3.1. Effective synthesis procedure
Fig. 3 indicates the SEM images of Cu and Zn- MOF samples, which have been synthesized by the IM and UAIM methods. In samples A and C, which are synthesized by the IM method, the particles have a tendency for severe agglomeration, which caused non-uniform morphology with a spherical shape in the structures. Also, the particle size distribution in these samples is in the bulk range. Although, the evidence of particle agglomeration can be seen in some cases of the samples synthesized by the UAIM method, but the dominant distribution of these particles is uniform (Fig. B and D). The morphology of these compounds is more likely rod-shaped in comparison with the samples synthesized by IM method. These images represent the less mean particle size distribution in the UAIM method than the IM method, so that the Cu and Zn- MOF samples synthesized by this method have mean particle size of 35 and 45 nm, respectively. According to the conditions used to synthesize these samples, it seems that the UAIM method leads to less particle size distribution, synthesis of samples with minimum agglomeration and uniform morphology. The homogenous morphology and small particle size distribution of the Cu- and Zn- MOF nanostructures developed by UAIM procedure in this method are remarkable compared to previous samples [17, 18]. Therefore, the UAIM has been chosen as an appropriate and effective technique compared to IM for the synthesis of the metal organic framework nanomaterials.
3.2. Structural formula
Fig. 4 shows the XRD patterns of Cu and Zn- MOF samples synthesized in optimal condittioms of UAIM methods. These patterns properly show the formation of Cu-MOF (JCPDS cards.no: 05-0661) and Zn-MOF (JCPDS cards.no: 234578). In both patterns, the wider and intense peaks in comparison with samples that have been already synthesized by different methods, are the strong evidence for the smaller crystalline sizes of the Cu and Zn- MOF samples synthesized in the present study [15, 19]. Regarding the UAIM method used in this study, it seems that this novel method has produced samples with desirable crystallographic properties. Furthermore, according to the data indexed from the XRD patterns, Cu and Zn- MOF samples have monoclinic and cubic crystallin structures, respectively.
The Co- and Cu-MOF samples synthesized by two different methods were characterized using the FTIR spectra (Fig. 5). In both samples, the absorption band at 3400 to 3500 cm−1 may be attributed to the coordinated water in the products [20]. The frequency peaks at 3000 cm−1 are ascribed to the stretching vibration of aromatic C-H. The absorption peak observed near 2600 to 2500 cm_1 confirms the presence of -COO- groups of the ionized ligand and the peak around 1400 to 1200 cm_1 is related to C-N bonds [21]. The absorption band near 900 cm_1 is corresponded to the asymmetric and symmetric stretching vibrations of aliphatic CH and the peaks at 400 to 600 cm_1 are assigned to the Cu- and Zn-O bonds. Based on results obtained from FTIR spectra and according to different configurations of the linkers [22], the structures of Figure 6 were suggested for Cu- and Zn-MOF nanostructures samples.
3.3. Desirable MOF-adsorbent
According to the N2 adsorption/desorption isotherms shown in Fig. 7, the Cu-MOF sample exhibited type I isotherm indicates the microporous distribution of pores [23], whereas the isotherms of Zn-MOF sample are similar to the type III, which shows the mesoporous behaviour in this sample [24]. According to the results of the BET technique, the Cu-MOF sample has a surface area of 410 m2/g with volume pore of 0.021 cm3, while the corresponding values for Zn-MOF were 1145 m2/g and 0.097 cm3, respectively.
The BJH method also approves the microporous and mesoporous distribution for Cu- and Zn-MOF samples, respectively. So, in accordance to the information obtained from this method, the mean pore size is 1.94 nm for Cu-MOF and 2.59 nm for Zn-MOF sample (See Fig. 8). Although, the synthesis technique developed for these samples is the same, but the cause of differences in the textural properties of these samples can be attributed to the type of applied MOF, and as a result, the variety of structural configurations that were discussed in the previous section. Since choosing the adsorbents with desirable textural properties is particularly important, thus, the Zn-MOF nanostructures synthesized in this study were selected as a novel candidate in order to study adsorption applications for Arsine gas adsorption.
3.4. Experimental design
Zn-MOF samples synthesized by the UAIM method, with regard to distinctive features, were used as a novel candidate for arsine gas adsorption. The adsorption studies of the process underwent systematic design using fractional factorial method. Since the gas adsorption process was studied by volumetric method, thus, the effective factors (Table 1) of which were selected based on previous studies including adsorbent dosage (A), temperature (B) and pressure (C) [25]. Considering these experimental parameters, 18 runs were carried out, and the resulting adsorption responses are shown in Table 2 (Each experiment was performed two replicates).
Table 1. Coded and uncoded levels of adsorbent dosage, temperature and pressure of UAIM method for fractional factorial design.
Table 2: Randomized complete fractional factorial design for arsine gas adsorption experiments of Zn-MOF prepared by UAIM method.
