 2001). Usually relative tendencies are analyzed in S - W plots since selection of the considered ranges in Doppler spectra depends strongly on the experimental conditions (Kajcsos et al., 2009). Pore sizes of microporous silicates (zeolites) and mesoporous substances are in the mentioned size range. Typical microporous structures (MFI and FER) are strictly crystalline, with interpenetrating channel systems with diameters of 0.45–0.55 nm (Baerlocher, 2001). Characteristic features of mesoporous MCM-41 and SBA-15 structures are their larger channel diameter (4–6 nm) and their partly amorphous pore wall structure (Schacht et al., 1998). Both of these types of ⁎
materials have achieved broad variety of applications, thus they have been widely characterized by o-Ps measurements, too (Goworek, 2014; Kajcsos et al., 2005). Combination of micro - and mesoporous strucures, namely those with so-called hierarchical porosity can also be studied (Zubiaga et al., 2016). Studies on mesoporous substances (e.g. MCM41) with positronium lifetime measurements are also reported (Zaleski et al., 2003). Beside pure silica and alumina based porous structures, further ones containing additional transition metal ions substituted isomorphically into the framework are also widely used in the practice. Addition of the transient metal ion modifies the electron densities in the cages, thereby it may influence the lifetime of o-Ps positronium, too (Süvegh et al., 2001). Further on adsorption of gas molecules (nitrogen, oxygen and even water from humidity of air) takes easily place in cages of microand mesoporous substances, modifying thereby the actual size of cages (Cabral-Prieto et al., 2006; Kajcsos et al., 2007). Thus these measurements should usually be performed after evacuation. Iron modified MCM-41 with iron in the pore walls (Wiertel et al., 2013) as well as with Fe2O3 (Surowiec et al., 2010) or with mixed oxide Mn-ferrites (Wiertel et al., 2014) in the channels have also been characterized with the method. Iron containing zeolites and mesoporous substances can advantageously be studied by another technique, with Mössbauer spectroscopy, too. This method provides information solely on the close environment of iron. Combination of the two techniques may strenghten the interpretation of data (Zaleski et al., 2003; Wiertel et al., 2013; Kajcsos et al., 1993). In our previous publications, general aspects of studying porous
Corresponding author. E-mail addresses: [email protected] (T.Q. Dũng), [email protected] (K. Lázár).
http://dx.doi.org/10.1016/j.radphyschem.2017.05.012 Received 9 February 2017; Received in revised form 24 April 2017; Accepted 12 May 2017 Available online 17 May 2017 0969-806X/ © 2017 Elsevier Ltd. All rights reserved.
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
media with positronium studies were discussed (Cabral-Prieto et al., 2006). Further on sorption and removal of adsorbed water in zeolites (Liszkay et al., 2001), and early stages of crystallization of zeolites (Bosnar et al., 2007) were also followed by the method. Preliminary results of positronium lifetime measurements on comparison of microporous FER, MFI with various iron contents and mesoporous SBA-15 plain silicate with its iron modified ferrisilicate counterpart have been reported recently (Dung et al., 2013). The microporous structures exhibit 2- 20 ns lifetimes for o-Ps, whereas 4 – 52 ns lifetimes were found for mesoporous SBA-15 in correspondence with the expectations of the extended Tao-Eldrup model. Addition of iron decreased the lifetimes in Fe-MFI and in Fe-SBA in comparison to the corresponding plain porous silicates. Iron species in the microporous structures were identified by Mössbauer spectroscopy as well (Dung et al., 2013). The present report accounts on further studies on various iron modified microporous (Fe)-FER and (Fe)-MFI as well as on mesoporus plain and iron modified SBA-15. In particular, lifetime measurements are completed with determinations of S and W parameters from Doppler measurements, and a broader scope in interpretation is also provided. In situ Mössbauer spectra are also shown to identify the species of iron present in the mesoporous Fe-SBA.
Table 1 List of samples. Type
Sample
Si/ Fe (at)
Si / Al (at)
Ref.
