利用各式光譜技術研究液相分子於固相表面之吸附與反應

Application of Spectroscopy Techniques to Measure the Adsorption and Chemical Reaction of Molecule Between Liquid and Solid Phase

范秀芳(Hsiu-Fang Fan)

指導教授 ： 林金全

國立臺灣大學/理學院/化學研究所/博士(2006年)

https://doi.org/10.6342/NTU.2006.00137

摘要

在第一部份的實驗當中，主要是針對一異相催化反應設計一光譜分析技術。在此部分的實驗中，主要偵測分析物於近紅外光的吸收圖譜，配合多變數分析及Matlab program同時取得此異相反應之催化反應速率及此分析物於反應系統中之擴散係數。所觀測之反應為Qudricyclane 在氯仿中以硫酸銅或氯化錫為催化劑轉變成Norbornadiene之過程。由本試驗所取得之反應速率為(3.7±0.1) x 10^-3 和(3.8±0.1) x 10^-3 s-1 分別對應是在硫酸銅及氯化錫之狀況.進而對硫酸銅系統下做溫度效應及催化劑粒徑大小分析：二級反應速率為 (8.47±±5.09)) x 10^-6 s-1M-1 ；三級反應速率為(2.19±0.77)x10^-8 s-1M-2;而此反應對應之反應活化能為24.8kJ/mol（一對一）286.2 kJ/mol（一對二）.
在第二部分的實驗當中，凝態腔體震盪吸收光譜技術被設計用於偵測分析物於介面之特性。此技術利用漸進波偵測介面物種之吸附情形。DP和DMPI吸附於介面之作用力，利用triethylamine滴定實驗間接證實：DP是利用氫鍵與矽氧群接合，其吸附平衡常數為(1.3±0.2)10^4 M-1；DMPI是利用靜電吸引力與矽氧群接合，其吸附平衡常數為(2.6±0.7)10^4 M-1。而在第二個實驗當中，Crystal Violet 被用於探測介面上矽氧分佈之特性，由此實驗中證實介面上之矽氧群有兩種不同的分佈情形，其分別之吸附平衡常數為(1.7±0.1)10^5 M-1, 及(1.2±0.2)10^3 M-1。第三個實驗中利用電化學之電雙層模型配合競爭蘭穆爾模型去求得Rhodamine B 於介面之吸附常數為：1.3x10^4±8.4x10^2, 6.3x10^3±7.9x10^2。
第三部份的實驗當中，螢光相關光譜被建立及測試。兩個波長下的激發體積及儀器參數均被取得：0.53fL和3.88 是在532nm之激發波長：0.83fL和4.63是在630nm之激發波長。

