63
64
Figure 4-1. Scheme of considered support effect on acetic acid hydrogenation.
To find out MOFs which have a sufficient tolerance for the acetic acid hydrogenation reaction, I prepared 12 types of MOFs and examined their acetic acid stabilities through acetic acid–
exposure experiments with XRD measurement. As a result, seven types of MOFs (MIL-125-NH2, UiO-66-NH2, HKUST-1, MIL-101, Zn-MOF-74, Mg-MOF-74 and MIL-121) were found out as a favorable catalytic support.
The behaviors of the acetic acid adsorption on the MOFs were evaluated by determination of acetic acid–desorption temperature. To determine the temperature, acetic acid molecules were introduced into MOFs which were applied for TPD-MS (temperature programmed desorption–
mass spectrometry) experiments. The desorption temperatures of the MOFs depended on the type of MOFs. Especially, MIL-125-NH2 and UiO-66-NH2 exhibited high acetic acid–
adsorption ability.
Pt NPs were loaded on the MOFs, TiO2 and Al2O3 in the same method of chapter 2 and 3. The TiO2 is known as one of the best support for the acetic acid hydrogenation reaction. The catalytic performances of the prepared Pt catalysts (Pt/MIL-125-NH2, Pt/UiO-66-NH2,
Pt/HKUST-1, Pt/MIL-101, Pt/Zn-MOF-74, Pt/Mg-MOF-74, Pt/MIL-121, Pt/TiO2 and Pt/Al2O3) for the acetic acid hydrogenation were examined using a home-made fixed bed flow reactor with a pressure of 10 atm. The Pt/MIL-125-NH2 and Pt/UiO-66-NH2 exhibited the high catalytic activities comparable to Pt/TiO2. In particular, the yield of ethanol on
Pt/MIL-125-65
NH2 was superior to those on the other catalysts. From these results, I demonstrated that the support effect of “substrate adsorption by a MOF”.
4-2. Experimental
4-2-1. Acetic acid tolerance Exposure experiment
All Pt/MOFs were placed at 9 ml of sample tube without a cap, respectively. The sample tubes were placed in 50ml of sample tube with about 3 ml of acetic acid, respectively. Then, the larger ones were tightly capped (as drawn in Figure 4-2) to be filled by acetic acid vapor and stayed for over 6h. Structures of the samples which did not be a suspension by the vapor were analyzed by XRD (Cu-ka).
Figure 4-2. Scheme of screening test of acetic acid tolerance of MOFs.
in situ XRD measurement
All Pt/MOFs were placed in glass capillaries and heated at 150 °C for 6 - 12 h under vacuum to remove water in their pores. Subsequently, acetic acid vapor was introduced at vapor pressure of room temperature for over 6 h. After that, the glass capillaries were sealed without exposure to the atmosphere. XRD measurements of the samples were conducted with rotating the capillary sample using a synchrotron radiation at BL44B2 in SPring-8 with the approval of RIKEN. In in situ temperature dependent XRD measurement, the capillary samples were heated
66
by two nitrogen-flow devices.
4-2-2. Characterization of a behavior of acetic acid adsorption TPD-MS
Acetic acid–introduced MOF samples were prepared by almost the same condition for the acetic acid durability test (precise). A long sample tube made of Pyrex® was connected for a vacuum line made of SUS and used for this preparation as shown in Figure 4-3.
Temperature-programmed desorption mass spectroscopy (TPD-MS) was conducted using BELCAT-A with BEL-Mass (Microtrac BEL). About 30 mg of each sample was fixed in a U-shaped sample tube by a glass wool and used for this experiment. TPD-MS experiments were performed at 50 °C–300 °C under He flow, the ramp rate was 10°C /min. The outlet gas was analyzed by the mass spectrometer.
Figure 4-3. Scheme of a preparation of an acetic acid–introduced MOF. The left photo shows the sample tube with a MOF actually used.