Microporous (A)
A-Fe-FER A- (Al,Fe)-FER A-Fe-MFI-50 A-Fe-MFI-100 B-Fe-SBA−15 B-SBA−15
16 64 50 100 65 –
24 200 200 – –
[20] [20] [21] [21] [23] [22]
Mesoporous (B)
acterization are presented (Lázár et al., 2004). List of samples is shown in Table 1. 2.2. Mesurements 2.2.1. Lifetime (PALS) and Doppler (DBS) positronium mesurements In all PAS measurements the radioactive positron source has been sandwiched between two disk-shaped pieces of samples. Na-22 sources were used for both DBS and PALS measurements. The zeolite powder was pressed (0.3 GPa) into disks of 8–13 mm diameter and 1–1.5 mm thickness. In the PALS studies the samples were processed and prepared sandwiching the Na-22 source wrapped in Al foil. Samples and source were placed in a sample holder in an Al vacuum chamber. Adsorbed water was removed from the pores by a combined evacuation and heat treatment before the measurements performed at room temperature. PALS mesurements with low-activity (0.5–2 MBq) Na-22 source were performed. A fast-fast coincidence PALS was equipped with XP 2020 URQ photomultipliers and BaF2 scintillators and NE111 plastic scintillator. Wide energy selection windows were set to enable high efficiency, still providing good time resolution (250 ps FWHM for 60Co). The spectra were obtained by employing time-to-pulse height converter units, in case of simultaneous use of two of them allowing also the setting of different time ranges simultaneously. The data collection chains on the PALS consisted of standard ORTEC and TENNELEC units. The data collection took place with computer-based multichannel cards - the 16k Oxford Microfast products. The time calibration values for the various settings varied from 20.1 ps/ch up to 200.3 ps/ch, respectively. In each spectrum integral counts number at 2.106 were recorded. The time base of the PALS setup was 8192 channels. A good reproducibility of the results was found by recording spectra over several days and repeating the experiments. The lifetimes and relative intensities were also evaluated using the LT v.9 fitting program (Kansy, 1996). All PALS spectra were decomposed to four lifetime components. In DBS technique the 511 keV annihilation energy, which is broadened by the Doppler shift by the total momentum of the positron– electron annihilation pair, is directly measured. The S parameter, which corresponds to positron annihilation with the valence electrons, is defined as the ratio of the central area of the 511 keV annihilation gamma ray peak to the total peak area. The W parameter, which corresponds to positron annihilation with the core electrons, is the ratio of the edge area of the DB peak to the total area electrons. In this DBS study the apparatus is typical and energies in the range of 40 keV–1.4 MeV were recorded by using a Canberra HPGe detector with 1.86 keV FWHM resolution at the 1.28 MeV line of positron source. A low-activity (0.5–2 MBq) Na-22 source were used. Each spectrum carried about 6 million counts.
2. Experimental 2.1. Samples 2.1.1. Microporous (type A) samples Fe-FER samples with Si/T atomic ratio of 16 have been synthesized. T stands for the modifying ion substituted isomorphously to tetrahedral sites into the siliceous framework. In the first case, for Fe-FER sample, solely Fe was used, resulting in an average composition of FeSi16O34. In the second case, for (Al,Fe)-FER sample, both Al and Fe (in Al:Fe =3 ratio) were introduced into the tetrahedral framework sites (FeAl3Si64O136). Thus, the amount of iron was only ¼th in the latter sample in comparison to the first Fe-FER. Further details of the synthesis and characterization can be found (Lázár et al., 1998). The extent of isomorphous substitution with Fe and Al ions located in tetrahedral sites was significanly smaller for Fe-MFI samples than for the previous Fe-FER ones, Si/T exceeded 40 in each cases. Si/Fe ratios are 50 and 100 for samples Fe-MFI-50 and 100, respectively. In other words, iron content was doubled in Fe-MFI-50 as compared to Fe-MFI100. The dominant substituting ion was Fe, with addition of Al in 1/3 portion. Thus the average nominal compositons were AlFe2Si100O206 and AlFe2Si200O406 respectively. To provide the possibility for obtaining Mössbauer spectra at these low iron contents 57Fe isotope was used for sample preparation. Futher details of preparation and characterization are reported in (Fejes et al., 2003). 2.1.2. Mesoporous SBA-15 (type B) samples SBA-15 was prepared by using mixture of P123 block copolymer, water, acidified with HCl. The siliceous pore structure was formed from gelation with tetra-ethyl-orthosilicate with subsequent hydrothermal treatment in an autoclave at 100 °C for 24 h, then filtered and dried at room temperature. The template was removed by calcination at 450 °C in oxygen for 5 h. The specific surface area of the product was 580 m2g−1, the average channel diameter was 5.7 nm as determined from BET measurements. Further details of preparation and characterization are described in the publications (Beck et al., 2008). Fe-SBA-15 was prepared with Si/Fe ~65 ratio (approximate nominal composition FeSi65O132). The preparation was similar to that applied for the previous plain SBA-15 with addition of iron into the starting synthesis mixture. Namely, P123 block copolymer was dissolved in 1.9 NHCl, then FeCl3·6H2O was added to the solution. Further stages were similar to that applied for the synthesis of plain SBA-15. The specific surface area of the Fe-SBA-15 was 715 m2 g−1, the average channel diameter was 8.6 nm. Further details of synthesis and char-
2.2.2. Mössbauer measurements Mössbauer measurements were performed by using a spectrometer using 1 GBq 57Co source. Fe-FER, Fe-MFI and Fe-SBA samples were treated at different conditions (evacuation at 650 K, reduction in hydrogen at 620 K) in an in situ Mössbauer cell (Lázár et al., 1984). Most of spectra were collected at room temperature. Parameters were extracted by decomposing the spectra to components of Lorentzian line 50
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Table 2 O-Ps lifetimes obtained from decomposition of PALS spectra (numbers in parentheses show errors).
