- Email: [email protected]
Effect of synthesis temperature on the ordered pore structure in mesoporous silica studied by positron annihilation spectroscopy
Applied Surface Science 363 (2016) 445–450
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Effect of synthesis temperature on the ordered pore structure in mesoporous silica studied by positron annihilation spectroscopy C.Y. Li, N. Qi, Z.W. Liu, B. Zhou, Z.Q. Chen ∗ , Z. Wang ∗ Hubei Nuclear Solid Physics Key Laboratory, Department of Physics, Wuhan University, Wuhan 430072, PR China
a r t i c l e
i n f o
Article history: Received 30 April 2015 Received in revised form 1 December 2015 Accepted 5 December 2015 Available online 12 December 2015 Keywords: Ordered mesoporous silica Synthesis temperature Pore structure N2 adsorption–desorption Positronium
a b s t r a c t Mesoporous silica with ordered pore structure was synthesized at various temperatures using TEOS as the silica source and P123 as the template. Both small angle X-ray scattering (SAXS) and high resolution transition electron microscopy (HRTEM) measurements verify the ordered pore structure of the synthesized SiO2 . With synthesis temperature increasing, the pore structure has slight damage at 130 ◦ C, while it shows ordered structure again at 150 ◦ C. When the synthesis temperature increases to 180 ◦ C, the ordered pore structure is severely destructed. The change of pore structure is further confirmed by scanning electron microscopy (SEM) measurements. Positron lifetime measurements reveal four lifetime components in the synthesized mesoporous SiO2 , and the two long lifetimes 3 and 4 correspond to the annihilation of o-Ps in the micropores and large pores of the material, respectively. The longest lifetime 4 tends to increase slightly with increasing synthesis temperature. However, its intensity I4 shows an overall decrease with exception at 150 ◦ C. At the synthesis temperature of 180 ◦ C, the intensity I4 decreases drastically to about 17.5%. This indicates variation of the size and fraction of pores with increasing synthesis temperature, and the pore structure is seriously destructed at 180 ◦ C. By comparing with the N2 adsorption–desorption measurements, it was found that the Goworek’s model is more suitable for the size estimation of cylindrical pores from the o-Ps lifetime, while Dull’s and Ito’s model is appropriate for the rectangular and spherical pores, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Mesoporous materials, from the natural materials such as diatomaceous earth, natural zeolite and polypite [1–3], to the synthesized porous materials [4–11], have found their wide applications in many fields and attracted more and more attentions. The research areas and related applications are further broadened especially after the appearance of the controllable ordered porous materials, which not only make up the deficiency of microporous zeolite materials [1] for the application in the macromolecules adsorption [4] and catalytic reaction [5], but also can be applied as the microreactor [6] of the controllable nano-materials in chemistry industries, biological technology [7], adsorption isolation [8], environment catalysis, [9], etc. The most commonly used material is ordered mesoporous silica [10–13], such as MCM-41, MCM-48, SBA-15, KIT-6, and SBA-16, which contains different pore size, surface area, pore volume and pore morphology.
∗ Corresponding authors. E-mail addresses: [email protected] (Z.Q. Chen), [email protected] (Z. Wang). http://dx.doi.org/10.1016/j.apsusc.2015.12.055 0169-4332/© 2015 Elsevier B.V. All rights reserved.
Ordered mesoporous silica can be easily synthesized at lower temperatures, even at the room temperature. However, at lower temperatures such as room temperature, large amounts of terminal hydroxyl groups will be irregularly stacked, which might cause the irregular rank of the wall structure of the ordered pore and result in the collapse of the pore structure. Therefore the mesopore in silica possesses a relatively low ordering. Higher synthesis temperature might result in a more stable and ordered mesopore structure. Nevertheless, a much too high temperature will also destroy the ordered pore structure. Therefore an appropriate synthesis temperature should be chosen to attain the balance between pore structure and pore stability. The pore structure such as pore size and their size distribution can be detected by the gas adsorption method [14]. This method is based on the fact that below the saturated vapor pressure of the adsorbate, it condenses to liquid in narrow pores. However, this method is only sensitive to mesopores (2–50 nm), and only open pores can be detected. In a closed-pore material, gases cannot enter the pore, so the adsorption method fails to work. Mercury intrusion porosimeter [15] is another approach to estimate the pore size. However, this porosimeter is sensitive to pores with even larger size. The pores with radius less than 10 nm will escape
446
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
from the detection by this method. Some other methods might be more useful for the study of pore structure in mesoporous materials. Positron annihilation spectroscopy has emerged as a powerful probe for the pore structure [16–21]. In many materials, positron will capture one electron to form a metastable hydrogen-like particle called positronium (Ps) atom. According to the spins of the positron and electron, the Ps atom exists in two states: the triplet spin state ortho-positronium (o-Ps) (s = 1, ms = −1, 0, +1) and the singlet spin state para-positronium (p-Ps) (s = 0, ms = 0). The formation probability of o-Ps is about three times of that of p-Ps. In vacuum, the o-Ps has a long lifetime of 142 ns, and it converts to 3 rays through self-annihilation. The p-Ps has a relatively shorter self-annihilation lifetime of 125 ps by emitting 2 rays. In porous materials, the annihilation of o-Ps depends on the structure of the pore [22]. When o-Ps is trapped in the pore, it will pick one electron from the wall of the pore and then annihilates into 2 rays, which is called pick-off annihilation. For pores with diameter smaller than 1 nm, the pick-off annihilation lifetime of oPs will be reduced to about 1–20 ns, which is closely related to the pore size. The positronium formation probability, which is reflected by the o-Ps intensity, is in accordance with the porosity. So, the formation and annihilation of positronium, especially the o-Ps, could be utilized to study the pore structure. This probe can detect micropores and mesopores, and can detect both closed pores and open pores, which shows advantage over other methods such as the gas adsorption method [14]. In this paper, we prepared a series of porous silica under acidic condition [23–25]. The effect of synthesis temperature on the pore structure was studied by positron annihilation measurements together with SAXS, HRTEM, SEM and N2 adsorption–desorption measurement.
2. Experiment Mesoporous silica was prepared using amphiphilic triblock copolymer Pluronic P123 (Mw = 5800, EO20 –PO70 –EO20 , Sigma Aldrich) as the structure agent and tetraethyl orthosilicate (TEOS, C8 H20 O4 Si) as the silica source [24]. The 37 wt% HCl aqueous solution is diluted to 2 mol/L with distilled water for further use. P123 was first dissolved in a mixture of distilled water and HCl solution, then it was stirred under constant agitation at room temperature for 30 min until P123 is completely dissolved. After that, H2 SO4 and sucrose were added to the mixture and was stirred for another 30 min. Finally, the silica source TEOS was added to the mixture and was stirred until the solution was distributed homogenously. The mole ratio of P123:TEOS:HCl:H2 SO4 :H2 O:sucrose is 1:48.3:649:40:6357:7.2. The mixture was stirred for 72 h at 323 K and then conveyed into the Teflon-lined autoclave under static conditions at the given synthesis temperature(80 ◦ C, 100 ◦ C, 130 ◦ C, 150 ◦ C, 180 ◦ C). The final silica/P123/sucrose gel was filtered, washed with distilled water (and ethanol) and dried at 373 K for 24 h. At the last, the final silica powder was obtained after calcination at 823 K for 6 h in the open air. The synthesized mesoporous silica powders were further hand milled in agate mortar for 2 h and then pressed under a static pressure of 6 MPa for 5 min to pellets with 1.5 mm thickness and a diameter of 15 mm. These pellets were dried over night at 100 ◦ C in the autoclave for other measurements. SAXS measurements were carried out using an X-ray diffractometer (X’Pert Pro, PANalytical, Netherlands) with Cu K˛ radiation operated at 40 kV and 40 mA. The incident X-ray wavelength was 1.5406 A˚ and the scanning angle 2 was from 0.6123 to 10◦ with the step of 0.001◦ . N2 adsorption–desorption measurements were performed at 77 K by using a Micromeritics ASAP
2020 gas-sorption analyzer. The sample was degassed in a vacuum at 180 ◦ C for the desorption process. Pore size distribution and average pore size were estimated from adsorption branch of the isotherms using the Barrett–Joyner–Halenda (BJH) method [26]. At the same time the specific surface area was calculated from the Brunauer–Emmett–Teller (BET) method [27] over the relative pressure P/P0 range of 0.05–0.25. Scanning electron microscope (SEM) (XL 30 (Philips), Amsterdam, Netherlands) and high resolution transmission electron microscopy (HRTEM) (JEOL JEM2010FEF (UHR), Tokyo, Japan) images with acceleration voltage at 200 kV were taken to study the pore morphology and pore ordering. Positron lifetime measurements were performed at room temperature using a conventional fast–fast coincidence lifetime spectrometer with a time resolution of 280 ps in full width at half maximum (FWHM). A 10 Ci 22 NaCl source was sandwiched between two pellet samples and the sample–source–sample sandwich was placed into a vacuum chamber with air pressure lower than 1 × 10−5 torr. The lower level of the energy window on the discriminator for the stop signal (0.511 MeV annihilation -ray) was set to as low as possible in order to collect the 3 annihilation signal of o-Ps. In our experiment, we have tried to measure the positron lifetime spectrum using a time range of 500 ns for the TAC. The resolved long lived lifetime values ( 3 and 4 ) and their intensities are almost the same for the two different time ranges. But time range of TAC of 500 ns leads to the larger uncertainties for the short lifetime components. So in our later measurements we just used the 200 ns range. The total channel number is 4096 and the time scale is 50.3 ps/channel. For each sample, two lifetime spectra were collected with a total count of 1.5 × 106 for each spectrum. The counting rate is about 26 count per second. Doppler broadening spectra of the positron annihilation radiation were measured simultaneously using a high purity Ge (HP-Ge)detector which was placed perpendicular to the direction of positron lifetime measurements. The energy resolution of the HP-Ge detector is about 1.2 keV in full width at half maximum at 511 keV. The S and W parameters are calculated from the spectrum, which represent the ratio of low-energy (|EL | ≤ 0.68 keV) region and high-energy region (2.8 keV ≤ |EL | ≤ 5.73 keV) to the total region of 511 keV annihilation peak, respectively. The total count of each Doppler broadening spectrum is more than 1 × 107 .
