ORCA labs 10 weak interactions...

14
Intermolecular and Weak Interactions 1 1 Computer Experiment 10: Intermolecular and Weak Interactions 1.1 Background In this section we treat intermolecular and weak interactions, this means noncovalent and non ionic interactions between different atoms or molecules. We will focus our attention to the two most familiar representatives of such interactions: Hydrogen bonds and van der Waals or dispersion interactions. Other weak intermolecular interactions like ππ stacking or dipoledipole interactions are closely related to these two basis types of interaction. 1.1.1 Hydrogen bonds Hydrogen bonds belong to the most important noncovalent intermolecular interactions. They are operative in determining molecular conformation, molecular aggregation, and the function of a vast number of chemical systems ranging from inorganic to biological. In the majority of cases hydrogen bonding occurs between a polar bond X δ H δ + and an electron rich region B δ . Hydrogen bonding can be understood as a donoracceptor interaction. Therefore we will denote XH as the (hydrogen) donor and B as the (hydrogen) acceptor. Definition: The first definition of the hydrogen bond was given by Pauling [1]: “Under certain conditions a hydrogen atom is attracted by rather strong forces to two atoms, instead of only one, so that it may be considered to be acting as a bond between them.” He further added that “it is now recognized that the hydrogen atom, with only one stable orbital (the 1s orbital), can form only one covalent bond, that the hydrogen bond is largely ionic in character, and that it is formed only between the most electronegative atoms.This definition implies that the hydrogen bond consists of two components: A hydrogen bond donor XH, where the H atom is covalently bonded to an atom X with a large electronegativity, and another electron rich atom B, the hydrogen bond acceptor. Typical examples are hydrogen bonds between water molecules or between amide groups in peptides. However, the second statement by Pauling limits the hydrogen bond to a few atoms with high electronegativity. Because nowadays many other bonding situations such as CH B are also referred to as hydrogen bond further definitions were suggested. The most recent originates from a IUPAC conference in the year 2005 [2]:

Transcript of ORCA labs 10 weak interactions...

Page 1: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     1  

1 Computer  Experiment  10:  Intermolecular  and  Weak  Interactions  

1.1 Background  In  this  section  we  treat  intermolecular  and  weak  interactions,  this  means  non-­‐covalent  

and  non  ionic  interactions  between  different  atoms  or  molecules.  We  will  focus  our  

attention  to  the  two  most  familiar  representatives  of  such  interactions:  Hydrogen  bonds  

and  van  der  Waals  or  dispersion  interactions.  Other  weak  intermolecular  interactions  

like  π-­‐π  stacking  or  dipole-­‐dipole  interactions  are  closely  related  to  these  two  basis  

types  of  interaction.  

1.1.1 Hydrogen  bonds  Hydrogen  bonds  belong  to  the  most  important  non-­‐covalent  intermolecular  interactions.  

They  are  operative  in  determining  molecular  conformation,  molecular  aggregation,  and  

the  function  of  a  vast  number  of  chemical  systems  ranging  from  inorganic  to  biological.  

In  the  majority  of  cases  hydrogen  bonding  occurs  between  a  polar  bond  Xδ-­‐-­‐Hδ+  and  an  

electron  rich  region  Bδ-­‐.  Hydrogen  bonding  can  be  understood  as  a  donor-­‐acceptor  

interaction.  Therefore  we  will  denote  X-­‐H  as  the  (hydrogen)  donor  and  B  as  the  

(hydrogen)  acceptor.  

Definition:  

The  first  definition  of  the  hydrogen  bond  was  given  by  Pauling  [1]:  “Under  certain  

conditions  a  hydrogen  atom  is  attracted  by  rather  strong  forces  to  two  atoms,  instead  of  

only  one,  so  that  it  may  be  considered  to  be  acting  as  a  bond  between  them.”  He  further  

added  that  “it  is  now  recognized  that  the  hydrogen  atom,  with  only  one  stable  orbital  (the  

1s  orbital),  can  form  only  one  covalent  bond,  that  the  hydrogen  bond  is  largely  ionic  in  

character,  and  that  it  is  formed  only  between  the  most  electronegative  atoms.”    