Microporous
Sample
τ3 ns
τ4 ns
I3 %
I4 %
A-Fe-FER
1.927 (0.024) 1.736 (0.041) 1.851 (0.020) 2.794 (0.046)
13.460 (0.350) 13.520 (0.500) 15.420 (0.300) 20.530 (0.730)
1.66 (0.04) 1.73 (0.05) 1.31 (0.02) 0.93 (0.04)
0.46 (0.01) 0.47 (0.01) 0.47 (0.01) 0.46 (0.01)
3.900 (0.130) 3.434 (0.074)
35.320 (0.190) 51.186 (0.098)
0.77 (0.05) 1.37 (0.02)
3.93 (0.24) 21.79 (0.29)
A-(Al,Fe)FER A-FeMFI−50 A-FeMFI−100 Mesoporous
B-FeSBA−15 B-SBA−15
Fig. 1. PALS spectra recorded for mesoporous SBA-15 and Fe-SBA-15. All spectrum are collected in 8 K channels, here illustrated in 3000 channels only.
shape. Data are referred to metallic alpha-iron. Accuracy of positional parameters (isomer shift and quadrupole splitting) is ± 0.03 mm/s.
syntheses treatments probably facilitated the removal of most of extraframework species. Considering the high Si/Fe ratio (i.e. the low iron content) most of pores and channels were probably open and accessible for o-Ps.
3. Results and discussion 3.1. Results
3.1.3. SBA-15 and Fe-SBA-15 samples PALS spectra recorded for mesoporous SBA-15 and Fe-SBA-15 are shown in Fig. 1. Spectra are distinctly different from those of the microporous samples. The deconvolution reveals that longer lifetimes of o-Ps are present. The increase is noticeable in τ3 and expressed in τ4, with 35 and 51 ns τ4 lifetimes for Fe-SBA-15 and SBA-15, respectively (see Table 2). Results for DBS measurements are shown in Fig. 2. It is clearly seen that the shapes of both DBS curves (in particular that recorded on ironfree SBA-15) are narrow, they are distinctly different from those recorded on microporous FER and MFI. Decomposition of in situ Mössbauer spectra recorded on Fe-SBA-15 (with 1.2 wt% Fe) is shown in Fig. 3. Mössbauer spectra of the sample recorded after evacuation demonstrates the incorporation of iron in ionic dispersion to the pore walls. Only a small part of iron is present as Fe2+, most of it retains the Fe3+ oxidation state in octahedral and tetrahedral coordinations. This assignments are in correspondence with results of other observations. For the interpretation the modest crystallinity of the pore walls in mesoporous silicates should also be considered (Lázár et al., 2004).