3. Results and discussion 3.1. SAXS measurement In order to verify the pore ordering of the synthesized silica, SAXS measurements were performed for the samples synthesized at various temperatures. The measured SAXS patterns are shown in Fig. 1. For most of the samples, the diffraction peaks are nearly the same. There are three main diffraction peaks at around 0.98, 1.64, and 1.85◦ . They are indexed as the (100), (110), and (200) diffraction peaks, which reflect the hexagonal space group (P6mm). This indicates that the porous silica has long-range ordering. In other words, our synthesized silica has ordered structure. It can be seen from Fig. 1 that the positions of the three diffraction peaks have some shifts at different synthesis temperatures. More importantly, the intensities of these peaks have significant change with increasing synthesis temperature. Actually for the sample synthesized at 130 ◦ C, all the three peaks (100), (110), and (200) become weaker, which indicates that the pore ordering is somewhat deteriorated at this temperature. However, for the sample synthesized at 150 ◦ C, the pores show high ordering again. When the synthesis temperature increases to 180 ◦ C, only the peak at 0.98◦ shows weak signal, and the other two peaks disappear completely. This suggests that pore ordering is
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
447
that at 180 ◦ C, the (100) peak is invisible, so the pore width cannot be estimated.
3.2. SEM measurement
Fig. 1. Small-angle X-ray scatting patterns for the synthesized ordered mesoporous silica.
severely damaged. The temperature of 180 ◦ C is probably too high to keep the ordered pore structure. A more suitable synthesis temperature will be 150 ◦ C or even lower. It has been also reported earlier that appropriate synthesis temperature was below 130 ◦ C [22,28–30]. The periodic interval can be estimated by the Bragg’s equation using the position of the main diffraction peak (100): 2d sin = k
(1)
The small angle X-ray scattering measurements can give us the periodic interval of the pores (pore width + pore wall) for the ordered mesoporous silica, which was labeled in Fig. 1. This interval can be used as a rough estimation of the variation of pore size. It increases from 9.10 nm to 10.19 nm at the synthesis temperature at 80–130 ◦ C, and then decrease a little to 9.72 nm at 150 ◦ C. As for
In order to observe the detailed morphology of these samples, SEM images were taken for the silica powder synthesized at different temperatures, which were shown in Fig. 2. The samples reveal worm-like morphology containing particles with irregular shape. The particles are approximately 0.4 m in diameter and 0.5–1.5 m in length. The surface of the particles change evidently with increasing synthesis temperature. The worm-like particles are relatively smooth for the samples synthesized at lower temperatures such as 80 and 100 ◦ C. However, when the synthesis temperature increases to 130 ◦ C, the surface is slightly deconstructed with some cracks. When the temperature increases to 150 ◦ C, the surface of the worm-like particle becomes smooth again. However, as for the sample synthesized at 180 ◦ C, the particle surface is almost full of cracks. SEM measurement reveals the variation of morphology with increasing synthesis temperature.
3.3. HRTEM measurement HRTEM images were taken for the synthesized silica from two different directions (perpendicular and parallel to the pore directions). They were shown in Fig. 3. The samples show a regular pore arrangement. The ordered 2D hexagonal cylindrical pores can be clearly seen. The diameter of the pore is about 10 nm. From the TEM images it can be clearly seen that the pore ordering becomes weaker with increasing synthesis temperature from 80 ◦ C to 130 ◦ C, but the pores show high order again at the relative higher temperature of 150 ◦ C. For the sample synthesized at much higher temperature of 180 ◦ C, the pore ordering is nearly invisible. This observation coincides with the SAXS and SEM measurements. All these results suggest that a relatively high temperature of 150 ◦ C maintains the
Fig. 2. Scanning electron diffraction images measured for the synthesized ordered mesoporous silica.
Fig. 3. High resolution transmission electron diffraction images measured for the synthesized ordered mesoporous silica from perpendicular and parallel to the channel directions.
448
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
Fig. 5. Peak-normalized positron lifetime spectrum for the synthesized ordered mesoporous silica at the synthetic temperature of 80 ◦ C, 130 ◦ C, 180 ◦ C.
radius increases to about 24 nm. For these large pores, the size is already close to the detection limit of N2 adsorption–desorption measurement, which is the reason for the distortion of the adsorption–desorption isotherms [33]. Fig. 4. Nitrogen adsorption–desorption isothermal of ordered mesoporous silica (a) and their relative pore size distribution (b) as a function of synthetic temperature.
well ordered mesoporous structure and might be the appropriate temperature to synthesize stable and ordered mesoporous silica. 3.4. N2 adsorption–desorption measurement In order to get more detailed pore parameters such as pore size, pore volume and surface area, the nitrogen adsorption–desorption measurements were performed at 77 K for the synthesized samples. The adsorption–desorption isotherms measured for each sample were plotted in Fig. 4(a). According to the classification of IUPAC [31], except for the sample synthesized at 180 ◦ C, the nitrogen adsorption–desorption isotherms are in accordance with the defined typical IV curves with H1 hysteresis loops and the clear capillary condensation, which occurs approximately at the relative pressure P/P0 of 0.6–0.9. This ratio is associated with the filling of mesopores by nitrogen, which suggests that the synthesized silica is mesoporous [32]. With the synthesis temperature increasing to 130 ◦ C, the relative pressure P/P0 shifts to a higher value and that indicates increase of the most probable pore radius. For the sample synthesized at 150 ◦ C, the adsorption–desorption isotherms almost overlaps with that of the sample synthesized at 100 ◦ C, and the relative pressure P/P0 shifts to a relatively lower value. However, when the synthesis temperature increases to 180 ◦ C, hysteresis loop of its adsorption–desorption isotherms is not the typical IV curves. The relative pressure P/P0 is approximately 1, which indicates presence of large macropores. Meanwhile, the pore size distribution curves of these five samples were calculated from the adsorption branches by Barrett–Joyner–Halenda (BJH) [26] model. They were plotted in Fig. 4(b). For the samples synthesized at temperatures below 180 ◦ C, they all exhibit relatively narrower pore size distribution. The most probable pore size has a little fluctuation around 10 nm, and the average pore radius is around 7–8 nm. With the synthesis temperature increasing from 80 to 150 ◦ C, the average pore radius first increases from 5.5 nm to 10.5 nm, then it decreases to 8.5 nm. This shows agreement with the results by SAXS measurements. The width of the distribution also increases with temperature increasing from 80 to 130 ◦ C, but it becomes narrower again at 150 ◦ C. For the sample synthesized at 180 ◦ C, of which the ordered pores are nearly invisible, it has a very wide pore size distribution, which starts at 5 nm and extends to 45 nm. The average pore
3.5. Positron annihilation measurements Positron annihilation spectroscopy was used to get more detailed information about the pore structure of synthesized silica. Fig. 5 shows the peak-normalized positron lifetime spectra measured for porous silica synthesized at 80 ◦ C, 130 ◦ C and 180 ◦ C. It is obvious that there is a very long positron lifetime component in the ordered mesoporous SiO2 , indicating that the sample contains large pores. Four lifetime components were derived from the positron annihilation lifetime spectra by PATFIT program [34]. The shortest lifetimes 1 (161 ps) can be ascribed to the annihilation lifetime of free positrons and p-Ps annihilation in silica, while the second lifetime 2 (510 ps) is due to positron annihilation in vacancies or voids. The two longer lifetime 3 (4–8 ns) and 4 (110–120 ns) are obviously o-Ps annihilation lifetime in the pores [35] of the material. In this paper, since our research interest is on the pore structure, we will put our attentions only on the long lifetime components 3 , 4 and their relative intensities I3 and I4 . It is well known that the o-Ps lifetime is closely related to the pore size in materials. The existence of two o-Ps lifetime components indicates that there are two different kinds of pores, which has small and large open volume, respectively. The longest lifetime 4 is apparently due to o-Ps annihilation in the 2D P6mm hexagonal ordered pores in the silica material. The shorter o-Ps lifetime 3 is the o-Ps lifetime in some micropores. However, these micropores are not detected by the N2 adsorption–desorption measurements. They might be the pipes that connect 2D P6mm hexagonal ordered pores. Tao and Eldrup et al. had already established a semi-empirical relationship between the o-Ps lifetime o−Ps and the free volume size R assuming a spherical infinite potential well of radius R0 with an electron layer of thickness R = R − R0 [36–38]:
−1 o−Ps =2 1−
R R 1 sin 2 + 2 R + R R + R
.
