NOTES ON INSTRUCTIONS: SC : Nuclear spin for the central atom. eg., if I=1 the instruction will be SC 1.0, for I=5/2 give SC 2.5 GX,GY,GZ : If all are equal the system has isotropic g-tensor. For an axial system GX and GY will be same, but different from GZ. Rhombic case will have all the three values different. AX,AY,AZ : Hyperfine splittings for the central atom in units of 10^-4 cm-1. An A-value of 50.0 G corresponds to g x 50.0 x 0.466923 x 10^-4 cm-1. If g = 2.000, A will be 46.7 x 10^-4 cm-1. Therefore, the input value will be 46.7 The program assumes same principal directions for g and A. QP,QDP : These are Q' and Q'', the quadrupole coupling para- meters. These are usually much smaller than A-values and can be assumed to be zero(default) for most purposes. GN : Nuclear Lande factor(gn) for the central atom. This appears in the nuclear Zeeman term which can be often neglected. Since the program does not distinguish between isotopes a population weighted value for gn can be given if desired. FR : The instrument frequency in cm-1 units. eg. 9.350 GHz corresponds to 9.350/29.979 = 0.3119 cm-1. TS,PS : Step sizes in deg. for theta and phi-integration. The default values are 90(TS) and 180(PS) and these suffice for the isotropic case(eg. solution spectrum). For axial case TS only need be given, PS being set by default. For rhombic case both should be given. The values are to be chosen based on the anisotropy in g and A. Large anisotropy will require small step size(say 1 deg) while small anisotropy can be simulated with higher values(say 10 deg) - see the two sample inputs. TS and PS values should be chosen carefully. Small values will mean greater computation time because more orientations have to be calculated when the step sizes are small. On the other hand, if too large a value is given for a highly anisotropic case, the simulated spectrum will have considerable "graining" - this is approximately equivalent to recording an experimental spectrum with a coarse powder. SF,EF : Starting and ending fields for the simulation (in Gauss). WI,WX,WY,WZ : Line widths. For simulation assuming isotropic widths only WI need be given. If anisotropic widths are desired give WX,WY and WZ instructions. Default is 5 Gauss for all. AF,GM : These take care of g-strain and A-strain (eg. in glassy or biological samples) in a very approximate way. A-strain can lead to M-I dependent line widths (ie., each hyperfine line may have a different width). Typical values for AF and GM are 0.001 and 3.0(Gauss) respectively. SL : Ligand nuclear spin. At present only one type of ligand is allowed. Each ligand requires an LP instruction followed by 5 lines of data (one number per line). The ligand h.f tensor is assumed to be axial whose principal axis is defined by giving the direction cosines with respect to the g/A tensor axes. Often the ligand tensor axis can be assumed to be along the bond. Two examples are given below: (1) A CuN4 complex having D4h symmetry - all N's (I=1) are equivalent and the g/A x,y axes are along the bonds and z axis normal to the plane. Typical A and B values are 13 and 10 (x10^-4) cm-1. The instructions will be: SL 1.0 LP 1.0 0. N3 0. . 13. . 10. . LP N2. . . . Cu . . . .N4 ----- y 0. . -1. . 0. . 13. N1 10. | LP | -1. | 0. x 0. 13. 10. LP 0. 1. 0. 13. 10. (2) This case has two F ligands(I=1/2). The molecule is planar and the x,y molecular( g/A ) axes bisect the bond angle. With A=80 and B=60 (x 10^-4 cm-1), the instructions will be SL 0.5 LP 0.7660 O O M--F1 -0.6428 . . lx = cos THETA 0. . . ly = -sin THETA 80. . . lz = 0 60. M ------- y LP . | . M--F2 0.7660 . | . lx = cos THETA 0.6428 . | . ly = sin THETA 0. F1 | F2 lz = 0 80. x 60. (angle F1 M x = angle x M F2 = THETA = 40 deg.) TE,PE : End values for theta and phi for part calculation. These need be given only if it is desired to accumulate the spectrum from several runs, because a continuous run takes too long, say several hours. In this case, for the second run onwards the CO instruction should be given in place of GO. The CO instruction signals continuation and should be given also for continuing an interrupted run. The same file name should be given for storing the plot data for the various incomplete runs. LS: Either a Lorentzian (LS 1.) or a Gaussian (LS 2.) can be chosen. Gaussian is faster because it cuts off more rapidly than the Lorentzian which has a long 'tail' - note that all numbers following the instructions are in floating point format.