Magnetochimica AA 2011-2012 Marco Ruzzi Marina Brustolon 4. The Zero Field Splitting spin Hamiltonian This lecture • In the previous lecture we have seen an example of Spin Hamiltonian, i.e. the so called Exchange Hamiltonian : • H 2JS S 1 2 • Now we will analyze a second Spin Hamiltonian, due to the dipolar interaction between two electrons magnetic moments. •This interaction is called Zero Field Splitting (ZFS) interaction, or also “Fine interaction” to distinguish it from “Hyperfine interaction”. We will show that the form of this latter spin Hamiltonian is : H SDS H ZFS ~ S DS Zero field splitting spin hamiltonian. ~ S is the total spin and D is a symmetric tensor (fine tensor). We will see : a. the physical origin of this Hamiltonian b. how to obtain it from the classical dipolar interaction expression. Physical origin of ZFS H ZFS ~ S DS This spin hamiltonian corresponds to two different interactions: 1. electron-electron dipolar interaction 2. mixing between ground and excited electronic states by spinorbit coupling The two interactions are different and independent from each other, but they give rise to the same type of spin Hamiltonian. ~ D Therefore the elements of tensor are in general due to both interactions. However, the spin-orbit coupling is usually predominant for transition metal ions, and the dipolar interaction for organic triplets. ZFS as originated from electron-electron dipolar interaction Let us suppose to have two unpaired electrons in two different molecular orbitals (either on the same molecule, or on different molecules). E 3M LUMO HOMO Biradical Triplet N O Between near electrons there will be a magnetic interaction, which is called “dipolar interaction”, or “fine interaction”. z r μ2 μ1 The Hamiltonian of the dipolar interaction can be obtained from the classical dipole-dipole interaction energy: Edip μ 0 μ1 μ 2 (μ1 r )(μ 2 r ) 3 3 4 r r5 where r is the distance between the two electrons. How to pass from the classical energy expression to the ZFS spin Hamiltonian In the following we will show how one can obtain the spin hamiltonian H ZFS ~ S DS from the classical dipolar interaction expression: Edip μ 0 μ1 μ 2 (μ1 r )(μ 2 r ) 3 3 4 r r5 2 r 1 From the classical energy to the spin hamiltonian: By substituting H ZFS μˆ i g i μ B si ˆ1 sˆ2 μ0 (sˆ1 rˆ)(sˆ2 rˆ) 2 s g1 g 2 μ B 3 3 4 r5 r Developing the products this expression becomes: Cost μ 0 g1 g2μ 2B 2 H ZFS ( r s1x s2 x s1 y s2 y s1z s2 z 5 4 r 3( xs1x ys1 y zs1z )( xs2 x ys2 y zs2 z )) and rearranging: Cost 2 2 2 2 2 2 (( r 3 x ) s s ( r 3 y ) s s ( r 3 z ) s1z s 2 z 1x 2 x 1y 2 y 5 r 3xy( s1x s 2 y s1 y s 2 x ) 3xz( s1x s 2 z s1z s 2 x ) 3 yz( s1 y s 2 z s1z s 2 y )) H ZFS It is convenient to express these products in terms of the total spin S = s1 + s2 and its components. S x (s x1 s x 2 ) S x2 (s x1 s x 2 ) 2 s x21 s x22 2s x1s x 2 and analogously for the y and z components. We can write: 1 1 S ( s x1 s x 2 ) 2s x1 s x 2 4 4 2 x In fact: s s x is y 2 s s x is y s s therefore s x 2 s s 1 sx 2 2 s s 1 sx 2 2 s s sy 2i s s 1 4 4 s s 1 2 sx 4 4 s 2 x The effect of the operator s2x is always to give ¼; therefore we can substitute ¼ to the operator. From: S x2 ( s x1 s x 2 ) 2 1 1 2s x1 s x 2 4 4 we obtain: S y2 ( s y1 s y 2 ) 2 1 1 2 s y1 s y 2 4 4 s x1 s x 2 S z2 ( s z1 s z 2 ) 2 1 1 2 s z1 s z 2 4 4 By a similar