3.1.1. Fe-FER and (Fe,Al)-FER samples Data extracted for the o-Ps components are shown in Table 2 (neglecting the components characteristic for p-Ps with lifetimes less than 1 ns). o-Ps lifetimes in Fe-FER and (Fe,Al)-FER are rather short, both τ3 are less than 2 ns, and τ4 parameters are the smallest ones among the data in the table. DBS spectra for Fe-FER and (Fe,Al)-FER samples are shown in Fig. 2. The DB energy distributions obtained for both Fe-FER and (Fe,Al)-FER adhere closely to the data obtained for the other (MFI) microporous samples. Mössbauer spectra for Fe-FER and (Fe,Al)-FER sample are presented in our previous publication (Dung et al., 2013). Their interpretation is shortly amended here as follow. Spectra of Fe-FER display dominant presence of framework substituted iron on tetrahedral positions, i.e. after evacuation at 620 K most of the channels and pores is empty. However, later treatments (e.g. reduction in hydrogen at 620 K) may result in relocation of iron from framework to extra-framework positions in a certain extent. Only evacuation was applied before the PALS measurements, thus significant presence of extra-framework iron should not be expected for this sample. Mössbauer spectra for the (Fe,Al)-FER sample exhibit slightly different features. Evacuation results in slightly increased extent from framework to extra-framework relocation connected to auto-reduction, as can be seen in the modest appearance of Fe2+. Thus, the (Fe,Al)-FER sample is less stable, the channels and pores can be contaminated in a certain extent with various FeO+, AlO+ etc. extra-framework fragments, too (Lázár et al., 1998).
3.2. Discussion 3.2.1. Comparison of samples The lifetime and intensity data are compared in Fig. 4.
3.1.2. Fe-MFI-50 and Fe-MFI-100 samples Data extracted for the o-Ps components (τ3, τ4) are shown in Table 2. τ3 lifetimes are silmilar or slightly exceed those obtained for the previous Fe-FER samples, whereas τ4 data are significantly larger. DBS spectra for Fe-MFI-50 and Fe-MFI-100 are shown in Fig. 2. The DB energy distributions are similar for all the four measured microporous samples. Preparation and Mössbauer analysis of MFI samples are described in detail (Fejes et al., 2003). Samples were exposed to deferration in order to remove extra-framework iron. The removal was successful only in part, both framework substituted and iron located in extra-framework sites could be distinguished in the corresponding Mössbauer spectra collected after evacuation (Fejes et al., 2003). In general, the post
Fig. 2. Doppler broadening spectra six samples by the conventional method.
51
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Fe-SBA-15
Before comparison of particular data it should be noticed in general that lifetime data shown in Fig. 1 and in Table 2 are extracted from 2γ annihilation data, originated mostly from annihilations proceeding after of o-Ps = > p-Ps conversions. However annihilation of o-Ps may proceed directly, too, in 3γ annihilations with arbitrary selection of energies and angles among the 3 γ quanta. Thus, the intensity data shown in Fig. 4 and Table 2 are inherently not correct in a strict sense since 3γ events are omitted (Kajcsos et al., 2005). As for the interpretation of lifetime data obtained on microporous Fe-FER and Fe-MFI substances, the particular pore and channel structure of the FER and MFI frameworks should be considered. The two structures are similar, basically composed from two intercrossing channel systems. The FER frameworks are dominated by an interpenetrating 10 member (-Si-O-) ring channel (with ca. 0.48 nm diameter), and another 8 member ring channel system. The MFI framework is slightly more spacious, both the intercrossing channels are composed from ten member (-Si-O-) rings – with one straight channel and with another zig-zag one (Baerlocher et al., 2001). Smaller cages and pores are also present in these structures, they are constructed from combinations of 5, 6 (-Si-O-) member frames. Results of earlier measurements can also be considered for evaluation of our recent PALS data. On pure MFI silicalites (without iron) similar data were obtained for τ3, e.g. 2.1 ns in average. A certain mean composed from τ4 (6.5 ns) and τ5 (34 ns) values given (Kajcsos et al., 2005), where PALS spectra were decomposed to five components, may correspond to our recent various 13 < τ4 < 21 ns values shown in Table 2. τ3 can probably be attributed to the lifetimes observed in the small 5, 6 (-Si-O-) member cages and voids, whereas τ4 is probably originated from o-Ps bouncing in the elongated channel systems. Shortening of τ4 lifetimes in FER (13.5 ns) in comparison to MFI (with 15.4 and 20.5 ns values) can probably be attributed to the difference in one of the channel diameters. FER is composed from an 8 and a 10 member ring channel system, whereas the MFI framework is slightly more spacious composed from two 10 member ring channels. Data of more recent measurements on MFI (Zubiaga et al., 2016) are also in correspondence with our presented interpretation. Mesoporous SBA-15 samples provide distinctly different PALS spectra as compared the microporous FER and MFI samples. The average channel diameters for SBA-15 and Fe-SBA-15 are 5.7 and 8.6 nm-s respectively as mentioned in the Experimental part. These values are larger with one order of magnitude than characteristic for the microporous systems. The observed longest τ4 lifetimes (35–50 ns) may correspond to 2–6 nm pore sizes by using the modified Tao-Eldrup approximation (Dull et al., 2001). The short τ3 lifetimes (3.4 – 3.9 ns) may indicate certain presence of smaller voids. However, the existence of mesoporous channels is dominant, the relative intensities for τ4 exceed more than one order of magnitude the values related to τ3. Data published on MCM-41 silicates (Surowiec et al., 2010) can be considered to compare our o-Ps lifetime values with data published on similar mesoporous silicates. Channel diameters in these MCM-41 systems were slightly smaller (3 – 4 nm), the corresponding τ3 and τ4 lifetimes were similar to our ones, 4 and 41 ns, respectively, (Surowiec et al., 2010). S and W parameters extracted from the Doppler measurements are shown in Fig. 5. The pore structure difference is clearly reflected in the S-W plot, the increase of channel diameters is connected to increase of the S parameter with simultaneous decrease of related W parameters.
