具有硬桿鏈段之共聚高分子自組裝囊胞的研究:耗散粒子動力學法

Abstract

高分子囊胞(polymersome)目前受到很多的關注與研究，主要原因是因為高分子囊胞有極大的潛在發展性，未來能夠應用在微反應器、藥物釋放以及細胞膜模擬上，是一種新穎的材料。而在製備上，高分子囊胞通常能由雙段鏈或多段鏈雙親性高分子自組裝聚合而成。本論文是選用具有硬桿鏈段之共聚高分子所自組裝的囊胞來作為研究對象，模擬的方法則是耗散粒子動力學法(dissipative particle dynamic, DPD)，本篇文章的重點將放在對於不同構造的高分子自組裝成高分子囊胞的條件與其不同的相行為，進而討論各自所形成的高分子囊胞之微結構性質、機械性質以及與未來應用很相關的膜融合動力學行為。 在論文的第一部份中，我們利用耗散粒子動力學法探討硬桿-軟鏈共聚高分子(RxCy)在疏硬桿溶劑下的相行為，針對不同長度的組合，在特定濃度下成功找出能夠形成高分子囊胞的結構條件。發現當硬桿鏈段太長時，硬桿鏈段會有排列的情形，其自組裝行為偏好形成圓片狀，特別一提的是，如果硬桿鏈段具有共軛高分子的特性時，硬桿鏈段也會有規整排列的情形，不利於形成高分子囊胞。另外，我們也分析高分子囊胞的結構因子，當軟鏈段長度增長時，囊胞中含硬桿鏈段的膜厚會變薄，硬桿鏈段的單位密度也會下降，主要是因為囊胞內外的軟鏈段伸展方式不同所致。在囊胞的機械性質方面，我們分析了R5Cy (y = 1~3)型的囊胞，發現在R5C2型的高分子囊胞有較高的膜張力，與較低的拉伸模數與彎曲模數，推測對於膜融合實驗的結果有相當的影響，也就是造成R5C2型的高分子囊胞成功融合的原因，另外，R5C1型的高分子囊胞融合行為只有進行至半融合狀態，而R5C3型的高分子囊胞則是無法進入融合的步驟。 在論文的第二部份中，延續了上一個部分的主題，繼續討論硬桿-軟鏈共聚高分子囊胞的融合機制，影響膜融合的因素有膜張力、硬桿-軟鏈共聚高分子的個別長度、與溶劑的親疏性以及硬桿鏈段上的共軛高分子作用力。對於囊胞的膜融合，基本上有四個主要步驟，當兩顆高分子囊胞的初始接觸之後，在接觸的區域當中，第一條高分子的跨越即達第一個步驟(Kissing)。當兩顆囊胞的疏溶劑鏈段層互相連結接觸之後，也就是當stalk形成之後，則達到第二個步驟(Adhesion)。第三個步驟是指當兩個膜互相融合的區域變成一個膜厚的時候，稱之為半融合狀態(Hemifusion)，最後一個步驟就是，膜融合的區域出現了足以讓兩顆高分子囊胞內溶劑連通的小洞時，則完成融合步驟(Fusion)。我們觀察了在融合的過程中硬桿鏈段的微觀區域排列，與磷脂質囊胞的融合有所不同。我們利用了膨脹法/萎縮法增加或減少高分子囊胞內部的溶劑量，在膜滲透性不佳的情況下，囊胞內的壓力也會改變，進而控制膜張力，證實了囊胞膜在高張力的情況下有利於融合的進行，也發現存在一個臨界膜張力能夠促使原本停留在半融合狀態的兩顆囊胞完成融合。另外，除了膜張力外，由覆蓋在囊胞內外層的軟鏈段所造成的等方向性立體阻礙也會影響融合結果，也就是較長的軟鏈段能夠形成一個較高的能量障礙阻止融合的進行，如果我們增加了溶劑對硬桿鏈段與溶劑對軟鏈段的疏溶劑程度，則可以降低融合的障礙。 在論文的第三部份中，主要是在探討軟鏈-硬桿-軟鏈三段鏈共聚高分子(CRC)在選擇性溶液中的相行為，以對稱的三段鏈共聚高分子為主(CmRxCm)，變換硬桿與軟鏈段的長度，可得到五種主要的聚集形狀，分別是球型微胞、蟲型微胞、圓盤塊狀，蜂巢狀二維聚集以及高分子囊胞，然而受到硬桿與軟鏈的長度的限制，能夠形成高分子囊胞的條件十分嚴苛。為了要增加以三段鏈共聚高分子自組裝成高分子囊胞的機會，針對結構的因素，我們測試了兩種方法，一種是改變共聚高分子的對稱性，也就是形成不對稱的共聚高分子(CmRxCn，m < n)，另一種是在硬桿鏈的正中間嵌入T鏈段分枝(Cm(RxTy)Cm)，進而討論調整不對稱共聚高分子的軟鏈段長度與T鏈段長度對於相行為的影響，並分析這些囊胞的物理特性，最後觀察高分子囊胞的融合行為。 在論文的第四部份中，著重在一種線型樹枝狀共聚高分子(LDBC)的新穎材料，這種新穎的材料因嵌有偶氮苯(Azobenzene)短硬桿鏈段而具有光敏感性質，我們利用耗散粒子動力學法進行模擬，基由樹枝狀的世代數、濃度、各鏈段長度變化以及偶氮苯硬桿的π電子共軛強度進行熱力學相行為的討論，所產生的自組裝聚集形狀十分多樣，如球型微胞、蟲型微胞、線狀微胞、漢堡狀微胞、片狀，碗狀以及高分子囊胞，大致上來說，在世代數較大的情況下較有利於形成高分子囊胞，世代數較小時，只能形成奈米線狀與奈米片狀的微胞，此模擬結果與合成實驗的結果是一致的。在紫外光照射實驗中，因紫外光波長會改變偶氮苯短硬桿鏈段的化學性質與結構，而造成線型樹枝狀共聚高分子囊胞因此而產生表面皺褶、囊胞萎縮甚至破裂，在模擬實驗中我們除了驗證了此一光敏感現象外，更進一步模擬紫外光照對藥物釋放的行為影響。