(2)
The thickness R is determined to be 1.656 A˚ for most of the molecular materials. While, for silica in our paper, R is estimated as 0.18 A˚ [39–41]. This equation is valid for o-Ps pick-off annihilation in free volume with width smaller than 1 nm [42,43]. For positronium localized in pores with diameter larger than 1 nm, some fraction of the o-Ps will have no chance to pick electron from the wall around the pore, so this part of o-Ps will undergo self-annihilation and emit 3-rays, which leads to a much longer lifetime of about 30–135 ns. Ito et al. have modified the above Tao–Eldrup model by taking into consideration the intrinsic three
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
449
self-annihilation of o-Ps in large pores, and proposed the following model which is also valid for o-Ps annihilation in small micropores less than 1 nm [43].
−1 o−Ps = 2 1−
+
1 Ra sin + 2 Ra + R
2R a Ra + R
1−
R − R b a R + R
1 , 142
(3)
Ra and b are best fitted with experimental data as 0.8 nm and 0.55, respectively [43]. By using the above two equations, we can thus estimate the pore size in our silica sample. For the sample synthesized at 80 ◦ C, the diameter of the small micropores corresponding to 3 is about 0.64 nm, while that of the large pore is about 16.4 nm. However, according to the N2 adsorption–desorption measurements the size of the large pore is only about 5.5 nm in width. This suggests that the model by Ito et al. gives overestimated size of the pore. The reason might be due to the shape of the pores in our sample. Ito et al. still use the spherical voids in the calculation, while our TEM observation indicates that the pores in our silica sample have ordered 2D hexagonal cylindrical geometry, thus the modified model by Ito is not applicable for this kind of pores. Another reason is that Ito’s model is based on previous experimental data. However, some of the data were obtained in air. This will also give overestimation of the pore size from the o-Ps lifetime. Besides the modified model by Ito et al, there are also some other extended Tao–Eldrup models to correlate o-Ps lifetime and pore size for different pore geometry, such as the rectangular tube [44,39], as shown below: (T ) = A −
S − T F(a, ı, T )F(b, ı, T )F(c, ı, T ), 4
where F(x, ı, T ) = 1 −
2ı + x
∞ i=1
(1/(i)) sin((2iı)/x)e((−ˇt
∞
i=1
e((−ˇt
2 )/(x2 kT ))
(4)
2 )/(x2 kT ))
, (5)
and ˇ = h2 /(16 m) = 0.188 eV nm2 . By using the above equation, the calculated pore width corresponding to 4 in silica sample synthesized at 80 ◦ C is about 6.6 nm. However, as for the 2D P6mm hexagonal ordered pores of synthesized silica, the cylindrical model by Goworek shall be much closer. In the revised paper, all the pore widths have been also calculated from the lifetime components by using Goworek’s model, which might be similar to that of the results of adsorption–desorption measurement. Fig. 6 shows the lifetimes 3 , 4 and their intensities I3 and I4 for the silica sample synthesized at different temperatures. With increase of the synthesis temperature, 3 keeps almost constant at about 4 ns, but the lifetime 4 shows irregular change, which assembles the change of pore size revealed by gas adsorption measurements. However, the intensities I3 and I4 change significantly with synthesis temperatures. The intensity I4 decreases gradually from 33.8% at 80 ◦ C to 17.5% at 180 ◦ C, with only one exception at 150 ◦ C, where I4 reaches to 31.6%. The trend of I3 is just opposite to that of I4 . Fig. 7 presents the radius of the pore in silica obtained from the oPs lifetime using the models proposed by Ito, Dull and Goworek. The results obtained from N2 adsorption–desorption measurements are also shown in the figure. As for the 2D P6mm hexagonal ordered pores of synthesized silica, the radius calculated by Ito’s spherical model is obviously overestimated, while that by Dull’s rectangular model will be better. The cylindrical model by Goworek shall be much closer to 2D P6mm hexagonal pore structure. The results from Goworek’s model are in good agreement with that of N2 adsorption–desorption measurements except for the point
Fig. 6. Variation of o-Ps lifetime 4 , 3 and their intensities I4 , I3 for the synthesized ordered mesoporous silica as a function of synthetic temperature.
at temperature of 180 ◦ C. This indicates that the pores have cylindrical tube-liked structure, which has been confirmed by HRTEM observation. As for the sample synthesized at 180 ◦ C, the pore diameter calculated by Ito’s model agrees well with that of N2 adsorption–desorption measurements. This may tell us that pore in silica synthesized at 180 ◦ C shall be sphere-like. However, the discrepancy between the pore size obtained by PALS and other methods for the sample synthesized at 180 ◦ C might be due to the underestimation of the o-Ps lifetime. Besides the large pores, the o-Ps lifetime also gives information on the small micropores. According to the Tao–Eldrup model, the size of micropore changes little with increasing synthesis temperature, but its concentration (reflected by I3 ) increases to more than 10% when the temperature increases to 180 ◦ C. The third lifetime reflects information of small micropores. The quick increase of I3 at 180 ◦ C indicates increase in the number of small micropores. At the same time, the intensity I4 shows abrupt decrease. Therefore we believe that these small micropores might be produced by the collapse of large pores. We also performed Doppler broadening measurements for the SiO2 to study the pore structure. If the positronium is formed in porous materials, one third of the positronium will be p-Ps. Since the p-Ps contains nearly zero momentum, its self-annihilation will contribute a very narrow peak to the Doppler broadening spectrum,
Fig. 7. The pore radius of the synthesized ordered mesoporous silica attained from N2 adsorption–desorption and PAL measurements according to different model as a function of synthetic temperatures.
