growth rates within the 3D features are in proportion to the concentrations of fi lmforming species (i.e., a source precursor or intermediate species generated from a precursor), under the assumption that the deposition reaction on the surface follows fi rst-order reaction kinetics, as is common in most low-pressure (LP) CVDs. In most LP-CVDs, the mean free path of the fi lm forming species is much larger than the feature size, so the governing mass-transport mechanism is Knudsen diffusion, and the gas-phase reactions consuming them are negligible. Accordingly, the growth rate profi le in the depth direction of a 3D feature can be written as a concentration profi le [ 14 ] φwhere C is the concentration of the fi lm-forming species at position x , C t is the concentration at the top (e.g., the opening) of 3D features, x is a position from the bottom, and φ is the Thiele modulus. L and S / V are the depth and the surface-tovolume ratio of the feature, respectively. k s and D are the surface reaction rate constant and diffusion coeffi cient of the fi lm-forming species, respectively. As L , S , and V are determined by the geometry of the 3D features, k s and D are the only parameters that control the thickness profi le. Thus, if we know the relationship between k s / D and the process conditions, it is possible to predict the conformality within the target 3D structures. It is also helpful to know the maximum ARs feasible with the process used. [ 15 ] To determine k s / D , we have to identify the chemical compositions of the fi lm-forming species and evaluate their D and k s in the target 3D feature under the operative condition. To identify the fi lm-forming species, ab initio computational calculation methods of gas-phase chemistry have been developed, enabling accurate prediction of the time evolutions of gas-phase species. [ 16 ] D can be also determined accurately by an empirical method, considering the operative conditions (e.g., temperature and pressure) and geometrical information about the 3D feature (e.g., mean pore size and the structural complexity, such as the tortuosity factor). [ 17,18 ] However, k s is diffi cult to determine theoretically due to many uncertainties in the ab initio computational calculation of surface chemistry (e.g., a signifi cant error in estimation of vibrational frequencies [ 19 ] ). Thus, k s is often estimated by order-of-magnitude. [20][21][22][23] A method is proposed to fabricate a high-aspect-ratio (HAR) microchannel with a microscopic gap and AR of more than 1000:1 applicable to a test structure for kinetic analysis of chemical vapor deposition (CVD). It has a parallelplate structure and is formed concisely by sticking a planar Si substrate and a patterned Si or silicon-on-insulator (SOI) substrate fabricated by single-step etching, by clamping them. The resulting feature exhibits a uniform gap and smooth surface morphology along its depth. When CVD is conducted into this HAR microchannel, the sticking probability ( η ) of fi lm-forming species can be detected by analyzi...