This  definition  implies  that  the  hydrogen  bond  consists  of  two  components:  A  hydrogen  

bond  donor  X-­‐H,  where  the  H  atom  is  covalently  bonded  to  an  atom  X  with  a  large  

electronegativity,  and  another  electron  rich  atom  B,  the  hydrogen  bond  acceptor.  Typical  

examples  are  hydrogen  bonds  between  water  molecules  or  between  amide  groups  in  

peptides.  

However,  the  second  statement  by  Pauling  limits  the  hydrogen  bond  to  a  few  atoms  with  

high  electronegativity.  Because  nowadays  many  other  bonding  situations  such  as  C-­‐H…B  

are  also  referred  to  as  hydrogen  bond  further  definitions  were  suggested.  The  most  

recent  originates  from  a  IUPAC  conference  in  the  year  2005  [2]:  

Page 2: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     2  

 

“Hydrogen  bonding  occurs  when  an  electron  deficient  hydrogen  that  is  bonded  to  an  

atom,  has  an  attractive  interaction  with  another  electron  rich  region  either  within  the  

same  or  another  molecular  entity”  

Hydrogen  bonding  can  occur  in  many  different  fashions.  Mostly  one  distinguishes  

between  strong,  moderate  and  weak  hydrogen  bonds.  Moderate  or  normal  hydrogen  

bonds  have  interaction  energies  between  20  and  60  kJ/mol,  common  examples  are  

hydrogen  bonds  between  water  molecules  or  amide  groups.  Weak  hydrogen  bonds  have  

interaction  energies  below  20  kJ/mol.  An  example  for  a  weak  hydrogen  is  the  interaction  

between  a  methyl  group  (C-­‐H  as  acceptor)  and  the  π-­‐electrons  of  benzene.  Strong  

hydrogen  bonds  are  characterized  by  the  fact  that  the  H  atom  lies  close  to  the  mid  point  

of  the  A…B  line  of  centers  (an  example  is  the  negatively  charged  complex  [F…H…F]-­‐).  The  

interaction  energy  is  larger  than  60  kJ/mol.    

A  further  important  point  concerning  to  hydrogen  bonds  is  their  cooperativity.  The  

properties  of  a  network  of  n  hydrogen  bonds  are  not  the  sum  of  n  isolated  hydrogen  

bonds.  This  cooperativity  or  nonadditivity  is  caused  by  the  polarizability  or  charge  

transfer  character  of  hydrogen  bonding.  One  example  for  this  cooperativity  is  the  

following  hydrogen  bonded  chain:  

Yδ-­‐-­‐Hδ+…Xδ-­‐-­‐Hδ+…Bδ-­‐    

Both  Yδ-­‐-­‐Hδ-­‐  and  Bδ-­‐  effect  a  polarization  of  the  middle  Xδ-­‐-­‐Hδ+  group.  Through  this  

opposite  polarization  both  hydrogen  bonds  became  stronger.        

Theory  of  hydrogen  bonds  

Hydrogen  bonding  is  a  quite  complex  interaction.  In  earlier  days  it  was  believed  that  

hydrogen  bonding  is  a  purely  electrostatic  interaction  between  Xδ-­‐-­‐Hδ+  and  Bδ-­‐  (In  fact  

electrostatic  models  are  often  able  to  predict  reasonable  geometries  for  hydrogen  

bonds).  However  a  pure  electrostatic  model  is  not  able  to  explain  e.g.  the  spectroscopic  

properties  of  hydrogen  bonds.  A  charge  transfer  from  Bδ-­‐  to  Xδ-­‐-­‐Hδ+    is  certainly  also  

involved  in  hydrogen  bonding  (This  is  the  reason  why  some  authors  call  Bδ-­‐  the  

(electron)  donor  and  Xδ-­‐-­‐Hδ+    the  (electron)  acceptor).  Furthermore,  of  course  it  is  not  

possible  to  treat  any  intermolecular  interactions  without  dispersion  and  repulsion  

forces.  Hence  hydrogen  bonding  is  the  sum  of  many  different  effects.  Morokuma,  in  