關鍵字

催化反應 ； 腔體震盪光譜 ； 螢光相關光譜

並列摘要

Part I
(A)
By using Fourier transform near infrared (NIR) absorption spectroscopy with the aid of a kinetic model, we have investigated the conversion of quadricyclane to norbornadiene as catalyzed by anhydrous CuSO4 and SnCl2 in chloroform. The reaction mixture is kept still to avoid the effect of sample heterogeneity. The NIR absorption spectra are acquired, at a position 2 mm above the catalyst surface, at 30-second intervals for a reaction period of 4 hours. The related concentrations of quadricyclane and norbornadiene in the temporal evolution are determined with the analysis of partial least squares. The evolution of quadricyclane, as numerically solved from the model, is expected to describe its behavior more accurately in the catalytic system than that obtained previously. In addition to the isomerization rate, the kinetic model takes into account the contribution of diffusion motion. The diffusion coefficients of quadricyclane can be explicitly determined to be 3.8x10-5 cm2s-1 in chloroform and 1.14 x10-5 and 2.85x10-6 cm2s-1 inside the CuSO4 and SnCl2 stacks, respectively. The diffusion motion is slowed down inside the solid stacks and thus the molecular mechanism cannot be suitable for this system. Given the diffusion coefficients, the pseudo-first order depletion rate constants are evaluated to be 3.7x10-3 and 3.8x10-3 s-1 for CuSO4 and SnCl2, respectively. The corresponding second-order depletion rate constant is determined to be 6.6x10-6 and 2.0x 10-6 s-1M-1 by considering the density and pore size of the catalyst; M denotes the total catalyst surface per unit effective volume of solvent. The reaction rates are found to be a factor of two slower as compared with others obtained in a continuously stirred solution. In the surface-mediated reaction, the catalytic isomerization is subjected to a one-site coordination (1:1 complex) between the reactant and the catalyst. Nevertheless, a two-site coordinated reaction cannot be justified unless a pore size dependence of the depletion rate is discerned.
(B)
The related concentrations of quadricyclane and norbornadiene in the temporal evolution are determined with the analysis of partial least squares. The concentration of quadricyclane, as numerically solved from the model, is expected to describe its behavior more accurately in the catalytic system than that obtained previously. In addition to the isomerization rate, the kinetic model takes into account the contribution of diffusion motion. In order to look into the motion inside the catalyst more accurately, the simulation model is developed into two-dimensional form. The branch model can describe the distribution of catalyst and solvent within catalyst layer. Under this model, the simulation result is as the same as that under one-dimensional model. In addition to branch model, another two-dimensional model that the distribution of catalyst and solvent within catalyst layer is at random is also built. Although the distribution within catalyst layer much conforms to the real condition, the problem of time and money consuming is much serious. Under these considerations, we adapt the branch model to simulate the experiment date in order to evaluate the diffusion coefficient, which is set the same both in solvent layer and catalyst layer. In the surface-mediated reaction, the catalytic isomerization is subjected to a one-site coordination (1:1 complex) between the reactant and the catalyst. Nevertheless, a two-site coordinated reaction cannot be justified unless a pore size dependence of the depletion rate is discerned. From the following experiments, we plot the pseudo-first order depletion rate constants vice the 6Wt/daVeff. The second-order depletion rate constant 8.47x10-6 s-1M-1 and third-order depletion rate constant2.19x10-8 s-1M-2 could be evaluated. When the ln(k) is plotted vice 1/RT, the slope is . Then the activation energy is acquired. The activation energy of one-site path is calculated to be 24.8 kJ/mol and that of two-site path is 286.2 kJ/mol. The value is assumed to be exact right because we get the almost the same value through another procedure.
Part II
(A)
An evanescent wave cavity ring down absorption spectroscopy is designed to measure the thermodynamic properties on the surface adsorption. Neutral trans-4-[4-(dibutylamino)-styryl]-1-(3-sulfopropyl) -pyridinum(DP) and charged trans-4-[4-(dibutylamino)styryl]-1-methyl
pyridinium iodide (DMP+I) as adsorbates are flowed through the CH3CN/fused silica interface. The surface density of adsorption at the interface is determined by measured absorbance. The concentration dependence of the surface density may be characterized with Langmuir isotherm model. The fit to the adsorption measurements at the interface yields saturated surface density, equilibrium constant and free energy of adsorption to be 7.0x10^13 cm-2, 1.3x10^4 M-1, and -23.5kJ/mol, respectively, for DP and 8.9x10^12cm-2,2.6x10^4 M-1, and -25.2kJ/mol for DMP+I. The DP is adsorbed to the SiOH sites by forming hydrogen bond, while the DMP+ cation is bound to the SiO- sites by electrostatic attraction. The force variety may be verified by addition of triethylamine(TEA), which is competitive with DP for the silanol site.
(B)
Combination of evanescent wave cavity ring down and dichroism absorption technique is a good way to probe the orientation of adsorbed molecule on interface. Crystal violet which has strong absorbance at 600nm is chosen to be measured in this experimental setup. The nearly flat structure and bulky size makes itself a good surface-geometry probe. The binding phenomenon of crystal violet on CH3CN/ fused silica interface could be defined into two types which could be approved by studying the adsorption plot. The fit to the adsorption measurements at the interface yields equilibrium constant and free energy of adsorption to be 1.7x10^5 M-1, and -30.1kJ/mol, respectively, for CV+ and type I silanol site and 1.2x10^3 M-1, and -17.9kJ/mol for CV+ and type II silanol site. By analyzing the tilt angle plot, the adsorption between CV+ and type I site could be said that the CV+ approach the type I SiO- freely and randomly because the average tilt angle reaches magic angle~35.3 degree. As the concentration increases, the tilt angle increases. It means that CV+ would get closer and closer to the surface as the surface density increases. This trend is seen both on CH3CN/fused silica and thin film/fused silica conditions.
(C)
Using Double layer model to acquire the adsorption phenomenon of R610 on water/fused silica surface has been done. The adsorption equilibrium constants are 1.3x104±8.4x102, 6.3x103±7.9x102 and 2.6x102±8.2x101 for R610 neutral form, R610+ charged form and Na+, irrespectively. By using double layer model, the information for both type of R610 molecule on the silica surface could be solved.
Part III
The FCS system is verified under our experimental setup by using R6G and Cy5. It is checked by acquired the FCS at fast time scale and slow time scale. The reasonable results have been acquired and compared with the theoretical value and the results from other literatures. In order to acquire high qualified FCS spectrum, the alignment and the excitation power should be optimized to prevent the misinterpretation. In our experimental setup with 532nm excitation light source, 1.25 oil objective and 63 um optical fiber and R6G as analyst, the optimized pump energy is about 60uW in order to get the best signal to noise ratio and better qualified FCS spectrum. The instrumental parameters under above experimental condition are 3.88 and 4.63 for 532nm and 630nm, irrespectively. For further experiment, the instrumental parameter should be fixed in order to acquire the diffusion time or bunching time of analyst for doing further studies.

並列關鍵字

catalytic reaction ； cavity ring down spectroscopy ； fluorescence correlation sepctroscopy

參考文獻

2. Whitten, D. G. J. Am. Chem. Soc. 1999, 121, 8146.

3. Michael, K. E. J. Am. Chem. Soc. 1994, 116, 6881.

4. Williams, D. J.; Handley, L. J. Am. Chem. Soc. 1990, 112, 7083

5. Kano, K. Chem. Lett. 1980, 421.

8. Winkler, T.; Dix, I.; Jones, P. G., Herges, R. Angew. Chem. Int. Ed. 2003, 42, 3541.