67
4-2-3. Catalytic reactions
The vapor-phase acetic acid hydrogenation was performed using 50 mg of catalysts with homemade fixed-bed reactor consisting of stainless steel pipes which were heated at 100 °C to prevent condensation, as shown in Figure 4-4 and 4-5. All catalysts were pretreated at 200 °C for 2h under H2 flow. The pretreatments were carried out in situ to avoid exposure to air before catalytic reaction. After the pretreatment, saturated vapor pressure of acetic acid vapor (17.1 °C, 10Torr) was introduced with 30 ccm of H2 at 125 °C as reaction gas. The gas products were analyzed by two Shimadzu GC-8A gas chromatograph equipped with a Porapak T column using a FID detector and an active carbon column using a TCD detector, respectively.
Total flow rate and gas hourly space velocity (GHSV) were 30 ml min–1 and 150,000 ml g–1 h–1. The acetic acid conversion was calculated by the following equation.
Figure 4-4. Home-made flow reactor for online gas analysis of acetic acid hydrogenation reaction.
68
Figure 4-5. Photos of (left) an exterior and (right) a sample tube of the home-made flow reactor
Conversions, selectivities and yields of the acetic acid reaction were calculated from these formula.
Conversion (%) = [acetic acid]rf – [acetic acid] / [acetic acid]rf ×100 Selectivity (%) = [product] / [sum of all products]
Yield (%) = Conversion (%) × Selectivity (%) ÷ 100
[…] means an amount of the substance at each temperature determined from FID or TCD measurement.
[…]rf means an amount of the substance determined from FID or TCD measurement without any catalysts.
4-2-4. IR measurement
In situ transmission infrared spectroscopy (TIRS) was applied to monitor the transient adsorption-desorption dynamics of acetic acid and ethanol on the catalytic materials. 3.5 mg of the catalyst powder was pressed into a round-shaped disc (5 mm in diameter) and placed onto a flow-through transmission cell made of quartz glass (Makuhari Rikagaku Garasu Inc.) with ZnSe windows. The ZnSe windows were heated at 60 oC with a thermostat to prevent the
69
condensation of gaseous components such as acetic acid and ethanol. Prior to IR measurements the mounted catalyst disc was pretreated in a helium flow at 300 mL/min to remove adsorbed species (mainly H2O).
In situ TIRS spectra were recorded on a INVENIO R spectrometer (Bruker Optics) equipped with a liquid nitrogen cooled mercury-cadmium-telluride (MCT) detector (D316, ZnSe Window) and an optical filter (F321). Spectra were recorded at 4 cm-1 of the spectral resolution and 60 kHz of the scanning velocity with 64 scans per spectrum. Modulation-excitation spectroscopy (MES) was combined with in situ TIRS by periodically changing between two different gas effluents; acetic acid + ethanol in He balance ↔ acetic acid in He balance at 300 mL/min. Switching between these effluents was repeated seven times. After reaching reproducible responses after two cycles, only the spectra of the last five cycles were averaged into one cycle to enhance the signal-to-noise (S/N) ratio and time resolution. The last spectrum in the period of acetic acid in He balance was used as a reference background. Phase-sensitive detection (PSD) was used to further remove the noise and to obtain kinetic information of responding surface species. The acquired time-domain spectra were mathematically treated by PSD to obtain phase-domain spectra according to the following equation:
𝐴𝑘(𝑣̃ ) cos (𝜑𝑘+ 𝜑𝑘𝑑𝑒𝑙𝑎𝑦(𝑣̃̃)) =2
𝑇∫ 𝐴(𝑡, 𝑣̃̃)0𝑇 sin(𝑘𝜔𝑡 + 𝜑𝑘) d𝑡 (1)
where T is the length of a cycle, ω is the demodulation frequency, φk is the demodulation phase angle, k is the demodulation index (k = 1 in this study), and A (t, ṽ) and Ak(ṽ) are the active species responses in the time and phase domains, respectively.