1977,  introduced  an  energy  decomposition  scheme  [3]  which  divides  the  interaction  

energy  EI  in  the  following  components:    

Page 3: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     3  

EI= E

S+ E

PL+ E

X+ E

CT+ E

DISP+ E

MIX         (1)  

Morokuma  explains  the  physical  meanings  of  these  components  as:   ES  is  the  

electrostatic  interaction,  i.e.  the  interaction  between  the  undistorted  electron  

distribution  of  the  donor  (X-­‐H)  and  that  of  the  acceptor  (B).  This  contribution  includes  

the  interaction  of  all  permanent  charges  and  multipoles,  such  as  dipole-­‐dipole,  dipole-­‐

quadrupole,  etc..   ES  is  oriented  and  has  a  long  range  (it  decreases  with  -­‐r-­‐3  for  dipole-­‐

dipole  and  with  -­‐r-­‐2  for  dipole–monopole  interactions).        

EPL  is  the  polarization  interaction,  i.e.  the  effect  of  the  distortion  (polarization)  of  the  

electron  distribution  of  the  donor  by  the  acceptor,  the  distortion  of  the  acceptor  by  the  

donor,  and  the  higher  order  coupling  resulting  from  such  distortions.  This  component  

includes  the  interactions  between  all  permanent  charges  or  multipoles  and  induced  

multipoles,  such  as  dipole-­‐induced  dipole,  quadrupole-­‐induced  dipole,  etc..   EPL  

decreases  with  –r-­‐4.  

EX  is  the  “exchange  repulsion”,  i.e.  the  interaction  caused  by  exchange  of  electrons  

between  the  acceptor  and  the  donor.  More  physically,  this  is  the  short-­‐range  repulsion  

due  to  overlap  of  electron  distribution  of  the  donor  with  that  of  the  acceptor.   EX  

increases  with  +r12.  

ECT  is  the  charge  transfer  or  electron  delocalization  interaction,  i.e.  the  interaction  

caused  by  charge  transfer  from  occupied  orbitals  of  the  acceptor  to  vacant  orbitals  of  the  

donor,  and  from  occupied  orbitals  of  the  donor  to  vacant  orbitals  of  the  acceptor,  and  the  

higher  coupled  interactions.  The   ECT  contribution  decreases  approximately  with  -­‐e-­‐r.  

EDISP  is  the  dispersion  energy,  the  stabilization  due  to  the  correlation  of  electronic  

motions  in  the  acceptor  and  the  donor.    This  interaction  is  described  by  simultaneous  

and  correlated  excitations  of  electrons  in  both  molecules.   EDISP  is  isotropic  and  

decreases  with  –r-­‐6.  

EMIX  includes  higher  order  interactions  between  various  components.    

One  must  keep  in  mind  that  each  hydrogen  bond  has  its  own  weighting  of  these  

components  (depending  on  the  acceptor,  the  donor  and  their  environment  (geometry).  

In  particular  the  different  distance  dependencies  of  the  components  are  important.    For  

Page 4: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     4  

example  a  hydrogen  bond  gets  more  electrostatic  if  the  distance  r  between  the  donor  

and  the  acceptor  is  increased.    

1.1.2 Van  der  Waals  interaction  In  Chemistry,  the  term  van  der  Waals  interaction  originally  referred  to  all  

intermolecular  interactions.  According  to  this  one  definition  a  van  der  Waals  interaction  

is:    

“All  intermolecular  interactions  and  interaction  between  two  atoms/groups  in  a  molecule  

that  are  not  directly  bonded  are  called  van  der  Waals  interactions.”  [2]  

 

However,  the  term  van  der  Waals  interaction  is  often  only  used  for  intermolecular  

interactions  which  arise  from  the  polarization  of  molecules  into  dipoles  or  multipoles.  In  

this  case  one  also  speaks  of  dispersion  or  London  interaction:        