procedure: From: 1 2 1 Sx 2 4 etc. we obtain: S x S y S y S x 2(s x1s y 2 s y1s x 2 ) s x1s y 2 s y1s x 2 1 2 (S x S y S y S x ) S x S z S z S x 2(s x1 s z 2 s z1 s x 2 ) s x1s z 2 s z1s x 2 1 2 (S x S z S z S x ) S y S z S z S y 2(s y1s z 2 s z1s y 2 ) s y1s z 2 s z1s y 2 1 2 (S y S z S z S y ) Cost 2 H ZFS 5 ((r 3x 2 ) s1x s 2 x (r 2 3 y 2 ) s1 y s 2 y (r 2 3z 2 ) s1z s 2 z r 3xy( s1x s 2 y s1 y s 2 x ) 3xz( s1x s 2 z s1z s 2 x ) 3 yz( s1 y s 2 z s1z s 2 y )) By replacing the single spin operators with the total spin operator we obtain: μ 0 g 2 μ 2B 1 2 2 2 2 2 2 2 2 2 H ( r 3 x ) S ( r 3 y ) S ( r 3 z ) S x y z 4 r 5 2 3xy( S x S y S y S x ) 3xz( S x S z S z S x ) 3 yz( S y S z S z S y ) This operator contains both spatial operators and spin operators. The spatial operators are given by products of the vectorial components of the vector joining the positions of the two electrons. To obtain a spin Hamiltonian, we must find the average values of the spatial operators over the spatial wavefunction representing the two electrons: (1,2) r 2 3 x 2 (1,2) r 2 3 x 2 etc. Finally, the products of spin operators can be put in a convenient tensorial form: H ZFS r 2 3x 2 5 r μ 0 g 2μ 2B [ S S S ] x y z 8 r 5 H ZFS ~ S DS 3xy r5 r2 3y2 r5 3xz r5 3 yz r5 r 2 3z 2 r5 Sx S y Sz H ZFS ~ S DS ~ D è simmetrico e senza traccia X Nel sistema di assi principali: D 2 2 2 H XS YS ZS ZFS x y 0 Y 0 0 Z z E’ conveniente riscrivere l’Hamiltoniano per lo stato di tripletto usando due parametri ottenuti dai precedenti: YX E 2 3 D Z 2 Con questi parametri l’Hamiltoniano si può riscrivere come: 1 H ZFS D( S z2 S 2 ) E( S x2 S y2 ) 3 e più convenientemente: 1 1 H ZFS D( S S ( S 1)) E( S2 S2 ) 3 2 2 z Magnetic energy levels and EPR spectra 1 1 H ZFS D( S z2 S ( S 1)) E( S2 S2 ) 3 2 The eigenfunctions and eigenvalues of this hamiltonian are: Eigenfunctions X = 1/2{ -} Y = 1/2{ +} Z = 1/2{ + } Eigenvalues 1 EX D E 3 1 EY D E 3 2 EZ - D 3 E X Y Z Now we consider also the presence of an external magnetic field. Let us assume an isotropic g factor. 0 ZFS and Zeeman interactions 1 1 2 2 H gμ B B S D( S S ( S 1)) E( S S ) 3 2 2 z The eigenfunctions of this term are the triplet functions S M s 11 1 2 ( ) 1 1 10 H SM S g B B0 S z SM S M S g B B0 SM S The eigenfunctions of this term are the linear combinations of the triplet functions X Y 11 1 1 2 11 1 1 2 Z 10 When both terms of the spin Hamiltonian are present, the eigenfunctions are a mixing (on one of the two basis). Depending on the relative magnitudes of the two interactions, is more convenient to use a basis or the other one. 1 1 H gμ B B S D( S S ( S 1)) E( S 2 S 2 ) 3 2 2 z On the basis of the ZFS eigenfunctions: X X Y Z Y Z 1 DE 3 gμ B BlZ igμ B BlY gμ B BlZ 1 DE 3 gμ B BlX gμ B BlX 2 D 3 igμ B BlY The elements out of diagonal depends on the direction cosines of the magnetic field in the XYZ axes of the fine tensor as lx, ly, lz. We can find easily the roots to diagonalize this secular determinant for B//X,Y,Z (this could correspond to a triplet in a single crystal oriented in the magnetic field). For example, for B//Z we have: lx =0 ly = 0 lz=1 therefore: X Y Z X 1 DE 3 gμ B B 0 Y gμ B B 1 DE 3 0 Z 0 0 2 EZ D 3 2 D 3 1 2 2 1/ 2 E D E ( gμ B B ) 3 The EPR allowed transitions are those with Ms= ±1. B // X X Y Z B// Y B B B// Z B B B B B X B0 Single crystal analysis B0 // X Y B 3 X gμ B 1000 Z 2000 Y 3000 4000 B0 (gauss) 5000 B Z B0 // Y X B 3Y gμ B 1000 2000 Z 3000 4000 B0 (gauss) B X B0 // Z Y 5000 B 3Z gμ B 1000 2000 3000 4000 B0 (gauss) 5000 1 G = 10-4 T EPR of triplets in a disordered matrix Disordered matrices B 1000 2000 3000 4000 B0 (gauss) 5000 X B0 Y Z Y Z X Z X B0 Y The principal values of the ZFS tensor can be obtained also from the features of the spectra in disordered phases. Disordered matrices 3XZFS 3YZFS 3ZZFS 2500 3000 3500 4000 4500 Field (G) assorbimento Disordered matrices Powder EPR spectrum of Naphtalene triplet B0 (E. Wasserman et al. J. Chem. Phys. 41, 1763 (1964)) dI/dB0 B0 B0 Some examples of ZFS parameters D / cm-1 |E|/cm-1 Benzene 0.158 0.0065 Naphtalene 0.104 Anthracene ±0.072 0.007 Biphenyl ±0.109 0.004 C60 0.0115 0.00047 C70 0.0052 0.00069 Fullero pyrrolidine 0.0090 0.0014 Triplets • 1. Ground state triplets: e.g. non kekulè molecules. • 2. Thermally accessible triplets (J~kT) • Photoexcited triplets Photoexcited triplets: a simple approach for ZFS calculations • For a photoexcited triplet * we can assume the simple model of an electron in HOMO and another one in LUMO. • Therefore we can calculate the ZFS parameters as deriving from the dipolar interaction between the spin distributions in LUMO and HOMO. Example: triplet of Butadiene LUMO HOMO μ1 μ2 We can separate the total dipolar interaction in the interactions between pairs of spin densities on atoms. Osservazioni: 1. L’interazione tra le densità di spin sullo stesso atomo di carbonio non c’è, perché per il principio di Pauli due elettroni con lo stesso spin non possono stare sullo stesso orbitale atomico. 2. L’interazione dipolare sarà tanto più grande quanto più piccolo è il volume nel quale sono distribuiti i due spin. 3. I tripletti del tipo n * hanno in generale un’interazione dipolare più grande rispetto ai tripletti * perché contribuiscono le distribuzioni di spin sullo stesso atomo, quindi molto vicine (il principio di Pauli in questo caso non si oppone!). Time resolved EPR (TR-EPR) spectra of photoexcited triplets Photoexcited triplets are born with non Boltzmann populations of the spin states. Therefore if the EPR spectrum is obtained shortly after the laser pulse (1 s) the EPR lines are part in enhanced absorption and part in emission. The parameters obtained are D and E , and the spin populations. The latter ones tell how the triplet was born (for example by ISC from the excited singlet). TR-EPR spectra of triplets 3XZFS 3YZFS 3ZZFS A 2500 3000 3500 4000 4500 EPR spectrum of a non polarized triplet Field (G) A E 2500 3000 A absorption 3500 4000 4500 Field (G) E Emission TR-EPR spectrum of a photoexcited polarized triplet ZFS as originated from spin-orbit coupling The Hamiltonian to be considered as a perturbation in addressing the ZFS in this case is the same as for g tensors: H geμ B S B μ B L B L S The second and third terms represent the Zeeman interaction of the orbital angular momentum and the spin-orbit coupling. The terms in L couple electronically excited states into the ground state and contribute in second order. There are four contribution to the second order energy. One of these is the term in 2, second order in S and independent of B. This gives rise to a ZFS ~ spin hamiltonian as in the previous slide: H S D S ZFS Quartet states Quintet states