4-3. Results and Discussion
No one knows the stability of a MOF to acetic acid even though there are few reports about MOF’s tolerance against kinds of solvent10, acid11, 12, 13 and base.11, 12. Then, as described in chapter 2, I prepared 12 kinds of MOFs (MIL-125-NH2, UiO-66-NH2, HKUST-1, MIL-101,
Zn-70
MOF-74, Mg-MOF-74, MIL-121, MIL-125, UiO-66, ZIF-8, ZIF-67 and DUT-5) as candidates which were known to have high thermal tolerance. These MOFs were expected to exhibit a different type or strength of interaction with acetic acid due to their ligands with different functional groups (e.g. -COOH of MIL-121 and –NH2 of UiO-66-NH2). Especially, I expected that the MIL-125-NH2 and UiO-66-NH2 which have basic side chain, i.e. -NH2, have an influence on acetic acid molecules during the hydrogenation reaction.
Tolerance of MOFs against acetic acid vapor with high temperature (about 200–300 C) should be examined before applying the MOFs to acetic acid hydrogenation reaction. In the beginning, the stability of the MOFs against acetic acid vapor was briefly evaluated using XRD measurement while an exposing MOFs experiment to the acetic acid vapor, which was conducted as a screening test. From comparison of the XRD patterns before and after the exposure as shown in Figure 4-6, it is confirmed that the XRD patterns of MIL-125-NH2, UiO-66-NH2, HKUST-1, MIL-101, Zn-MOF-74, Mg-MOF-74 and MIL-121 remain its pattern even after the exposure to acetic acid vapor. This result indicates that these MOFs maintain its structures even in the acetic acid atmosphere. On contrary, a missing or broadening of peaks were observed on the XRD pattern of MIL-125, UiO-66 and DUT-5. This result indicates that the structure of MIL-125, UiO-66 and DUT-5 were broken by the acetic acid vapor. Note that the XRD patterns of ZIF-8 and ZIF-67 were not measured due to complete dissolution of their crystal by the acetic acid vapor. From these results, I firstly found that seven types MOFs, i.e.
MIL-125-NH2, UiO-66-NH2, HKUST-1, MIL-101, Zn-MOF-74, Mg-MOF-74 and MIL-121 exhibit a certain stability against acetic acid vapor. I also found that MIL-125, UiO-66, DUT-5, ZIF-8 and ZIF-67 are too weak against the acid vapor. The result of the screening experiment is summarized in Table 4-1. Interestingly, MIL-125-NH2 and UiO-66-NH2 showed the acetic acid tolerances although MIL-125 and UiO-66 did not exhibit it. Note that the type of functional groups of ligands is the only difference between MIL-125– MIL-125-NH2 and66– UiO-66-NH2. This result suggests that a functional group of ligand, i,e, –NH2, play a significant role
71
on the acetic acid tolerance of MOFs. Considering that the –NH2 can act as base whereas the acetic acid is a monoprotic acid, a kind of an acid–base interaction between them may happen when the MOFs with –NH2 is exposed to an acetic acid atmosphere. I suggest that the coordinate bond in MIL-125 and UiO-66 was protected from an attack of acetic acid molecules by an acid–base interaction of –NH2 in their ligands.
72
Figure 4-6a. XRD spectra (λ=1.080Å) of (a) MIL-125-NH2, (b) UiO-66-NH2, (c) HKUST-1, (d) MIL-101, (e) Zn-MOF-74 and (f) Mg-MOF-74 before (blue) and after (red) the exposure into acetic acid vapor.
73
Figure 4-6b. XRD spectra (λ=1.080Å) of (a) MIL-121, (b) MIL-125, (c) UiO-66 and (d) DUT-5 before (blue) and after (red) the exposure into acetic acid vapor
74
Table 4-1. Result of acetic acid tolerance test of MOFs
However, the MOFs were putted in atmospheric humidity before and during the brief experiment. Therefore, I cannot assure that the acetic acid vapor was replaced with the pre-adsorbed water molecule. All I can assure is that the 5 MOFs (MIL-125, UiO-66, DUT-5, ZIF-8 and ZIF-67) are very weak against the acetic acid vapor from the acetic acid tolerance test.