450
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
Acknowledgment This work was supported by the National Natural Science Foundation of China under Grant Nos. 11475130, 11305117 and 11275143. References
Fig. 8. Variation of S parameters for the synthesized ordered mesoporous silica as a function of synthesis temperature.
which leads to increase in the S parameter. From Fig. 8, we can see that the variation of S parameter is very similar to the change of I4 . Due to the contribution of p-Ps annihilation to the Doppler broadening spectrum, the S parameter is closely related to the Ps intensity. Although in our paper the intensity of o-Ps is the sum of I3 and I4 , Io−Ps is approximately equal to I4 with a minimum contribution from I3 . The p-Ps intensity is one third of o-Ps fraction. So, the comparison of S to I4 is more reasonable. Comparing the total five measurements, each one has its own superiority on exhibiting the different aspects of porous materials. Small-angle X-ray scatting measurement is usually used to verify the degree of ordering, which is the basic measurement of ordered mesoporous pores. SEM measurement is appropriate for the pore morphology. HRTEM measurement is a more directly view to express the pore structure. N2 adsorption–desorption measurements are the relative standard method to characterize the pore parameters quantitatively by Barrett–Joyner–Halenda (BJH) model and the Brunauer–Emmett–Teller (BET) method, which could provide information about the pore size, surface area, pore volume, etc. Positron annihilation measurement could provide the pore size, size distribution, pore density, etc. Besides, it can detect both mesopore and micropore, no matter it is open or closed. Our results show that positron is a superb probe to get comprehensive information about pores. 4. Conclusion Mesoporous silica were synthesized and the pore structure was studied by SAXS, SEM, HRTEM, N2 adsorption–desorption measurements and positron annihilation spectroscopy. SAXS and HRTEM measurements both confirm the ordered pore structure in the synthesized silica, and the pores exhibit 2D hexagonal morphology with diameter of about 10 nm. With increasing synthesis temperature, the pore ordering was slightly damaged at 130 ◦ C. However, at 150 ◦ C, the pores show ordered structure again. When the synthesis temperature is as high as 180 ◦ C, the ordered pores are severely damaged. N2 adsorption–desorption measurements reveal that the ordered pore radius is around 6 nm, but it increases to 24 nm when the pores collapse at 180 ◦ C. Positron lifetime measurements indicate that there are two kinds of pores in the synthesized samples, one is the large pores which are detected by TEM and N2 adsorption–desorption measurements, another is the micropores with radius less than 1 nm. The change of the size of the large pore obtained from the o-Ps lifetime using the cylindrical and semi-empirical model shows good agreement with the results of N2 adsorption–desorption measurements. The positron lifetime results also reveal that the collapse of large pores at 180 ◦ C leads to drastic decrease of the pore density, and some large pores break into small micropores.
[1] M.A. Al Ghouti, M.A.M. Khraisheh, S.J. Allen, M.N. Ahmad, J. Environ. Manag. 69 (2003) 229–238. [2] E. Mahmoud, R.F. Lobo, Microporous Mesoporous Mater. 189 (2014) 97–106. [3] L.H. Chen, X.Y. Li, J.C. Rooke, Y.H. Zhang, X.Y. Yang, Y. Tang, F.S. Xiao, B.L. Su, J. Mater. Chem. 22 (2012) 17381–17403. [4] X. Zhang, J.C. Wang, K.M. Lacki, A.I. Liapis, J. Colloid Interface Sci. 277 (2004) 483–498. [5] B.S. Li, W. Ma, J.J. Liu, C.Y. Han, S.L. Zuo, X.F. Li, Catal. Commun. 13 (2011) 101–105. [6] Y.Z. Chen, M.H. Wu, K.Y. Wang, B. Chen, S.Z. Yao, H.F. Zou, L.H. Nie, J. Chromatogr. A 1218 (2011) 7982–7988. [7] S. Hudson, R.F. Padera, R. Langer, D.S. Kohane, Biomaterials 29 (2008) 4045–4055. [8] H. Song, R.M. Rioux, J.D. Hoefelmeyer, R. Komor, K. Niesz, M. Grass, P.D. Yang, G.A. Somorjai, J. Am. Chem. Soc. 128 (2006) 3027–3037. [9] K. Soni, B.S. Rana, A.K. Sinha, A. Bhaumik, M. Nandi, M. Kumar, G.M. Dhar, Appl. Catal. B: Environ. 90 (2009) 55–63. [10] H. Takahashi, B. Li, T. Sasaki, C. Miyazaki, T. Kajino, S. Inagaki, Chem. Mater. 12 (2000) 3301–3305. [11] T.W. Kim, F. Kleitz, B. Paul, R. Ryoo, J. Am. Chem. Soc. 127 (2005) 7601–7610. [12] B.R. Jermy, D.R. Cho, K.V. Bineesh, S.Y. Kim, D.W. Park, Microporous Mesoporous Mater. 115 (2008) 281–292. [13] T.W. Kim, R. Ryoo, K.P. Gierszal, M. Jaroniec, L.A. Solovyov, Y. Sakamotod, O. Terasakid, J. Mater. Chem. 15 (2005) 1560–1571. [14] C.G. Sonwane, S.K. Bhatia, J. Phys. Chem. B 104 (2000) 9099–9110. [15] M.J. Mosquera, J. Pozo, L. Esquivias, T. Rivas, B. Silva, J. Non-Cryst. Solids 311 (2002) 185–194. [16] J. Lahtinen, P. Hautojarvi, J. Phys. Chem. B 101 (1997) 1609–1614. [17] H.J. Zhang, Z.Q. Chen, S.J. Wang, A. Kawasuso, N. Morishita, Phys. Rev. B 82 (2010) 035439. [18] M. Koshimizu, K. Shimokita, H.S. Zhou, I. Honma, K. Asai, J. Am. Chem. Soc. 112 (2008) 8779–8783. [19] M. Liu, A.G. Wong-Foy, R.S. Vallery, W.E. Frieze, J.K. Schnobrich, D.W. Gidley, A.J. Matzger, Adv. Mater. 22 (2010) 1598–1601. [20] R. Zaleski, A. Kierys, M. Dziadosz, J. Goworek, I. Halasz, RSC Adv. 2 (2012) 3729–3734. [21] A. Kierys, R. Zaleski, M. Gorgol, J. Goworek, Microporous Mesoporous Mater. 179 (2013) 104–110. [22] K.S. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, Pure Appl. Chem. 57 (1985) 603–619. [23] M. Barczak, M.O. Cieniuch, P. Borowski, Z. Fekner, E. Zieba, J. Therm. Anal. Calorim. 108 (2012) 1093–1099. [24] C.C. Ting, H.Y. Wu, S. Vetrivel, D. Saikia, Y.C. Pan, G.T.K. Fey, H.M. Kao, Microporous Mesoporous Mater. 128 (2010) 1–11. [25] L.X. Li, H.H. Song, X.H. Chen, Microporous Mesoporous Mater. 92 (2006) 9–14. [26] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380. [27] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–318. [28] T. Benamor, L. Vidal, B. Lebeau, C. Marichal, Microporous Mesoporous Mater. 153 (2012) 100–114. [29] Z. Zheng, Y. Jie, J.H. Hero, M.C. Ignacio, Microporous Mesoporous Mater. 8 (2013) 103–111. [30] S. Boskovic, A.J. Hill, T.W. Turney, G.W. Stevens, M.L. Gee, A.J. O’Connor, J. Am. Chem. Soc. 113 (2009) 21283–21292. [31] M.E. Davis, Nature 417 (2002) 813–821. [32] A. Sousa, K.C. Souza, S.C. Reis, R.G. Sousa, D. Windmoller, J.C. Machado, E.M.B. Sousa, J. Non-Cryst. Solids 354 (2008) 4800–4805. [33] V.B. Dmitry, N.P. Valentin, A.L. Alexei, C.W. Frank, J. Mater. Chem. 14 (2004) 3370–3377. [34] P. Kirkegaard, N.J. Pederson, M. Eldrup, RisReport M2740, Ris National Laboratory, Denmark, 1989. [35] R. Paulin, G. Ambrosino, J. Phys. (France) 29 (1968) 263–268. [36] S.J. Tao, J. Chem. Phys. 56 (1972) 5499–5510. [37] M. Eldrup, D. Lightbody, J.N. Sherwood, Chem. Phys. 63 (1981) 51–58. [38] D.M. Schrader, Y.C. Jean (Eds.), Positron and Positronuim Chemistry, Elsevier, Amsterdam, 1988 (Chapter 5). [39] T.L. Dull, W.E. Frieze, D.W. Gidley, J. Phys. Chem. B 105 (2001) 4657–4662. [40] R. Zaleski, J. Wawryszczuk, T. Goworek, Radiat. Phys. Chem. 76 (2007) 243–247. [41] S. Thranert, D. Enke, G. Dlubek, R. Krause-Rehberg, Mater. Sci. Forum 607 (2009) 169. [42] S.J. Lue, D.T. Lee, J.Y. Chen, C.H. Chiu, C.C. Hu, Y.C. Jean, J.Y. Lai, J. Membr. Sci. 325 (2008) 831–839. [43] K. Ito, H. Nakanishi, Y. Ujihira, J. Phys. Chem. B 103 (1999) 4555–4558. [44] D.W. Gidley, W.E. Frieze, T.L. Dull, A.F. Yee, E.T. Ryan, H.M. Ho, Phys. Rev. B 60 (1999) 5157–5160.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Effect of synthesis temperature on the ordered pore structure in mesoporous silica studied by positron annihilation spectroscopy C.Y. Li, N. Qi, Z.W. Liu, B. Zhou, Z.Q. Chen ∗ , Z. Wang ∗ Hubei Nuclear Solid Physics Key Laboratory, Department of Physics, Wuhan University, Wuhan 430072, PR China
a r t i c l e
i n f o
Article history: Received 30 April 2015 Received in revised form 1 December 2015 Accepted 5 December 2015 Available online 12 December 2015 Keywords: Ordered mesoporous silica Synthesis temperature Pore structure N2 adsorption–desorption Positronium
