Ligand-coated semiconductor nanoplatelets (NPLs), also called nanoribbons, can adopt a wide variety of shapes in solution or when deposited from a solution onto a substrate. Following [Guillemeney et al (2022)](http://doi.org /10.1038/s42004-021-00621-z), the most common shapes for CdSe-type NPLs are cylinders, helices, and helicoids. More exotic shapes have been observed with other crystalline materials, e.g. double spiral-like shapes have been reported by Guillemeney et al (2024) for InS. While these shapes can be generated from molecular dynamics (MD) simulations, as has been done by [Monego et al (2024)](http://doi.org/ 10.1073/pnas.2316299121), it is computationally very expensive for experimentally relevant NPL sizes. In addition, MD-generated shapes invariably contain fluctuations. While modeling shape-dependent NPL properties, e.g. scattering patterns, it is often very useful to rapidly generate a desired shape with certain geometric properties, e.g. pitch and radius in case of helical ribbons. This package allows us to directly generate, without any simulation, a wide variety of idealized NPL shapes. The shapes are in the form of a mesh and/or constituent atoms of the NPL.
-
Shapes with constant curvatures
- Cylinders
- Helicoids (right and left handed)
- Ribbons with a helical centerline (right and left handed)
- Degenerate helices
-
Shapes with non-constant centerline curvatures. The curvature and twist along the centerline can be specified as a function of the distance along the centerline. In addition to new shapes, this feature allows creation of lightly perturbed shapes around a base shape with constant curvature, which may be useful for studying experimentally observed shape-polydispersity.
-
Spiral shaped ribbons can be created and joined end-to-end to form double spirals. Many different spirals are supported:
- Generalized Archimedean spirals (includes Archimedes spiral, Fermat spiral, hyperbolic spiral, and lituus),
- Circle involute
- Polynomial spirals (includes Nielsen and Cornu). The curvature of Nielsen spirals vary exponentially (increasing or decreasing) with arclength, which could of interest in studying the physics of certain NPL models.
Parts of this code has been used in the following publication(s):
- L. Guillemeney, S. Dutta, R. Valleix, G. Patriarche, B. Mahler, and B. Abécassis, Ligand tail controls the conformation of indium sulfide ultrathin nanoribbons Journal of the American Chemical Society 146, 22318 (2024).
This package is based on the formalism presented in the following papers.
- D. Grossman, E. Sharon, and H. Diamant, Elasticity and fluctuations of frustrated nanoribbons Physical Review Letters 116, 258105 (2016).
- D. Grossman, E. Sharon, and E. Katzav, Shape and fluctuations of positively curved ribbons Physical Review E 98, 022502 (2018).
- M. Zhang, D. Grossman, D. Danino, and E. Sharon, Shape and fluctuations of frustrated self-assembled nano ribbons Nature Communications 10, 3565 (2019).
Dependencies. numpy, scipy, rotlib. It is not necessary to separately install them, pip will install whatever is required (taking care of the appropriate versions).
In a new virtual environment, install using the command
pip install "ribbon @ git+https://github.com/saridut/ribbon.git"Documentation is available on GitHub pages of this repo.
import numpy as np
from ribbon import Ribbon, write_xyz
import matplotlib.pyplot as plt
length = 36.0 # Length of the ribbon
width = 5.0 # Width of the ribbon
thickness = 0.0 # Set ribbon thickness to zero for a single atomic layer
#Set some atoms in the reference configuration. For more realistic cases
#this could be chosen from an appropriate lattice.
atom_coords = []
for x in np.arange(0, length, 0.4):
for y in np.arange(-width/2, width/2, 0.4):
atom_coords.append([x,y,0])
atom_coords = np.asarray(atom_coords)
#Mesh size along length
dl = 0.1
#Mesh size along width
dw = 0.1
#Mesh size along thickness
dt = 0.0 # No thickness
radius = 2.0 #Radius of the helical ribbon
pitch = 10 #Pitch of the helical ribbon
#Create the ribbon.
ribbon = Ribbon(length, width, thickness, dl, dw, dt)
ribbon.set_atom_refpos(atom_coords)
ribbon.set_shape('HelicalRibbon', radius=radius, pitch=pitch)
ribbon.create(orient_along=[0,0,1]) #Orient helix axis along z.