Then, in situ XRD measurement of the seven types of MOFs (MIL-125-NH2, UiO-66-NH2,
HKUST-1, MIL-101, Zn-MOF-74, Mg-MOF-74 and MIL-121) with completely degassing was conducted using a synchrotron radiation at SPring-8. In this experiment, the degassing and acetic acid–introduction of MOFs were performed in a glass capillary without exposure to atmospheric air. The only MOF samples and acetic acid vapor were sealed in the capillary. XRD measurements for the samples and the same ones without acetic acid were performed as shown in Figure 4-7. The XRD patterns for seven types of MOFs remained even after exposure to acetic acid. From this result, the acetic acid tolerance of the seven types of MOFs were confirmed more precisely.
75
Figure 4-7. XRD patterns (λ=1.080Å) of before (blue) and after (red) acetic acid introduction into (a) Pt/MIL-125-NH2, (b) Pt/UiO-66-NH2, (c) Pt/MIL-101, (d) Pt/HKUST-1, (e) Pt/Zn-MOF-74, (f) Pt/Mg-MOF-74 and (g) Pt/MIL-121.
76
Subsequently, in situ temperature dependent XRD patterns were also measured in the range of 25–500 °C using the samples with acetic acid as the same used in the in situ XRD measurement.
All the XRD patterns of the MOFs remained at 200 °C even in the acetic acid atmosphere as shown in Figure 4-8. In particular, the obtained XRD patterns of MIL-125-NH2, UiO-66-NH2, Mg-MOF-74 and MIL-121 maintains even at over 350 °C. From this result, I found that the seven types of MOFs can maintain its structure at least 200 °C even in the acetic acid atmosphere. I also found that the intensity of XRD peaks in several MOFs were varied by the heating. Especially, the intensity of almost all of peaks in MIL-101 drastically decreased at 25 and 50 °C. Moreover, the peaks were revived in the range of 75–250 °C. This result suggests that the peak intensities of MIL-101 are weakened by the acetic acid adsorption and recovered by the desorption. Similarly, despite of the drastic results, i.e. imcomplete recovery of the peak intensities was obtained on MIL-125-NH2, HKUST-1, Mg-MOF-74 and MIL-121.
77
Figure 4-8a. In situ temperature dependent XRD patterns of (a) MIL-125-NH2, (b) UiO-66-NH2, (c) HKUST-1 and (d) MIL-101
78
Figure 4-8b. in situ temperature dependent XRD spectra of (a) Zn-MOF-74, (b) Mg-MOF-74 and (c) MIL-121
Loading Pt NPs on the seven types of MOF supports was conducted using arc plasma deposition method as shown in Chapter 2. Note that this method ensures that the loaded Pt NPs directly contact with the MOFs without any reagents. The resulting samples were called as
79
Pt/MIL-125-NH2, Pt/UiO-66-NH2, Pt/HKUST-1, Pt/MIL-101, Pt/Zn-MOF-74, Pt/Mg-MOF-74 and Pt/MIL-121. To compare with one of the best catalyst support among basic metal oxides, TiO2 (Degussa, P-25) was chosen according to a literature3. γ-Al2O3 was also selected as a reference which is unsuitable for the acetic acid hydrogenation reaction3.
The all prepared catalysts for the acetic acid hydrogenation reaction were characterized by XRD, STEM-EDS and ICP-AES as described in Chapter 2. In addition, the number of active sites of Pt NPs in the catalysts were evaluated by H2-pulse chemisorption. The obtained characterization data are summarized in Table 4-2.
Table 4-2. Characterization data of Pt/MOFs, Pt/TiO2 and Pt/Al2O3.
80
To evaluate acetic acid adsorption properties of these MOFs, I chose determining desorption temperature of acetic acid molecules as a measure of acetic acid adsorption abilities of them. In particular, I aimed to judge whether the MOFs can keep holding acetic acid molecules on its surface at around reaction temperature or not. The desorption temperatures were determined using TPD-MS. In the measurements, the desorption of pre-introduced acetic acid from each MOF was detected by mass spectrometry. Figure 4-9 shows the result of TPD-MS measurement of the MOFs. Each peak corresponds to the desorption of pre-introduced acetic acid. As a result, I firstly demonstrated that the desorption temperatures on MOFs are different from the type of MOFs, in particular, MIL-125-NH2 and UiO-66-NH2 show notable desorption peaks at relatively high temperatures (about 140 C and 100 C, respectively) nonetheless most of the MOFs could not keep holding acetic acid molecules at high temperatures. This means acetic acid adsorption properties of MOFs are different from the type of MOFs and, in particular, MIL-125-NH2 and UiO-66-NH2 have relatively high adsorption abilities for acetic acid. The high adsorption abilities of the two MOFs may come from acid-base interaction (or hydrogen bond) between –NH2 and acetic acid. In acetic acid hydrogenation reaction, as previously reported,2,3,8,9 the product selectivity is highly dependent on the adsorption ability of catalytic supports of Pt NPs. Rachmady et al. reported that TiO2 which can adsorb acetic acid is one of the suitable for ethanol production. From this point of view, the two MOFs which showed high acetic acid are promising for ethanol production from acetic acid hydrogenation in gas phase.