a b s t r a c t Mesoporous silica with ordered pore structure was synthesized at various temperatures using TEOS as the silica source and P123 as the template. Both small angle X-ray scattering (SAXS) and high resolution transition electron microscopy (HRTEM) measurements verify the ordered pore structure of the synthesized SiO2 . With synthesis temperature increasing, the pore structure has slight damage at 130 ◦ C, while it shows ordered structure again at 150 ◦ C. When the synthesis temperature increases to 180 ◦ C, the ordered pore structure is severely destructed. The change of pore structure is further confirmed by scanning electron microscopy (SEM) measurements. Positron lifetime measurements reveal four lifetime components in the synthesized mesoporous SiO2 , and the two long lifetimes 3 and 4 correspond to the annihilation of o-Ps in the micropores and large pores of the material, respectively. The longest lifetime 4 tends to increase slightly with increasing synthesis temperature. However, its intensity I4 shows an overall decrease with exception at 150 ◦ C. At the synthesis temperature of 180 ◦ C, the intensity I4 decreases drastically to about 17.5%. This indicates variation of the size and fraction of pores with increasing synthesis temperature, and the pore structure is seriously destructed at 180 ◦ C. By comparing with the N2 adsorption–desorption measurements, it was found that the Goworek’s model is more suitable for the size estimation of cylindrical pores from the o-Ps lifetime, while Dull’s and Ito’s model is appropriate for the rectangular and spherical pores, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Mesoporous materials, from the natural materials such as diatomaceous earth, natural zeolite and polypite [1–3], to the synthesized porous materials [4–11], have found their wide applications in many fields and attracted more and more attentions. The research areas and related applications are further broadened especially after the appearance of the controllable ordered porous materials, which not only make up the deficiency of microporous zeolite materials [1] for the application in the macromolecules adsorption [4] and catalytic reaction [5], but also can be applied as the microreactor [6] of the controllable nano-materials in chemistry industries, biological technology [7], adsorption isolation [8], environment catalysis, [9], etc. The most commonly used material is ordered mesoporous silica [10–13], such as MCM-41, MCM-48, SBA-15, KIT-6, and SBA-16, which contains different pore size, surface area, pore volume and pore morphology.
∗ Corresponding authors. E-mail addresses: [email protected] (Z.Q. Chen), [email protected] (Z. Wang). http://dx.doi.org/10.1016/j.apsusc.2015.12.055 0169-4332/© 2015 Elsevier B.V. All rights reserved.
Ordered mesoporous silica can be easily synthesized at lower temperatures, even at the room temperature. However, at lower temperatures such as room temperature, large amounts of terminal hydroxyl groups will be irregularly stacked, which might cause the irregular rank of the wall structure of the ordered pore and result in the collapse of the pore structure. Therefore the mesopore in silica possesses a relatively low ordering. Higher synthesis temperature might result in a more stable and ordered mesopore structure. Nevertheless, a much too high temperature will also destroy the ordered pore structure. Therefore an appropriate synthesis temperature should be chosen to attain the balance between pore structure and pore stability. The pore structure such as pore size and their size distribution can be detected by the gas adsorption method [14]. This method is based on the fact that below the saturated vapor pressure of the adsorbate, it condenses to liquid in narrow pores. However, this method is only sensitive to mesopores (2–50 nm), and only open pores can be detected. In a closed-pore material, gases cannot enter the pore, so the adsorption method fails to work. Mercury intrusion porosimeter [15] is another approach to estimate the pore size. However, this porosimeter is sensitive to pores with even larger size. The pores with radius less than 10 nm will escape
446
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
from the detection by this method. Some other methods might be more useful for the study of pore structure in mesoporous materials. Positron annihilation spectroscopy has emerged as a powerful probe for the pore structure [16–21]. In many materials, positron will capture one electron to form a metastable hydrogen-like particle called positronium (Ps) atom. According to the spins of the positron and electron, the Ps atom exists in two states: the triplet spin state ortho-positronium (o-Ps) (s = 1, ms = −1, 0, +1) and the singlet spin state para-positronium (p-Ps) (s = 0, ms = 0). The formation probability of o-Ps is about three times of that of p-Ps. In vacuum, the o-Ps has a long lifetime of 142 ns, and it converts to 3 rays through self-annihilation. The p-Ps has a relatively shorter self-annihilation lifetime of 125 ps by emitting 2 rays. In porous materials, the annihilation of o-Ps depends on the structure of the pore [22]. When o-Ps is trapped in the pore, it will pick one electron from the wall of the pore and then annihilates into 2 rays, which is called pick-off annihilation. For pores with diameter smaller than 1 nm, the pick-off annihilation lifetime of oPs will be reduced to about 1–20 ns, which is closely related to the pore size. The positronium formation probability, which is reflected by the o-Ps intensity, is in accordance with the porosity. So, the formation and annihilation of positronium, especially the o-Ps, could be utilized to study the pore structure. This probe can detect micropores and mesopores, and can detect both closed pores and open pores, which shows advantage over other methods such as the gas adsorption method [14]. In this paper, we prepared a series of porous silica under acidic condition [23–25]. The effect of synthesis temperature on the pore structure was studied by positron annihilation measurements together with SAXS, HRTEM, SEM and N2 adsorption–desorption measurement.
2. Experiment Mesoporous silica was prepared using amphiphilic triblock copolymer Pluronic P123 (Mw = 5800, EO20 –PO70 –EO20 , Sigma Aldrich) as the structure agent and tetraethyl orthosilicate (TEOS, C8 H20 O4 Si) as the silica source [24]. The 37 wt% HCl aqueous solution is diluted to 2 mol/L with distilled water for further use. P123 was first dissolved in a mixture of distilled water and HCl solution, then it was stirred under constant agitation at room temperature for 30 min until P123 is completely dissolved. After that, H2 SO4 and sucrose were added to the mixture and was stirred for another 30 min. Finally, the silica source TEOS was added to the mixture and was stirred until the solution was distributed homogenously. The mole ratio of P123:TEOS:HCl:H2 SO4 :H2 O:sucrose is 1:48.3:649:40:6357:7.2. The mixture was stirred for 72 h at 323 K and then conveyed into the Teflon-lined autoclave under static conditions at the given synthesis temperature(80 ◦ C, 100 ◦ C, 130 ◦ C, 150 ◦ C, 180 ◦ C). The final silica/P123/sucrose gel was filtered, washed with distilled water (and ethanol) and dried at 373 K for 24 h. At the last, the final silica powder was obtained after calcination at 823 K for 6 h in the open air. The synthesized mesoporous silica powders were further hand milled in agate mortar for 2 h and then pressed under a static pressure of 6 MPa for 5 min to pellets with 1.5 mm thickness and a diameter of 15 mm. These pellets were dried over night at 100 ◦ C in the autoclave for other measurements. SAXS measurements were carried out using an X-ray diffractometer (X’Pert Pro, PANalytical, Netherlands) with Cu K˛ radiation operated at 40 kV and 40 mA. The incident X-ray wavelength was 1.5406 A˚ and the scanning angle 2 was from 0.6123 to 10◦ with the step of 0.001◦ . N2 adsorption–desorption measurements were performed at 77 K by using a Micromeritics ASAP
2020 gas-sorption analyzer. The sample was degassed in a vacuum at 180 ◦ C for the desorption process. Pore size distribution and average pore size were estimated from adsorption branch of the isotherms using the Barrett–Joyner–Halenda (BJH) method [26]. At the same time the specific surface area was calculated from the Brunauer–Emmett–Teller (BET) method [27] over the relative pressure P/P0 range of 0.05–0.25. Scanning electron microscope (SEM) (XL 30 (Philips), Amsterdam, Netherlands) and high resolution transmission electron microscopy (HRTEM) (JEOL JEM2010FEF (UHR), Tokyo, Japan) images with acceleration voltage at 200 kV were taken to study the pore morphology and pore ordering. Positron lifetime measurements were performed at room temperature using a conventional fast–fast coincidence lifetime spectrometer with a time resolution of 280 ps in full width at half maximum (FWHM). A 10 Ci 22 NaCl source was sandwiched between two pellet samples and the sample–source–sample sandwich was placed into a vacuum chamber with air pressure lower than 1 × 10−5 torr. The lower level of the energy window on the discriminator for the stop signal (0.511 MeV annihilation -ray) was set to as low as possible in order to collect the 3 annihilation signal of o-Ps. In our experiment, we have tried to measure the positron lifetime spectrum using a time range of 500 ns for the TAC. The resolved long lived lifetime values ( 3 and 4 ) and their intensities are almost the same for the two different time ranges. But time range of TAC of 500 ns leads to the larger uncertainties for the short lifetime components. So in our later measurements we just used the 200 ns range. The total channel number is 4096 and the time scale is 50.3 ps/channel. For each sample, two lifetime spectra were collected with a total count of 1.5 × 106 for each spectrum. The counting rate is about 26 count per second. Doppler broadening spectra of the positron annihilation radiation were measured simultaneously using a high purity Ge (HP-Ge)detector which was placed perpendicular to the direction of positron lifetime measurements. The energy resolution of the HP-Ge detector is about 1.2 keV in full width at half maximum at 511 keV. The S and W parameters are calculated from the spectrum, which represent the ratio of low-energy (|EL | ≤ 0.68 keV) region and high-energy region (2.8 keV ≤ |EL | ≤ 5.73 keV) to the total region of 511 keV annihilation peak, respectively. The total count of each Doppler broadening spectrum is more than 1 × 107 .