as rec.
vac. 650 K
H2, 620 K
-4
-2
0 2 velocity, mm/s
4
Fig. 3. In situ Mössbauer spectra of Fe-SBA-15 after various treatments (as rec: as received, vac. 650: evacuation at 650 K for 2 h, H2 620 K: treatment in hydrogen at 620 K, for 2 h). Green: Fe2+, blue: tetrahedrally coordinated Fe3+, brown: octahedrally coordinated Fe3+.
3.2.2. Effects of insertion of iron into the framework In advance it can be noticed that most of the iron in our samples was isomorphically substituted into tetrahedral framework sites, as Mössbauer studies revealed for all the five iron containing samples. Thus most of ions was evenly distributed in the walls of pores and channels. A minor portion of iron was probably removed from these framework positions to extra-framework sites as slight appearance of Fe2+ component upon autoreduction taking place at evacuation at
Fig. 4. Change of lifetime and intensity of the 3rd and 4th components of the samples.
52
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
The results obtained on microporous FER and MFI samples reveal a certain dependence of lifetimes on the various pore and channel sizes. These results are in correspondence with those obtained in earlier studies performed on samples without iron. Thus, in the first approximation addition of iron does not influence significantly the observed lifetimes in microporous systems. However, a slight decrease of lifetimes can be observed in the MFI samples. The effect of addition of iron in mesoporous SBA-15 system is more expressed, more significant decrease of the corresponding life time can be detected. Simultaneously, addition of iron has an influence on the S parameter of Doppler broadening spectra, too. Namely the S parameter decreases considerably upon addition of iron both in mesoporous SBA-15 and in microporous MFI. Acknowledgements
Fig. 5. W parameters vs. S parameters for six samples of two type zeolite.
elevated indicated. This amount was small (less than c.a 20% of total iron), and this removal from framework positions did not resulted in appearance of separate iron oxide (hematite, magnetite, etc.) phases. Thus, noticeable filling of pores with agglomerated extra framework moieties probably does not take place in a significant extent in our samples of low Fe contents. Comparison of PALS and DBS data obtained on the first set of samples, Fe-FER and (Fe,Al)-FER reveals that the increase of iron content by four times (Table 1), does not result in any significant difference neither in the values of lifetimes, nor in values of S and W parameters. Thus the expected quenching effect of the iron cannot be demonstrated on our Fe-FER and (Al,Fe)-FER samples. In contrast, the expectation is fulfilled in a certain extent on our next set of MFI samples. The expected increase of the lifetimes with the decrease of the iron content is clearly manifested in comparison of both τ3 and τ4 lifetimes. Namely, the former one increases from 1.8 to 2.8, the latter one from 15 to 20 ns, respectively upon cutting the iron content in half. The effect of decrease of iron content in one of the DBS parameters, namely in the noticeable increase of S parameter is also reflected. Considering our mesoporous SBA-15 samples the effect of addition of iron is significant. Substitution of 1.2 w% iron into the SBA-15 structure resulted in significant decrease of lifetime of o-Ps experienced in the 6 – 8 nm channels, τ4 dropped from 51 ns to 35 ns. Simultaneously, value of S parameter depends also noticeably on the iron content as Fig. 5 attests. The response of S parameters to the increase of iron content is in accordance with the expectation. Namely S parameter is more closely related to the density of electrons close to the valency level, where the densities can be influenced by addition of iron. W parameter is related to levels of electrons occupying deeper levels with more limited accessibility, thus, the latter parameter is less sensitive to addition of iron. For comparison with literature data, it can be mentioned that expressed decrease of lifetimes was reported upon addition of iron into mesoporous MCM-41 structures. However the amounts of iron in these Fe-MCM-41 samples exceeded significantly the iron content present in our Fe-SBA-15 sample (Wiertel et al., 2013). Before concluding, it should be noted here that, for FER and MFI samples, the intensities (I4) of the longest lifetime components are very small. This should be further researched.