81
Figure 4-9. Mass spectra at M = 60 of (a) MIL-125-NH2, (b) UiO-66-NH2, (c) HKUST-1, (d) MIL-101 and (e) Zn-MOF-74, Mg-MOF-74 and MIL-121, which had been introduced the acetic acid vapor. The components and the structures of the MOFs are also shown as inset.
The Pt/MOFs (Pt/MIL-125-NH2, Pt/UiO-66-NH2, Pt/MIL-101, Pt/HKUST-1, Pt/Zn-MOF-74, Pt/Mg-MOF-Pt/Zn-MOF-74, Pt/MIL-121), Pt/TiO2 and Pt/Al2O3 were applied for the acetic acid hydrogenation reaction using home-made fixed bed flow reactor. All samples were pretreated at 200 C for 2h under H2 flow to activate the surface of Pt NPs. To promote the reaction, a
82
pressure of 10 atm was applied during the experiment. The all products were detected by using two gas chromatographs with a TCD detector or a FID detector. Figure 4-10 indicates the acetic acid conversions of the catalysts on the acetic acid hydrogenation. As expected from the result of TPD-MS measurement and previously reports, Pt/MIL-125-NH2 and Pt/UiO-66-NH2
showed remarkable activities (almost 100% conversion over 250 C) nevertheless the other Pt/MOFs showed very low activities. In comparison with Pt/TiO2 which is one of the best catalyst, I found that Pt/MIL-125-NH2 and Pt/UiO-66-NH2 exhibited high activities as well as Pt/TiO2, and moreover, Pt/MIL-125-NH2 showed a superior catalytic activity compared to the Pt/TiO2 in low temperature region. In addition, Pt/MIL-125-NH2 showed the highest yield of ethanol among all samples in all temperature region as shown in Figure 4-11. This is the first example that Pt/MOFs are applied to acetic acid hydrogenation reaction. Considering that the other Pt/MOFs showed very low activities as the same as Pt/Al2O3, an acetic acid adsorption ability of a support appears to take part in the catalytic activity of the loaded Pt NPs.
Figure 4-10. Acetic acid conversions of Pt/MIL-125-NH2, Pt/UiO-66-NH2, Pt/MIL-101, Pt/HKUST-1, Pt/Zn-MOF-74, Pt/Mg-MOF-74, Pt/MIL-121, Pt/TiO2 and Pt/Al2O3 at 10 atm.
83
Figure 4-11. The yields of ethanol on Pt/MIL-125-NH2, Pt/UiO-66-NH2, Pt/MIL-101, Pt/HKUST-1, Pt/Zn-MOF-74, Pt/Mg-MOF-74, Pt/MIL-121, Pt/TiO2 and Pt/Al2O3 at 10 atm.