3. Results and discussion 3.1. SAXS measurement In order to verify the pore ordering of the synthesized silica, SAXS measurements were performed for the samples synthesized at various temperatures. The measured SAXS patterns are shown in Fig. 1. For most of the samples, the diffraction peaks are nearly the same. There are three main diffraction peaks at around 0.98, 1.64, and 1.85◦ . They are indexed as the (100), (110), and (200) diffraction peaks, which reflect the hexagonal space group (P6mm). This indicates that the porous silica has long-range ordering. In other words, our synthesized silica has ordered structure. It can be seen from Fig. 1 that the positions of the three diffraction peaks have some shifts at different synthesis temperatures. More importantly, the intensities of these peaks have significant change with increasing synthesis temperature. Actually for the sample synthesized at 130 ◦ C, all the three peaks (100), (110), and (200) become weaker, which indicates that the pore ordering is somewhat deteriorated at this temperature. However, for the sample synthesized at 150 ◦ C, the pores show high ordering again. When the synthesis temperature increases to 180 ◦ C, only the peak at 0.98◦ shows weak signal, and the other two peaks disappear completely. This suggests that pore ordering is
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
447
that at 180 ◦ C, the (100) peak is invisible, so the pore width cannot be estimated.
3.2. SEM measurement
Fig. 1. Small-angle X-ray scatting patterns for the synthesized ordered mesoporous silica.
severely damaged. The temperature of 180 ◦ C is probably too high to keep the ordered pore structure. A more suitable synthesis temperature will be 150 ◦ C or even lower. It has been also reported earlier that appropriate synthesis temperature was below 130 ◦ C [22,28–30]. The periodic interval can be estimated by the Bragg’s equation using the position of the main diffraction peak (100): 2d sin = k
(1)
The small angle X-ray scattering measurements can give us the periodic interval of the pores (pore width + pore wall) for the ordered mesoporous silica, which was labeled in Fig. 1. This interval can be used as a rough estimation of the variation of pore size. It increases from 9.10 nm to 10.19 nm at the synthesis temperature at 80–130 ◦ C, and then decrease a little to 9.72 nm at 150 ◦ C. As for
In order to observe the detailed morphology of these samples, SEM images were taken for the silica powder synthesized at different temperatures, which were shown in Fig. 2. The samples reveal worm-like morphology containing particles with irregular shape. The particles are approximately 0.4 m in diameter and 0.5–1.5 m in length. The surface of the particles change evidently with increasing synthesis temperature. The worm-like particles are relatively smooth for the samples synthesized at lower temperatures such as 80 and 100 ◦ C. However, when the synthesis temperature increases to 130 ◦ C, the surface is slightly deconstructed with some cracks. When the temperature increases to 150 ◦ C, the surface of the worm-like particle becomes smooth again. However, as for the sample synthesized at 180 ◦ C, the particle surface is almost full of cracks. SEM measurement reveals the variation of morphology with increasing synthesis temperature.
3.3. HRTEM measurement HRTEM images were taken for the synthesized silica from two different directions (perpendicular and parallel to the pore directions). They were shown in Fig. 3. The samples show a regular pore arrangement. The ordered 2D hexagonal cylindrical pores can be clearly seen. The diameter of the pore is about 10 nm. From the TEM images it can be clearly seen that the pore ordering becomes weaker with increasing synthesis temperature from 80 ◦ C to 130 ◦ C, but the pores show high order again at the relative higher temperature of 150 ◦ C. For the sample synthesized at much higher temperature of 180 ◦ C, the pore ordering is nearly invisible. This observation coincides with the SAXS and SEM measurements. All these results suggest that a relatively high temperature of 150 ◦ C maintains the
Fig. 2. Scanning electron diffraction images measured for the synthesized ordered mesoporous silica.
Fig. 3. High resolution transmission electron diffraction images measured for the synthesized ordered mesoporous silica from perpendicular and parallel to the channel directions.
448
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
Fig. 5. Peak-normalized positron lifetime spectrum for the synthesized ordered mesoporous silica at the synthetic temperature of 80 ◦ C, 130 ◦ C, 180 ◦ C.
radius increases to about 24 nm. For these large pores, the size is already close to the detection limit of N2 adsorption–desorption measurement, which is the reason for the distortion of the adsorption–desorption isotherms [33]. Fig. 4. Nitrogen adsorption–desorption isothermal of ordered mesoporous silica (a) and their relative pore size distribution (b) as a function of synthetic temperature.
well ordered mesoporous structure and might be the appropriate temperature to synthesize stable and ordered mesoporous silica. 3.4. N2 adsorption–desorption measurement In order to get more detailed pore parameters such as pore size, pore volume and surface area, the nitrogen adsorption–desorption measurements were performed at 77 K for the synthesized samples. The adsorption–desorption isotherms measured for each sample were plotted in Fig. 4(a). According to the classification of IUPAC [31], except for the sample synthesized at 180 ◦ C, the nitrogen adsorption–desorption isotherms are in accordance with the defined typical IV curves with H1 hysteresis loops and the clear capillary condensation, which occurs approximately at the relative pressure P/P0 of 0.6–0.9. This ratio is associated with the filling of mesopores by nitrogen, which suggests that the synthesized silica is mesoporous [32]. With the synthesis temperature increasing to 130 ◦ C, the relative pressure P/P0 shifts to a higher value and that indicates increase of the most probable pore radius. For the sample synthesized at 150 ◦ C, the adsorption–desorption isotherms almost overlaps with that of the sample synthesized at 100 ◦ C, and the relative pressure P/P0 shifts to a relatively lower value. However, when the synthesis temperature increases to 180 ◦ C, hysteresis loop of its adsorption–desorption isotherms is not the typical IV curves. The relative pressure P/P0 is approximately 1, which indicates presence of large macropores. Meanwhile, the pore size distribution curves of these five samples were calculated from the adsorption branches by Barrett–Joyner–Halenda (BJH) [26] model. They were plotted in Fig. 4(b). For the samples synthesized at temperatures below 180 ◦ C, they all exhibit relatively narrower pore size distribution. The most probable pore size has a little fluctuation around 10 nm, and the average pore radius is around 7–8 nm. With the synthesis temperature increasing from 80 to 150 ◦ C, the average pore radius first increases from 5.5 nm to 10.5 nm, then it decreases to 8.5 nm. This shows agreement with the results by SAXS measurements. The width of the distribution also increases with temperature increasing from 80 to 130 ◦ C, but it becomes narrower again at 150 ◦ C. For the sample synthesized at 180 ◦ C, of which the ordered pores are nearly invisible, it has a very wide pore size distribution, which starts at 5 nm and extends to 45 nm. The average pore
3.5. Positron annihilation measurements Positron annihilation spectroscopy was used to get more detailed information about the pore structure of synthesized silica. Fig. 5 shows the peak-normalized positron lifetime spectra measured for porous silica synthesized at 80 ◦ C, 130 ◦ C and 180 ◦ C. It is obvious that there is a very long positron lifetime component in the ordered mesoporous SiO2 , indicating that the sample contains large pores. Four lifetime components were derived from the positron annihilation lifetime spectra by PATFIT program [34]. The shortest lifetimes 1 (161 ps) can be ascribed to the annihilation lifetime of free positrons and p-Ps annihilation in silica, while the second lifetime 2 (510 ps) is due to positron annihilation in vacancies or voids. The two longer lifetime 3 (4–8 ns) and 4 (110–120 ns) are obviously o-Ps annihilation lifetime in the pores [35] of the material. In this paper, since our research interest is on the pore structure, we will put our attentions only on the long lifetime components 3 , 4 and their relative intensities I3 and I4 . It is well known that the o-Ps lifetime is closely related to the pore size in materials. The existence of two o-Ps lifetime components indicates that there are two different kinds of pores, which has small and large open volume, respectively. The longest lifetime 4 is apparently due to o-Ps annihilation in the 2D P6mm hexagonal ordered pores in the silica material. The shorter o-Ps lifetime 3 is the o-Ps lifetime in some micropores. However, these micropores are not detected by the N2 adsorption–desorption measurements. They might be the pipes that connect 2D P6mm hexagonal ordered pores. Tao and Eldrup et al. had already established a semi-empirical relationship between the o-Ps lifetime o−Ps and the free volume size R assuming a spherical infinite potential well of radius R0 with an electron layer of thickness R = R − R0 [36–38]:
−1 o−Ps =2 1−
R R 1 sin 2 + 2 R + R R + R
.