This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under Grant number 103.04–2013.11 and was conducted on the Memorandum of Understanding for scientific co-operation between the Centre for Nuclear Techniques, Vietnam and the Wigner Research Centre for Physics, Hungarian Academy of Sciences. The authors are indebted to late Zsolt Kajcsos, who provided substantial help in the early stages. Technical assistance of Csilla Bogdán in preparation of samples is also appreciated. References Baerlocher, Ch, Meier, W.M., Olson, D.H., 2001. Atlas of Zeolite Framework Types. Elsevier. Beck, A., Horváth, A., Stefler, Gy, Katona, R., Geszti, O., Tolnai, Gy, Liotta, L.F., Guczi, L., 2008. Formation and structure of Au/TiO2 and Au/CeO2 nanostructures in mesoporous SBA-15. Catal. Today 139, 180–187. Bosnar, S., Kosanovic, C., Subotic, B., Bosnar, D., Kajcsos, Zs, Liszkay, L., Lohonyai, L., Molnár, B., Lázár, K., 2007. On the potential of positron lifetime spectroscopy for the study of early stages of zeolites formation from their amorphous precursors. Radiat. Phys. Chem. 76, 252–256. Dull, T.L., Frieze, W.E., Gidley, D.W., Sun, J.N., Yee, A.F., 2001. Determination of pore size in mesoporous thin films from the annihilation lifetime of positronium. J. Phys. Chem. B 105, 4657–4662. Dung, T.Q., Lázár, K., Havancsák, K., Kajcsos, Zs, 2013. o-Ps lifetimes in iron containing micro- and mesoporous media. Mater. Sci. Forum 733, 197–202. Fejes, P., Kiricsi, I., Lázár, K., Marsi, I., Rockenbauer, A., Korecz, L., Nagy, J.B., Aiello, R., Testa, F., 2003. Attempts to produce uniform Fe(III) siting in Fe-content ZSM-5 zeolites Determination of framework/extra-framework ratio of Fe(III) in zeolites by EPR and Mössbauer spectroscopy. Appl. Catal., A: General. 242, 247–266. Gidley, D.W., Peng, H.-G., Vallery, R.S., 2006. Positron annihilation as a method to characterize porous materials. Annu. Rev. Mater. Res. 36, 49–79. Goworek, T., 2014. Positronium as a probe of small free volumes in crystals, polymers and porous media. Ann. Univ. Mariae Curie-Skłodowska 69, 1–110. Jean, Y.C., Mallon, P.E., Schrader, D.M. (Eds.), 2003. Principles and Applications of Positron & Positronium Chemistry. World Scientific. Kajcsos, Zs, Lázár, K., Brauer, G., 1993. Ion-exchange with Pd/Pt and Fe and their reduction to metallic state in zeolites: positron annihilation and Mössbauer studies. J. De. Phys. IV 3, 197–200. Kajcsos, Zs, Liszkay, L., Duplatre, G., Varga, L., Lohonyai, L., Pászti, F., Szilágyi, E., Lázár, K., Kótai, E., Pál-Borbély, G., Beyer, H.K., Caullet, P., Patarin, J., Azenha, M.E., Gordo, P.M., Lopes Gil, C., de Lima, A.P., Ferreira Margues, M.F., 2005. Positronium trapping in porous solids: means and limitations for structural studies. Acta Phys. Pol. A 107, 729–737. Kajcsos, Zs, Liszkay, L., Duplátre, G., Lázár, K., Lohonyai, L., Varga, L., Gordo, P.M., de Lima, A.P., Lopes de Gil, C., Ferreira Marques, M.F., Bosnar, D., Bosnar, S., Kosanovic, C., Subotic, B., 2007. Competitive positron and positronium trapping in porous media. Radiat. Phys. Chem. 76, 231–236. Kajcsos, Zs, Liszkay, L., Duplatre, G., Varga, L., Lohonyai, L., Kosanovic, C., Bosnar, S., Subotic, B., Lázár, K., Bosnar, D., Havancsák, K., Gordo, P.M., 2009. Critical parameters of positron and