The catalytic activities comparable, or superior, to those on Pt/TiO2were demonstrated using Pt/MIL-125-NH2 and Pt/UiO-66-NH2. To confirm whether the high catalytic activities are derived from the interaction between the Pt NPs and the MOFs having the acetic acid–
adsorption ability, the samples of MIL-125-NH2 and UiO-66-NH2 without Pt NPs were applied for the acetic acid hydrogenation reaction. Figure 4-12a and b represent the temperature dependence of acetic acid conversions on Pt/MIL-125-NH2/MIL-125-NH2 and Pt/UiO-66-NH2/UiO-66-NH2. Both of results were compared to that on Pt/Al2O3 as a reference. The catalytic activity on Pt/MIL-125-NH2 and Pt/UiO-66-NH2 were much superior to those on MIL-125-NH2 and UiO-66-NH2, respectively. The MIL-125-NH2 produced a small amount of ethyl acetate (as shown in Figure 4-12c and e) whereas the UiO-66-NH2 produced acetone (as shown in Figure 4-12d and f) with almost 100% selectivity, respectively. According to literatures3 and my result described below, the acetone production, i.e. ketonization on the surface is ascribed to the behavior of a material like TiO2 which can adsorb acetic acid in the reaction condition.
84
On contrary, the 100% selectivity of ethyl acetate production without metal NPs has not been reported as far as I surveyed. At least, this is the first example of synthesis of ethyl acetate from acetic acid using MOFs.
Pt/Al2O3 showed a very low catalytic activity even though the catalyst involves Pt NPs. The result suggests that the remarkable catalytic activity of Pt/MIL-125-NH2 andPt/UiO-66-NH2is originated from a synergistic effect of Pt NPs and MOF support which can adsorb acetic acid molecules on its surface. In other words, the high acetic acid hydrogenation ability is possibly to be derived from the interface between Pt NPs and MIL-125-NH2 or UiO-66-NH2, which means the most of catalytic active sites exist at the interface. Considering the strength of acetic acid adsorption properties, the catalytic activities at the interface between Pt NPs and MOFs were supposed to depend on the adsorption abilities of MOFs. From these results, I first systematically demonstrate that the support effect originating from substrate adsorption effect of MOF.
85
Figure 4-12. The temperature dependence of acetic acid conversions on (a) Pt/MIL-125-NH2, MIL-125-NH2 and Pt/Al2O3, and (b) Pt/UiO-66-NH2, UiO-66-NH2 and Pt/Al2O3. The yield of products on (c) MIL-125-NH2 and(d) UiO-66-NH2. The selectivity of products on (e) MIL-125-NH2 and(f) UiO-66-NH2.
Figure 4-13 shows that the catalytic conversions and selectivities of Pt/MIL-125-NH2, Pt/UiO-66-NH2 and Pt/TiO2, which exhibited high activities.From comparison of Pt/UiO-66-NH2 and Pt/TiO2, I found that their behaviors of the reaction are very similar besides the conversion
86
behavior in the following points: 1) the ethanol production is rapidly declining whereas the one of ethyl acetate rapidly increases in the region of low–middle conversion, i.e. 140–220 C. Note that the yield of ethanol was low while the selectivity of the ethanol was very high at below 140
C. The ethyl acetate is produced by an esterification of the unreacted acetic acid and the ethanol from the acetic acid hydrogenation. Therefore, the reaction behavior suggests that the Pt/UiO-66-NH2 and Pt/TiO2 produce a lot of ethanol even in the low temperature region, i.e. near 140-160 C although the most of the produced ethanol are consumed by a reaction with the unreacted acetic acid. 2) The ethyl acetate production is steeply decreased as the conversion reaches near 100%. This result indicates that a reactant of the esterification, i.e. acetic acid, runs short by the very high conversion at the high temperature. 3) After the accomplishment of the esterification, ethane production is strongly enhanced. Ethane is derived from the over reduction of ethanol, which is driven by extremely high temperature. As a result of these points, the high-selective synthesis of ethanol is not achieved in all temperature range. In other words, the high-selective synthesis of ethanol is strongly hindered by the reaction of the unreacted acetic acid and the produced ethanol.
On the other hands, the reaction steps or mechanism on Pt/MIL-125-NH2 which showed the highest ethanol production amount is completely different, including that ethanol was the main product in almost all temperature range. Especially, the yield of ethyl acetate was much less than that on the other catalysts. Therefore, I predict that the reaction of ethyl acetate production on Pt/MIL-125-NH2 is strongly suppressed by a reason, which brings the superior activity to Pt/MIL-125-NH2. Therefore, I succeed in discovery of the most suitable support for ethanol synthesis from acetic acid hydrogenation.