“Interaction  between  two  atoms/groups  arising  primarily  due  to  dispersive  forces  is  called  

London  interaction.”  [2]  

In  this  script  we  will  use  the  term  van  der  Waals  interaction  as  synonum  for  London  or  

dispersion  interaction.  Even  though  dispersion  interactions  are  the  weakest  non-­‐

covalent  interactions  they  are  very  important  in  chemistry,  since  they  are  the  only  

attractive  force  at  large  distance  present  between  neutral  atoms  like  noble  gases  or  

nonpolar  molecules  like  alkanes.  For  example  without  dispersion  interaction  it  would  be  

no  possibility  to  obtain  noble  gases  in  a  liquid  form.  Furthermore  dispersion  forces  

become  stronger  as  the  atoms  or  molecules  become  larger.  This  is  due  to  the  increased  

polarizability  of  molecules  with  larger,  more  dispersed  electron  clouds.  

Theory  of  the  Dispersion  interaction  

Dispersion  or  London  force  is  an  attractive  interaction  which  arises  from  an  interplay  

between  electrons  belonging  to  the  densities  of  two  otherwise  non-­‐interacting  atoms  or  

molecules.  It  originates  from  the  fact  that  electron  density  is  even  in  nonpolar  molecules  

not  evenly  distributed  in  space.  At  intermediate  distances  the  motion  of  electrons  in  one  

molecule  induces  slight  perturbations  in  the  otherwise  evenly  distributed  electron  

densities  of  the  neighbouring  molecule.  This  correlation  leads  to  a  temporary  dipole  

moment.  The  induced  dipole,  in  turn,  induces  a  polarization  of  the  electron  density  of  the  

other  molecule.  Through  this  it  comes  to  an  attractive  interaction  between  the  two  

molecules.  

As  one  knows  from  Physical  Chemistry  such  induced  dipole-­‐induced  dipole  interactions  

decay  with  –r-­‐6  (r  is  the  distance  between  the  two  molecules).  For  completeness  one  

Page 5: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     5  

should  mention  that  the  presence  of  interactions  from  higher  order  electric  moments  

like  induced  quadrupole-­‐induced  dipole  or  induced  quadrupole-­‐induced  quadrupole  

interactions  leads  also  to  terms  which  vary  with  –r-­‐8,  –r-­‐10  and  so  on.    

The  repulsion  interaction  between  the  electrons  of  the  two  interacting  molecules  acts  in  

the  opposite  direction.  The  Pauli  repulsion  or  exchange  interaction  arises  with  +r12.  

Dispersion  and  exchange  interaction  are  the  sole  electronic  interactions  which  occur  in  

all  molecules.  Together  they  give  the  well  known  Lennard-­‐Jones  Potential  

V(r)!

1r12"

1r 6                 (2)  

where  higher  order  dispersion  interactions  are  neglected.    

Calculation  of  interaction  energies:  

The  most  common  Ansatz  for  the  calculation  of  interaction  energies  of  hydrogen  or  van  

der  Waals  bonded  complexes  is  the  supermolecular  approach.  In  this  Ansatz  the  total  

intrinsic  interaction  energy   EI  of  a  complex  AB  which  is  build  from  the  monomers  A  

and  B  is  defined  as:  

EI= E(R

AB)!E(R

A)!E(R

B)             (3)  

E R

AB( )  is  the  total  energy  of  the  complex  AB,   E R

A( )  and   E R

B( )  are  the  total  energies  of  the  isolated  molecules  A  and  B.  It  should  be  noticed  that  the  supermolecular  Ansatz  

describes  the  total  interaction  energy  of  the  complex.  For  example  it  is  not  possible  to  

determine  individual  hydrogen  bond  energies  if  there  is  more  than  one  hydrogen  bond  

broken  upon  dissociation.  Moreover  the  supermolecular  Ansatz  can  not  be  used  for  the  

calculation  of  intramolecular  interactions.    

