Recently oblique angle deposition has been used in a wide range of important, unique applications. The column angle of the obliquely deposited columnar structures is an important parameter that determines their mechanical, optical, and chemical properties. Unfortunately this angle can be greatly affected by materials and processing conditions which are too complex to model and predict. Existing models such as the tangent rule and cosine rule are independent of materials and processing conditions and therefore in general have a limited ability to predict the column angle. We present a semi empirical model that includes the effects of materials and processing conditions. We show that our model is able to accurately predict column angle analytically for a wide range of obliquely deposited amorphous Ge for two different sets of processing conditions. We also show how this model can be used to predict other useful quantities, such as porosity and column width.

 The model uses the fact that the deposition on a line (or wire) results in a fan structure due to a self-shadowing effect with a fan angle that depends on materials and processing conditions. We first show how columnar structures can be generated by analytically applying global shadowing between the fan structures growing on adjacent lines. The columnar structures obtained possess geometrical properties (such as column width and column merging) that are consistent with columnar structures observed in experiment and simulation. We show how the exact shape and time evolution of the columnar structure can be calculated based on the knowledge of the fan shape. Once the exact shape of the columnar structure is known, various useful quantities can also be obtained: column angle, porosity, and column width. We experimentally verified our model by depositing amorphous Ge on line seeded substrate and on a flat substrate. The model agrees with experiments done on both substrates.

 Finally, we describe relatively simple experimental setups that can be used for fast and convenient measurement of the fan geometry at various processing conditions, such as flux rates and temperatures. These fan geometry data obtained on normal incident flux can then be used to predict the columnar structure geometries for the whole range of incident flux angles and for all the various processing conditions.