Reflection nebulae–dense cores–illuminated by surrounding stars offer a unique opportunity to directly test our quantitative model of grain alignment based on radiative torques ( RATs ) and to explore new effects arising from additional torques .
In this paper , we first perform detailed modeling of grain alignment by RATs for the IC 63 reflection nebula illuminated both by a nearby \gamma Cas star and the diffuse interstellar radiation field .
We calculate linear polarization p _ { \lambda } of background stars by radiatively aligned grains and explore the variation of fractional polarization ( p _ { \lambda } / A _ { V } ) with visual extinction A _ { V } across the cloud .
Our results show that the variation of p _ { V } / A _ { V } versus A _ { V } from the dayside of IC 63 to its center can be represented by a power-law ( p _ { V } / A _ { V } \propto A _ { V } ^ { \eta } ) with different slopes depending on A _ { V } .
We find a shallow slope \eta \sim - 0.1 for A _ { V } < 3 and a very steep slope \eta \sim - 2 for A _ { V } > 4 .
We then consider the effects of additional torques due to H _ { 2 } formation and model grain alignment by joint action of RATs and H _ { 2 } torques .
We find that p _ { V } / A _ { V } tends to increase with an increasing magnitude of H _ { 2 } torques .
In particular , the theoretical predictions obtained for p _ { V } / A _ { V } and peak wavelength \lambda _ { \max } in this case show an improved agreement with the observational data .
Our results reinforce the predictive power of the RAT alignment mechanism in a broad range of environmental conditions and show the effect of pinwheel torques in environments with efficient H _ { 2 } formation .
Physical parameters involved in H _ { 2 } formation may be constrained using detailed modeling of grain alignment combined with observational data .
In addition , we discuss implications of our modeling for interpreting latest observational data by Planck and other ground-based instruments .