A microfluidic system for controlling reaction networks in time
A microfluidic system for controlling reaction networks in time
复制标题
DOI：
10.1002/anie.200390203
复制
发表时间：
2003-01-01
影响因子：
16.6
通讯作者：
Ismagilov, RF
中科院分区：
文献类型：
--
作者：
Song, H;Tice, JD;Ismagilov, RF
我们在这里提出了一个微流控系统，可用于控制网络的许多化学反应的毫秒级。它允许控制每个反应何时开始，每个反应在分离或与其他反应结合之前进行多长时间，以及何时分析或淬灭每个反应。该系统使用流体流动将空间（毛细管长度）线性转换为时间（反应时间）。对于开放系统中的化学反应，这种转换是简单且公知的：试剂A的溶液和试剂B的溶液在它们之间的反应开始的初始点d= 0（t= 0）作为稳定流注入微流体通道中。当反应混合物由流体流以恒定速度U输送时，每个空间点d对应于时间点t，即反应时间，其中t= d/U。如果实施这样的系统，则可以简单地通过创建承载反应混合物的会聚和发散通道的网络以及改变流速以调节反应和相互作用时间来控制多个化学反应的相互作用。如果反应伴随着光学信号（例如荧光或吸收的变化），则可以从单个空间分辨的光学图像获得整个网络中反应的时间分辨测量。微流体通道的网络[1-3]对于这种距离到时间的转换特别有吸引力，因为它们可以很容易地制造并用于操纵小体积的试剂;它们对于化学和生物学分析和合成变得至关重要。[1-3]微流体装置中的流动是层流;它发生在雷诺数Re（0.01-100）的低值下。Re定义为IU 1/m，其中l [m]是毛细管的直径，U [ms <$1]是流速，1 [kg m <$3]是密度，m [kg m <$1 s <$1]是流体的粘度。[4]由于两个原因，这种层流使得难以在微流体装置中实现距离到时间的转换（图1a）。首先，混合是缓慢的-两股流并排注入通道流，仅通过扩散混合，[4-6]因此d= 0不对应于反应的明确定义的起始点（t= 0）。大量的研究工作致力于解决这个缓慢混合的问题。[2，3，7]湍流提供快速混合和低分散，[8]但湍流发生在Re> 2000的值时，仅在高流速（10 ms/h）下在微通道中达到。实现这样的流速需要不期望的高样品消耗（101 mLsg-1）和高压。其次，溶质沿着通道的分散是大的-流动剖面是抛物线，并且试剂以一定范围的速度输送。[4]因此，给定的距离d对应于反应时间t= d/U的范围。我们开发了一个简单的微流体系统，克服了这两个问题-它传输的解决方案与快速混合，没有分散（图1b）。该系统使用矩形横截面的微通道网络和疏水表面，使用快速成型在聚二甲基硅氧烷（PDMS）。[9]我们使用注射泵控制通过每个通道的体积流速。通过将试剂定位在由与水不混溶的油分离的水性塞（大到足以阻塞通道的液滴）内来消除分散。在商业[10]和实验室[11-13]系统中，不混溶流体已被用于定位试剂。在这里，我们描述的方法，用于形成多个解决方案的试剂塞，使用混沌平流，以实现特别快速（102毫秒）内的插头混合，并分裂和合并…
We present here a microfluidic system that may be used to control networks of many chemical reactions on the millisecond scale. It allows to control when each reaction begins, for how long each reaction evolves before it is separated or combined with other reactions, and when each reaction is analyzed or quenched. The system uses flow of fluids to linearly transform space (length of capillaries) into time (reaction time). For a chemical reaction in an open system this transformation is simple and well known: A solution of reagent A and a solution of reagent B are injected as steady streams into a microfluidic channel at initial point d= 0 where the reaction between them begins (t= 0). As the reaction mixture is transported by the fluid stream at a constant velocity U, every spatial point d corresponds to a time point t, the reaction time, where t= d/U. If such a system is implemented, interactions of multiple chemical reactions in time could be controlled simply by creating a network of converging and diverging channels carrying reaction mixtures, and varying flow velocities to adjust reaction and interaction times. If the reactions are accompanied by an optical signal (eg changes in fluorescence or absorption), time-resolved measurements of the reactions in the entire network could be obtained from a single spatially resolved optical image. Networks of microfluidic channels [1–3] are especially attractive for this distance-to-time transformation because they can be easily fabricated and used to manipulate small volumes of reagents; they are becoming essential for chemical and biological analysis and synthesis.[1–3] Flow in microfluidic devices is laminar; it occurs at low values of the Reynolds number, Re (∼ 0.01–100). Re is defined as lU1/m, where l [m] is the diameter of the capillary, U [ms À1] the velocity of the flow, 1 [kg mÀ3] the density, and m [kg mÀ1 sÀ1] the viscosity of the fluid.[4] This laminar flow makes it difficult to implement the distance-to-time transformation in microfluidic devices for two reasons (Figure 1a). First, mixing is slow—two streams injected into a channel flow side-by-side with mixing only by diffusion,[4–6] therefore d= 0 does not correspond to a welldefined starting point (t= 0) of the reaction. Significant research efforts have been devoted to solving this problem of slow mixing.[2, 3, 7] Turbulent flows provide both rapid mixing and low dispersion,[8] but turbulence occurs at values of Re> 2000 reached in microchannels only at high flow rates (∼ 10 msÀ1). Achieving such flow velocities requires undesirably high sample consumption (∼ 1 mLsÀ1) and high pressures. Second, the dispersion of solutes along the channel is large—the flow profile is parabolic, and the reagents are transported at a range of velocities.[4] Therefore, a given distance d corresponds to a range of reaction times t= d/U. We developed a simple microfluidic system that overcomes both problems—it transports solutions with rapid mixing and no dispersion (Figure1b). This system uses networks of microchannels with rectangular cross sections and hydrophobic surfaces fabricated using rapid prototyping in polydimethylsiloxane (PDMS).[9] We controlled the volumetric flow rates through each channel using syringe pumps. Dispersion was eliminated by localizing the reagents within aqueous plugs (droplets large enough to block the channel) separated by a water-immiscible oil. Immiscible fluids have been used to localize reagents in both commercial [10] and laboratory [11–13] systems. Here we describe methods for forming plugs of multiple solutions of reagents, for using chaotic advection to achieve especially rapid (∼ 2 ms) mixing within the plugs, and for splitting and merging …