BSSE  and  Counterpoise  Correction:  

Calculation  of  the  interaction  energy  by  means  of  the  supermolecular  Ansatz  involves  an  

important  inconsistency.  When  the  total  energy  for  the  isolated  monomers  A  or  B  is  

calculated  only  the  basis  functions  of  the  particular  molecule  (A  or  B)  are  available  to  

describe  each  electronic  spatial  orbital.  Whereas  when  the  total  energy  of  the  complex  

AB  is  calculated  the  basis  functions  of  one  molecule  can  help  compensate  for  the  basis  

set  incompleteness  on  the  other  molecule,  and  vice  versa.  In  effect  the  basis  set  of  the  

complex  is  larger  than  that  one  of  each  monomer.    This  produces  an  artificial  lowering  of  

the  complex  energy  relative  to  that  of  the  separated  monomers  and  as  a  consequence  

thereof  an  overestimation  of  the  interaction  energy.  This  effect  is  called  basis-­‐set  

superposition  error  (BSSE).  It  should  be  clear  that  the  larger  the  used  basis  set  the  

Page 6: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     6  

smaller  is  the  BSSE.  In  the  limit  of  a  complete  basis  set  the  BSSE  would  vanish  since  

adding  basis  functions  would  not  lead  to  an  improvement  in  this  case.  However,  

calculations  with  a  (nearly)  complete  basis  set  are  not  possible  (or  not  reasonable)  in  

the  majority  of  cases.    

In  order  to  get  accurate  values  for  the  interaction  energies  it  is  necessary  to  find  a  

practicable  correction  for  the  BSSE.    The  most  common  used  method  for  an  

approximated  determination  of  the  BSSE  is  the  Counterpoise  correction  [4].  In  the  

counterpoise  correction  one  estimates  the  above  described  artificial  lowering  of  the  

total  energy  of  the  complex  AB.  For  this  purpose  four  additional  calculations  are  

necessary:  

1.) One  calculates  E(R´A,a),  the  energy  of  the  isolated  monomer  A  with  the  geometry  

R´A  it   has   in   the   complex  AB.   For   this   calculation  only   the  basis   functions  a   for  

molecule  A  are  used.  

2.) One   calculates   E(R´A,ab),   the   energy   of   the   isolated   monomer   A   with   the  

geometry   it  has   in   the  complex  AB.  However,  now  the  complete  basis  ab  of   the  

complex  AB,   that  means  the  basis  functions  for  molecule  A  and  B  are  used.  The  

basis  functions  b  of  molecule  B  are  located  at  the  corresponding  positions  of  the  

nuclei,  but  the  B  nuclei  are  not  present.  These  positions  in  space  where  only  basis  

functions   but   no   nuclei   are   located   are   often   called   ghost   atoms   or   ghost  

orbitals.  

3.) According  to  step  1.)  one  calculates  E(R´B,b),  the  energy  of  B  with  the  geometry  it  

has  in  the  complex  AB    (of  course  with  the  basis  functions  b).  

4.) According   to   step   2.)   one   calculates   the   energy  E(R´B,ab)   of  B   in   the   complete  

basis  ab  of  AB.  

For  the  BSSE  we  get:  

EBSSE  =  {E(R´A,a)  -­‐  E(R´A,ab)}  +  {E(R´B,b)  -­‐  E(R´B,ab)}       (4)  

And  for  the  counterpoise  corrected  interaction  energy  EIcc  we  get:  

EIcc  =  EI  +  EBSSE                 (5)  

 The  calculation  of  E(R´Aa)  and  E(R´B,b)  is  quite  easy.  One  deletes  the  coordinates  of  A  

respectively  B  from  the  optimized  structure  AB  and  performs  a  normal  calculation  with  

the  chosen  basis  set.  The  calculation  of  E(R´A,ab)  and  E(R´B,ab)  can  be    more  

Page 7: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     7  

complicated  because  the  ghost  atoms  must  be  defined.  ORCA  offers  a  reasonably  

convenient  solution:  

If  one  wants  a  ghost  atom  with  specific  basis  functions  at  certain  coordinates  it  is  only  

necessary  to  add  in  the  input  file  a  colon  “  :  “  behind  the  atom  symbol.    