#The (n,3) ndarray ribbon.atom_pos contains the atom positions of
#the helical ribbon.
#Write out atom positions for visualization in Ovito/VMD/etc.
write_xyz(ribbon.atom_pos[:,0], ribbon.atom_pos[:,1], ribbon.atom_pos[:,2],
atom_symbol='C', filename='out.xyz', title='')
#Geometrical properties: Radius, pitch, etc. Below, theta is the angle between
#the principal curvature direction and the length direction of the ribbon.
print(f"R = {ribbon.get_radius():g}\n"
f"P = {ribbon.get_pitch():g}\n"
f"kg = {ribbon.get_gauss_curvature():g}\n"
f"km = {ribbon.get_mean_curvature():g}\n"
f"theta = {ribbon.get_theta():g}")
#Plotting the midline and the midsurface as a wireframe
figh, axh = plt.subplots(nrows=1, ncols=1, figsize=(12,9),
subplot_kw={'projection':'3d', 'proj_type': 'persp'}
)
axh.plot(ribbon.mline[:,0], ribbon.mline[:,1], ribbon.mline[:,2], '-k', lw=1.2)
axh.plot(ribbon.mline[0,0], ribbon.mline[0,1], ribbon.mline[0,2], 'ob')
axh.plot(ribbon.msurf[0,:,0], ribbon.msurf[0,:,1], ribbon.msurf[0,:,2], '-g', lw=1.0)
axh.plot(ribbon.msurf[-1,:,0], ribbon.msurf[-1,:,1], ribbon.msurf[-1,:,2], '-g', lw=1.0)
axh.plot(ribbon.msurf[:,0,0], ribbon.msurf[:,0,1], ribbon.msurf[:,0,2], '-r', lw=1.0)
axh.plot(ribbon.msurf[:,-1,0], ribbon.msurf[:,-1,1], ribbon.msurf[:,-1,2], '-r', lw=1.0)
axh.plot_wireframe(ribbon.msurf[:,:,0], ribbon.msurf[:,:,1],
ribbon.msurf[:,:,2], linestyles='-', linewidths=0.5,
rstride=2, cstride=8, color='0.5')
axh.set_aspect('equal', 'box')
axh.set_axis_off()
plt.show()Executing the above code gives
R = 2
P = 10
kg = 6.93889e-18
km = 0.25
theta = 0.672159In the image below, the left panel shows the wireframe of the midsurface and the atom positions visualized using Ovito are shown in the middle panel. On the wireframe, the dark line marks the centerline and the blue dot indicates the origin.
The atoms can also be chosen based on the crystal structure. In this case, the positions need to be generated using a crystallographic tool or manually. The right panel of the above image shows a 3.5 monolayer CdSe nanoribbon with the same code as before except that the Cd and Se atom positions in the reference state were obtained from ASE. The visualization was performed using ChimeraX.
import numpy as np
from ribbon import SpiralArchimedes, double_spiral, extrude, write_xyz
#First create a single Archimedean spiral with parameters b = 0.2 and
#r0 = 0.
length = 20.0 # Length of the spiral
width = 5.0 # Width of the ribbon
b = 0.2 # Spiral parameter
r0 = 0.0 # Spiral parameter
ds = 0.1 # Spiral arclength spacing
#Discretized arclength
s = np.linspace(0, L, 1+int(L/ds))
spl = SpiralArchimedes(b, r0) #Single spiral
#Create double spiral by joining end-to-end. This will make the length
#of the resulting object 2*length.
dspl = double_spiral(spl, L, ds=ds, same=False, end=1, f=0.0)
#Double spiral coordinates. dspl[0] is arclength, dspl[1] is curvature.
dspl_x = dspl[2] #x coordinates
dspl_y = dspl[3] #y coordinates
#Extrude. Atoms will be placed at x and y and extruded along z.
dz = 0.1 #Atom spacing along z
X, Y, Z = extrude(dspl_x, dspl_y, width, dz)
#Write atom positions
write_xyz(X, Y, Z, atom_symbol='C', filename='dspl.xyz', title='')Below is the double spiral ribbon visualized in Ovito.