Polymersomes have attracted great attention for their potential applications such as nano-reactor, drug release or cell imitator. The research objectives of this thesis are to explore the conditions for the formations of polymersomes self-assembled from rod-containing amphiphilic copolymers of various architectures. Structural, transport, and mechanical properties are studied to supplement the experimental findings. The fusion mechanisms between polymersomes are also explored. Due to the distinctive structural differences for various rod-containing block copolymers, different fusion behaviors are observed. In the first part (chapter 3), polymersomes formed by rod-coil diblock copolymer (RxCy) is fundamentally different from that of coil–coil diblock copolymers due to the effect that the rod block has limited ability to stretch and to accommodate packing within self-assembled structures. RxCy denotes the polymer comprises of x rod-like beads and y coil-like beads. The morphological phase diagram of RxCy in selective solvents and the essential physical properties of the RxCy-polymersomes are studied by using dissipative particle dynamics. Our simulation results show that small-sized polymersomes can only take shape for rod-coils with short enough rod-block length. The extended chain crystalline phases or high-order smectics of the rod domain disrupts the formation of polymersome. The detailed membrane structures of RC-polymersomes are also investigated and it is found that the rods within the membrane are highly interdigitated which is essentially different from the ordered bilayer of the liposomes. Moreover, the structural and mechanical properties of RxCy -polymersomes behave in an unexpected manner as the coil-block length (y) is adjusted. The membrane tension exhibit a maximum while the stretching and bending moduli display a minimum at y = 2 as y varies from 1 to 3. In addition, R5C2-polymersomes fuse most easily. Whereas R5C1-polymersomes do not proceed beyond the hemisfusion stage and R5C3-polymersomes can not even move past the initial kissing stage. In the second part (Chapter 4), the fusion mechanism of polymersomes self-assembled by rod-coil copolymers is investigated by dissipative particle dynamics. The influences of membrane tension, coil-block length, rod-block length, mutual compatibility between solvent and rod-coil block, and π-π interaction on the fusion pathway are explored. The fusion process of spontaneously formed polymersomes generally consists of four stages. In the kissing stage, hopping of rod-blocks forms connection between two vesicles of one-legged rod-coil copolymer. In the adhesion stage, a stalk is developed by a few link-up rods and then a stretched diaphragm with rods lying parallel to the stretching direction is formed in the hemi-fusion stage. Eventually, a pore is developed and expanded in the fusion stage. If the membrane tension (τ) is adjusted by deflation/inflation, the hemi-fusion diaphragm disappears. As τ is reduced, multiple stalks take shape and lead to the formation of inverted micelles, which is the rate-determining step and raises the fusion time substantially. As τ is elevated, the neck is developed after the stalk formation. The fusion time is significantly accelerated. τ of spontaneously formed vesicles varies with coil-block length, rod-block length, solvent quality, and π-π interaction. There exists a critical value of τ below which the fusion process cannot be completed and a hemi-fused polymersome is formed. In addition to τ, the anisotropic steric interactions within the rod layers also resist hopping of longer rod-blocks. The coil layers develop a barrier impeding fusion between vesicles with longer coil-blocks. Consequently, lowering the solvent quality for the coil-block or rod-block facilitates the fusion process because the coil layer becomes thinner. In the third part (Chapter 5), self-assembly behaviors of coil-rod-coil copolymers in a selective solvent are explored by dissipative particle dynamics. The morphological phase diagram as a function of rod length and coil length shows five distinct types of aggregates, including spherical micelle, worm-like micelle, disk-like aggregate, honeycomb structure, and polymersome. Small polymersomes are formed at rather poor alignment associated with the monolayered rod domain. Some of the rods are even lying perpendicular to the radial direction. For symmetric copolymers (CmRxCm), the condition of vesicle formation is restricted to short coil and rod lengths. To favor the formation of CRC-polymersome, two architecture modifications are adopted. One is to increase the coil length asymmetrically to be CmRxCn, where n>m. The other one is to tether a T-block onto the middle of the rod-block as Cm(RxTy)Cm copolymers. For those CRC-polymersomes, structural, transport, and mechanical properties of the vesicular membrane are determined, including membrane thickness, area density of coil blocks, order parameter, solvent permeability, frequency of flip-flop, membrane tension, and stretching and bending moduli. The influences of the coil length (n) and tethered block length (y) on membrane properties are examined. Finally, the mechanism of membrane fusion between CRC-polymersomes is investigated. The fusion process involves four stages and in the contact region the rods lying perpendicular to the direction of the rod layer play the key role. The encounter of two vesicles may result in fused, hemifused, or non-fused polymersome. The final fate is determined by the competition between membrane tension and steric barrier of coil corona. The fusion outcome may change if the tension is altered by manipulating the lumen pressure. In the last part (Chapter 6), azobenzene-containing linear-dendritic block-copolymers (LDBC) with varied generation numbers were synthesized recently. This photosensitive LDBC consists of a linear solvophilic block (R) and solvophilic dendrons of which the periphery is attached with a solvophobic coil-rod diblock (B-Y). The self-assembly and its photoresponsive transformation are explored by dissipative particle dynamics. Dependent on the generation number, polymer concentration, block lengths, and π-π