positronium annihilation in grainy and porous solids: ionic crystals and zeolites. Phys. Status Solidi C 6, 2540–2545. Kansy, J., 1996. Microcomputer program for analysis of positron annihilation lifetime spectra. Nucl. Instrum. Methods, Phys. Res. A374, 235–244. Lázár, K., Matusek, K., Mink, J., Dobos, S., Guczi, L., Vizi-Orosz, A., Markó, L., Reiff, W.M., 1984. Spectroscopic and catalytic study on metal carbonyl clusters supported on Cab-O-Sil. I. Impregnation and decomposition of Fe3(CO)12. J. Catal. 87, 163–178. Lázár, K., Lejeune, G., Ahedi, R.K., Shevade, S.S., Kotasthane, A.N., 1998. Interpreting the oxidative catalytic activity in iron-substituted ferrierites using in situ Mössbauer spectroscopy. J. Phys. Chem. B 102, 4865–4870.
4. Conclusions Positronium lifetime and Doppler broadening measurements have been performed on microporous FER and MFI structures containing framework subtituted iron in different amounts and on mesoporous SBA-15 without and with iron introduced to the pore walls. The measurements have been complemented with Mössbuer studies to prove the extent of isomorphous incorporation of iron into the different structures. The amount of iron added to the structures was modest or less (Si/Fe ratio varied from 16 to 100). 53
Radiation Physics and Chemistry 139 (2017) 49–54
T.Q. Dũng, K. Lázár
Süvegh, K., Vankó, G., Domján, A., Vértes, A., 2001. Temperature dependent electron acceptor properties of zeolite Y: a combined Mössbauer and positron lifetime spectroscopy study. Mater. Sci. Forum Vols. 363–365, 266–268. Wiertel, M., Surowiec, Z., Budzyński, M., Gac, W., 2013. Positron annihilation studies of mesoporous iron modified MCM-41 silica. Nukleonika 58, 245–250. Wiertel, M., Surowiec, Z., Gac, W., Budzynski, M., 2014. Positron annihilation in MnFe2O4/MCM-41 nanocomposite. Acta Phys. Pol. A 125, 793–797. Zaleski, R., Wawryszczuk, J., Borowka, A., Goworek, J., Goworek, T., 2003. Temperature changes of the template structure in MCM-41 type materials; positron annihilation studies. Microporous Mesoporous Mater. 62, 47–60. Zubiaga, A., Warringham, R., Boltz, M., Cooke, D., Crivelli, P., Gidley, D., Perez-Ramirez, J., Mitchell, S., 2016. The assessment of pore connectivity in hierarchical zeolites using positron annihilation lifetime spectroscopy: instrumental and morphological aspects. Phys. Chem. Chem. Phys. 18, 9211.
Lázár, K., Calleja, G., Melero, J.A., Martinez, F., Molina, R., 2004. Influence of synthesis routes on the state of Fe-species in SBA-15 mesoporous materials. Stud. Surf. Sci. Catal. 154, 805–812. Liszkay, L., Kajcsos, Zs, Duplátre, G., Lázár, K., Pál-Borbély, G., Beyer, H.K., 2001. Positronium interactions in zeolites: effect of adsorbed water. Mater. Sci. Forum 363–365, 377–379. Cabral-Prieto, A., Garcia-Sosa, I., Jimenez-Becerril, J., Lopez-Castanares, R., OleaCardoso, O., 2006. Positron annihilation in Co2+-exchanged zeolite LTA. Microporous Mesoporous Mater. 93, 199–204. Schacht, S., Janicke, M., Schüth, F., 1998. Modeling X-ray patterns and TEM images of MCM-41. Microporous Mesoporous Mater. 22, 485–493. Surowiec, Z., Wiertel, M., Zaleski, R., Budzyński, M., Goworek, J., 2010. Positron annihilation study of iron oxide nanoparticles in mesoporous silica, MCM-41 template. Nukleonika 55, 91–96.
54