(2)
The thickness R is determined to be 1.656 A˚ for most of the molecular materials. While, for silica in our paper, R is estimated as 0.18 A˚ [39–41]. This equation is valid for o-Ps pick-off annihilation in free volume with width smaller than 1 nm [42,43]. For positronium localized in pores with diameter larger than 1 nm, some fraction of the o-Ps will have no chance to pick electron from the wall around the pore, so this part of o-Ps will undergo self-annihilation and emit 3-rays, which leads to a much longer lifetime of about 30–135 ns. Ito et al. have modified the above Tao–Eldrup model by taking into consideration the intrinsic three
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
449
self-annihilation of o-Ps in large pores, and proposed the following model which is also valid for o-Ps annihilation in small micropores less than 1 nm [43].
−1 o−Ps = 2 1−
+
1 Ra sin + 2 Ra + R
2R a Ra + R
1−
R − R b a R + R
1 , 142
(3)
Ra and b are best fitted with experimental data as 0.8 nm and 0.55, respectively [43]. By using the above two equations, we can thus estimate the pore size in our silica sample. For the sample synthesized at 80 ◦ C, the diameter of the small micropores corresponding to 3 is about 0.64 nm, while that of the large pore is about 16.4 nm. However, according to the N2 adsorption–desorption measurements the size of the large pore is only about 5.5 nm in width. This suggests that the model by Ito et al. gives overestimated size of the pore. The reason might be due to the shape of the pores in our sample. Ito et al. still use the spherical voids in the calculation, while our TEM observation indicates that the pores in our silica sample have ordered 2D hexagonal cylindrical geometry, thus the modified model by Ito is not applicable for this kind of pores. Another reason is that Ito’s model is based on previous experimental data. However, some of the data were obtained in air. This will also give overestimation of the pore size from the o-Ps lifetime. Besides the modified model by Ito et al, there are also some other extended Tao–Eldrup models to correlate o-Ps lifetime and pore size for different pore geometry, such as the rectangular tube [44,39], as shown below: (T ) = A −
S − T F(a, ı, T )F(b, ı, T )F(c, ı, T ), 4
where F(x, ı, T ) = 1 −
2ı + x
∞ i=1
(1/(i)) sin((2iı)/x)e((−ˇt
∞
i=1
e((−ˇt
2 )/(x2 kT ))
(4)
2 )/(x2 kT ))
, (5)
and ˇ = h2 /(16 m) = 0.188 eV nm2 . By using the above equation, the calculated pore width corresponding to 4 in silica sample synthesized at 80 ◦ C is about 6.6 nm. However, as for the 2D P6mm hexagonal ordered pores of synthesized silica, the cylindrical model by Goworek shall be much closer. In the revised paper, all the pore widths have been also calculated from the lifetime components by using Goworek’s model, which might be similar to that of the results of adsorption–desorption measurement. Fig. 6 shows the lifetimes 3 , 4 and their intensities I3 and I4 for the silica sample synthesized at different temperatures. With increase of the synthesis temperature, 3 keeps almost constant at about 4 ns, but the lifetime 4 shows irregular change, which assembles the change of pore size revealed by gas adsorption measurements. However, the intensities I3 and I4 change significantly with synthesis temperatures. The intensity I4 decreases gradually from 33.8% at 80 ◦ C to 17.5% at 180 ◦ C, with only one exception at 150 ◦ C, where I4 reaches to 31.6%. The trend of I3 is just opposite to that of I4 . Fig. 7 presents the radius of the pore in silica obtained from the oPs lifetime using the models proposed by Ito, Dull and Goworek. The results obtained from N2 adsorption–desorption measurements are also shown in the figure. As for the 2D P6mm hexagonal ordered pores of synthesized silica, the radius calculated by Ito’s spherical model is obviously overestimated, while that by Dull’s rectangular model will be better. The cylindrical model by Goworek shall be much closer to 2D P6mm hexagonal pore structure. The results from Goworek’s model are in good agreement with that of N2 adsorption–desorption measurements except for the point
Fig. 6. Variation of o-Ps lifetime 4 , 3 and their intensities I4 , I3 for the synthesized ordered mesoporous silica as a function of synthetic temperature.
at temperature of 180 ◦ C. This indicates that the pores have cylindrical tube-liked structure, which has been confirmed by HRTEM observation. As for the sample synthesized at 180 ◦ C, the pore diameter calculated by Ito’s model agrees well with that of N2 adsorption–desorption measurements. This may tell us that pore in silica synthesized at 180 ◦ C shall be sphere-like. However, the discrepancy between the pore size obtained by PALS and other methods for the sample synthesized at 180 ◦ C might be due to the underestimation of the o-Ps lifetime. Besides the large pores, the o-Ps lifetime also gives information on the small micropores. According to the Tao–Eldrup model, the size of micropore changes little with increasing synthesis temperature, but its concentration (reflected by I3 ) increases to more than 10% when the temperature increases to 180 ◦ C. The third lifetime reflects information of small micropores. The quick increase of I3 at 180 ◦ C indicates increase in the number of small micropores. At the same time, the intensity I4 shows abrupt decrease. Therefore we believe that these small micropores might be produced by the collapse of large pores. We also performed Doppler broadening measurements for the SiO2 to study the pore structure. If the positronium is formed in porous materials, one third of the positronium will be p-Ps. Since the p-Ps contains nearly zero momentum, its self-annihilation will contribute a very narrow peak to the Doppler broadening spectrum,
Fig. 7. The pore radius of the synthesized ordered mesoporous silica attained from N2 adsorption–desorption and PAL measurements according to different model as a function of synthetic temperatures.