Sample
(Level)
|
Std
order
|
Center Pt
|
A
(mg)
|
B
(°C)
|
C
(bar)
|
REP
|
Adsorption
(mmol/g)
|
a
|
9
|
1
|
+1
|
+1
|
-1
|
1
|
64.4
4.2
|
2
|
b
|
5
|
1
|
-1
|
+1
|
0
|
1
|
73.8
3.9
|
2
|
c
|
6
|
1
|
-1
|
0
|
-1
|
1
|
5.2
5.3
|
2
|
d
|
3
|
1
|
0
|
0
|
+1
|
1
|
7.1
6.8
|
2
|
e
|
2
|
0
|
+1
|
-1
|
1
|
1
|
8.2
8.3
|
2
|
f
|
8
|
1
|
0
|
+1
|
-1
|
1
|
3.1
2.0
|
2
|
g
|
4
|
1
|
0
|
-1
|
1
|
1
|
8.1
8.0
|
2
|
h
|
7
|
1
|
-1
|
+1
|
-1
|
1
|
1.7
1.4
|
2
|
i
|
1
|
1
|
+1
|
0
|
+1
|
1
|
7.8
7.6
|
2
|
3.5. Systematic study of procedure
Fig. 9 shows a different residual plot for adsorption studies of the Zn-MOF samples. Since the positive and negative levels are approximately equal in all of these plots, thus, it is concluded that dispersions of adsorption experiments are quite randomized and chance of each one is equal to another. The above cases approve the scientific design of experiments for adsorption studies [26].
The analysis of variance is used in order to study the effects of different experimental parameters including temperature, adsorbent, and pressure on adsorption response. As reported in Table 3, although all three studied parameters affect the arsine gas adsorption, but with regard to Pvalues obtained for each, the effect of temperature is more significant than the other parameters. Also, the effects of experimental parameters and the interaction among them are confirmed by Pareto charts in Fig. 10.
According to the arsine adsorption results reported in Table 2, conditions e, g with highest adsorption rate were selected as optimal conditions. Although, the adsorption rate in condition e is higher than condition g, but the difference has not such an impact on adsorption results. Regarding the effect of pressure on gas adsorption, which has been evaluated in earlier studies, since the lowest value is selected for pressure in conditions f and h, thus, the arsine gas adsorption is significantly reduced. In conditions a and b, with same temperature and different adsorbent and pressure, the adsorption results show the high arsine adsorption rate for condition a. Since in this case, the adsorbent rate is greater than the pressure, thus, the difference in the adsorbent dosage has a more effect on the adsorption. This difference is in accordance with the results of the analysis of variance, which approves the great effect of the adsorbent dosage than the pressure. The adsorption values in c, d, and i are varied depending on the difference in values of experimental parameters in these cases compared to the optimal conditions.
Table 3 Analyses of variance for arsine gas adsorption of Zn-MOF samples synthesized by UAIM method.
Source
|
DF
|
Seq SS
|
Adj SS
|
Adj MS
|
Pvalue
|
A
|
1
|
93.924
|
88.5639
|
29.5213
|
0.004
|
B
|
1
|
86.44
|
66.97
|
25.14
|
0.01
|
C
|
1
|
68.09
|
36.78
|
13.42
|
0.04
|
A*B
|
1
|
49.57
|
20.12
|
17.39
|
0.06
|
A*C
|
1
|
79.14
|
59.12
|
19.74
|
0.03
|
B*C
|
1
|
87.45
|
69.22
|
27.41
|
0.009
|
R-Sq: 97.24% R-Sq(pred): 99.05% R-Sq(adj): 98.44%
|
3.6. Optimization parameters
Regarding the fractional factorial design evaluated in this study and taking into account the regression equations obtained as an output, three-dimensional images are shown in Fig. 11, which show the relationship between experimental parameters related to the arsine gas adsorption process. These images are in good agreement with the results of Table 2. With respect to the regression equation of adsorption (ADS: 1158-514A-686B-500C+44D), we able to conclude that the proposed theoretical model of this study confirms the experimental results presented in Table2.
Since the purpose of this study is to find adsorbents with high adsorption rates, RSM optimization has been used, and values for each of the experimental parameters (adsorbent dosage, temperature and pressure) with the desirability of 0.9 were depicted in Fig. 12. According to the formula described in Table 1, these values are converted to non-coded values which data is reported in Table 3. By comparing the predicted values of arsine gas adsorption by RSM with previous samples, it seems that the adsorbents developed in this study have a higher adsorption rate than recently adsorbent such as carbon materials [27], graphene oxide [28], copper-exchanged zeolite [29], monolayer MoS2 [30], Hf2CO2 monolayer [7] and carbon nanotubes [31] (Fig. 13). Increasing the efficiency of Zn-MOF sample developed in this study are related to the systematic studies as well as optimization procedure which these superiorities could distinguish these adsorbents from previous adsorbents.
Table 4 Response optimization the experimental parameters for Arsine gas adsorption obtained by RSM results.
Response
|
Goal
|
Lower
|
Target
|
Upper
|
Experimental parameters
|
Desirability
|
Predict response value
|
A
(mg)
|
B
(°C)
|
C
(bar)
|
Adsorption (mmol/g)
|
maximize
|
8.20
|
9.00
|
12.00
|
0.032
|
25.00
|
4.34
|
0.9071
|
8.74
|