450
C.Y. Li et al. / Applied Surface Science 363 (2016) 445–450
Acknowledgment This work was supported by the National Natural Science Foundation of China under Grant Nos. 11475130, 11305117 and 11275143. References
Fig. 8. Variation of S parameters for the synthesized ordered mesoporous silica as a function of synthesis temperature.
which leads to increase in the S parameter. From Fig. 8, we can see that the variation of S parameter is very similar to the change of I4 . Due to the contribution of p-Ps annihilation to the Doppler broadening spectrum, the S parameter is closely related to the Ps intensity. Although in our paper the intensity of o-Ps is the sum of I3 and I4 , Io−Ps is approximately equal to I4 with a minimum contribution from I3 . The p-Ps intensity is one third of o-Ps fraction. So, the comparison of S to I4 is more reasonable. Comparing the total five measurements, each one has its own superiority on exhibiting the different aspects of porous materials. Small-angle X-ray scatting measurement is usually used to verify the degree of ordering, which is the basic measurement of ordered mesoporous pores. SEM measurement is appropriate for the pore morphology. HRTEM measurement is a more directly view to express the pore structure. N2 adsorption–desorption measurements are the relative standard method to characterize the pore parameters quantitatively by Barrett–Joyner–Halenda (BJH) model and the Brunauer–Emmett–Teller (BET) method, which could provide information about the pore size, surface area, pore volume, etc. Positron annihilation measurement could provide the pore size, size distribution, pore density, etc. Besides, it can detect both mesopore and micropore, no matter it is open or closed. Our results show that positron is a superb probe to get comprehensive information about pores. 4. Conclusion Mesoporous silica were synthesized and the pore structure was studied by SAXS, SEM, HRTEM, N2 adsorption–desorption measurements and positron annihilation spectroscopy. SAXS and HRTEM measurements both confirm the ordered pore structure in the synthesized silica, and the pores exhibit 2D hexagonal morphology with diameter of about 10 nm. With increasing synthesis temperature, the pore ordering was slightly damaged at 130 ◦ C. However, at 150 ◦ C, the pores show ordered structure again. When the synthesis temperature is as high as 180 ◦ C, the ordered pores are severely damaged. N2 adsorption–desorption measurements reveal that the ordered pore radius is around 6 nm, but it increases to 24 nm when the pores collapse at 180 ◦ C. Positron lifetime measurements indicate that there are two kinds of pores in the synthesized samples, one is the large pores which are detected by TEM and N2 adsorption–desorption measurements, another is the micropores with radius less than 1 nm. The change of the size of the large pore obtained from the o-Ps lifetime using the cylindrical and semi-empirical model shows good agreement with the results of N2 adsorption–desorption measurements. The positron lifetime results also reveal that the collapse of large pores at 180 ◦ C leads to drastic decrease of the pore density, and some large pores break into small micropores.
[1] M.A. Al Ghouti, M.A.M. Khraisheh, S.J. Allen, M.N. Ahmad, J. Environ. Manag. 69 (2003) 229–238. [2] E. Mahmoud, R.F. Lobo, Microporous Mesoporous Mater. 189 (2014) 97–106. [3] L.H. Chen, X.Y. Li, J.C. Rooke, Y.H. Zhang, X.Y. Yang, Y. Tang, F.S. Xiao, B.L. Su, J. Mater. Chem. 22 (2012) 17381–17403. [4] X. Zhang, J.C. Wang, K.M. Lacki, A.I. Liapis, J. Colloid Interface Sci. 277 (2004) 483–498. [5] B.S. Li, W. Ma, J.J. Liu, C.Y. Han, S.L. Zuo, X.F. Li, Catal. Commun. 13 (2011) 101–105. [6] Y.Z. Chen, M.H. Wu, K.Y. Wang, B. Chen, S.Z. Yao, H.F. Zou, L.H. Nie, J. Chromatogr. A 1218 (2011) 7982–7988. [7] S. Hudson, R.F. Padera, R. Langer, D.S. Kohane, Biomaterials 29 (2008) 4045–4055. [8] H. Song, R.M. Rioux, J.D. Hoefelmeyer, R. Komor, K. Niesz, M. Grass, P.D. Yang, G.A. Somorjai, J. Am. Chem. Soc. 128 (2006) 3027–3037. [9] K. Soni, B.S. Rana, A.K. Sinha, A. Bhaumik, M. Nandi, M. Kumar, G.M. Dhar, Appl. Catal. B: Environ. 90 (2009) 55–63. [10] H. Takahashi, B. Li, T. Sasaki, C. Miyazaki, T. Kajino, S. Inagaki, Chem. Mater. 12 (2000) 3301–3305. [11] T.W. Kim, F. Kleitz, B. Paul, R. Ryoo, J. Am. Chem. Soc. 127 (2005) 7601–7610. [12] B.R. Jermy, D.R. Cho, K.V. Bineesh, S.Y. Kim, D.W. Park, Microporous Mesoporous Mater. 115 (2008) 281–292. [13] T.W. Kim, R. Ryoo, K.P. Gierszal, M. Jaroniec, L.A. Solovyov, Y. Sakamotod, O. Terasakid, J. Mater. Chem. 15 (2005) 1560–1571. [14] C.G. Sonwane, S.K. Bhatia, J. Phys. Chem. B 104 (2000) 9099–9110. [15] M.J. Mosquera, J. Pozo, L. Esquivias, T. Rivas, B. Silva, J. Non-Cryst. Solids 311 (2002) 185–194. [16] J. Lahtinen, P. Hautojarvi, J. Phys. Chem. B 101 (1997) 1609–1614. [17] H.J. Zhang, Z.Q. Chen, S.J. Wang, A. Kawasuso, N. Morishita, Phys. Rev. B 82 (2010) 035439. [18] M. Koshimizu, K. Shimokita, H.S. Zhou, I. Honma, K. Asai, J. Am. Chem. Soc. 112 (2008) 8779–8783. [19] M. Liu, A.G. Wong-Foy, R.S. Vallery, W.E. Frieze, J.K. Schnobrich, D.W. Gidley, A.J. Matzger, Adv. Mater. 22 (2010) 1598–1601. [20] R. Zaleski, A. Kierys, M. Dziadosz, J. Goworek, I. Halasz, RSC Adv. 2 (2012) 3729–3734. [21] A. Kierys, R. Zaleski, M. Gorgol, J. Goworek, Microporous Mesoporous Mater. 179 (2013) 104–110. [22] K.S. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, Pure Appl. Chem. 57 (1985) 603–619. [23] M. Barczak, M.O. Cieniuch, P. Borowski, Z. Fekner, E. Zieba, J. Therm. Anal. Calorim. 108 (2012) 1093–1099. [24] C.C. Ting, H.Y. Wu, S. Vetrivel, D. Saikia, Y.C. Pan, G.T.K. Fey, H.M. Kao, Microporous Mesoporous Mater. 128 (2010) 1–11. [25] L.X. Li, H.H. Song, X.H. Chen, Microporous Mesoporous Mater. 92 (2006) 9–14. [26] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380. [27] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–318. [28] T. Benamor, L. Vidal, B. Lebeau, C. Marichal, Microporous Mesoporous Mater. 153 (2012) 100–114. [29] Z. Zheng, Y. Jie, J.H. Hero, M.C. Ignacio, Microporous Mesoporous Mater. 8 (2013) 103–111. [30] S. Boskovic, A.J. Hill, T.W. Turney, G.W. Stevens, M.L. Gee, A.J. O’Connor, J. Am. Chem. Soc. 113 (2009) 21283–21292. [31] M.E. Davis, Nature 417 (2002) 813–821. [32] A. Sousa, K.C. Souza, S.C. Reis, R.G. Sousa, D. Windmoller, J.C. Machado, E.M.B. Sousa, J. Non-Cryst. Solids 354 (2008) 4800–4805. [33] V.B. Dmitry, N.P. Valentin, A.L. Alexei, C.W. Frank, J. Mater. Chem. 14 (2004) 3370–3377. [34] P. Kirkegaard, N.J. Pederson, M. Eldrup, RisReport M2740, Ris National Laboratory, Denmark, 1989. [35] R. Paulin, G. Ambrosino, J. Phys. (France) 29 (1968) 263–268. [36] S.J. Tao, J. Chem. Phys. 56 (1972) 5499–5510. [37] M. Eldrup, D. Lightbody, J.N. Sherwood, Chem. Phys. 63 (1981) 51–58. [38] D.M. Schrader, Y.C. Jean (Eds.), Positron and Positronuim Chemistry, Elsevier, Amsterdam, 1988 (Chapter 5). [39] T.L. Dull, W.E. Frieze, D.W. Gidley, J. Phys. Chem. B 105 (2001) 4657–4662. [40] R. Zaleski, J. Wawryszczuk, T. Goworek, Radiat. Phys. Chem. 76 (2007) 243–247. [41] S. Thranert, D. Enke, G. Dlubek, R. Krause-Rehberg, Mater. Sci. Forum 607 (2009) 169. [42] S.J. Lue, D.T. Lee, J.Y. Chen, C.H. Chiu, C.C. Hu, Y.C. Jean, J.Y. Lai, J. Membr. Sci. 325 (2008) 831–839. [43] K. Ito, H. Nakanishi, Y. Ujihira, J. Phys. Chem. B 103 (1999) 4555–4558. [44] D.W. Gidley, W.E. Frieze, T.L. Dull, A.F. Yee, E.T. Ryan, H.M. Ho, Phys. Rev. B 60 (1999) 5157–5160.