Example  input  for  calculation  of  E(R´A,ab)  (AB  is  a  water  dimmer):  

 Computational  Methods:  

Before  one  starts  with  the  calculation  of  hydrogen  bonded  or  van  der  Waals  bonded  

systems  one  must  choose  a  suitable  computational  method.  Nowadays  density  

functional  theory  is  the  most  common  used  method  for  standard  computational  

calculations.  However,  DFT  methods  have  some  shortcomings  which  narrow  their  ability  

to  describe  such  situations.  The  main  disadvantage  of  DFT  methods  in  this  context  is  

their  deficient  description  of  dispersion.  

Since  the  van  der  Waals  interaction  is  a  pure  dispersion  effect  it  should  be  obvious  that  

DFT  is  the  wrong  choice  for  the  calculation  of  such  interactions.  One  must  mention  that  

one  can  get  quite  good  results  for  von  der  Waals  complexes  with  some  functionals.  

However,  it  is  not  clear,  if  this  is  “the  right  result  for  the  right  reason”.      

The  situation  is  more  complicated  for  hydrogen  bonds  since  their  interaction  energy  can  

be  described  as  the  sum  of  different  contributions.  Depending  on  the  situation  DFT  

methods  are  more  or  less  suitable:  For  weak  hydrogen  bonds,  where  dispersion  

interactions  are  more  important,  DFT  is  a  bad  choice,  whereas  it  delivers  good  results  

for  normal  hydrogen  bonds  which  are  a  mainly  electrostatic  interaction.  However,  even  

in  these  cases  one  should  verify  DFT  results  with  a  method  with  a  better  description  of  

electron  correlation.    

Coupled  Cluster  methods  like  CCSD(T)  with  a  large  basis  set  would  be  an  excellent  

choice.  However,  due  to  its  huge  computational  costs  such  calculations  are  not  

practicable  in  most  cases.  Instead  of  one  common  uses  MP2  calculations  for  the  

calculation  of  van  der  Waals  and  hydrogen  bonded  complexes.  MP2  combines  a  good  

description  of  correlation  effects  like  dispersion  with  moderate  computational  costs.  

! RKS B3LYP TZVP TightSCF * xyz 0 1 O 0.081307 -0.360387 0.357004 H -0.259589 0.505266 0.499968 H 0.926815 -0.246293 -0.052696 O: 2.753726 -0.032808 -0.943497 H: 3.521598 -0.305170 -0.469305 H: 2.859663 -0.336280 -1.829698 end

Page 8: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     8  

Further  Reading:  

The  physical  background  of  the  underlying  forces  of  weak  interactions  is  well  explained  

in  many  Physical  Chemistry  textbooks.  A  more  detailed  description  of  hydrogen  bonding  

can  be  obtained  from  an  excellent  review  article  from  Thomas  Steiner  [5]  or  some  

textbooks  [6-­‐9].  Accurate  computation  of  weak  interactions  is  still  a  challenge  in  

Quantum  Chemistry.  For  the  problems  of  DFT  with  the  description  of  weak  interactions  

we  especially  refer  to  the  book  by  Koch  and  Holthausen  [10].  

1.2  Description  of  the  Experiment  

1.2.1 Exercise  1:  Investigation  of  the  water  dimer    In  the  first  exercise  we  will  study  the  water  dimer.  Since  it  is  well  known  that  different  

conformations  are  possible  we  will  begin  the  calculation  at  different  starting  points.  

Proceeding:  

• Build  a  z-­‐matrix  for  different  conformations  of  the  water  dimer.  Try  to  create  at  least  

one   linear,  one  cyclic  and  one  bifurcated  dimer.  290  pm  is  a  reasonable  start  value  

for  the  distance  ROO  between  the  two  oxygen.  

• Perform   geometry   optimizations   of   the   different   conformers.   Use   the   B3LYP,   RHF,    

MP2  and  CCSD(T)  methods1.  Update  the  basis  set  from  SVP  to  TZVP  to  TZVPP  to  aug-­‐

TZVPP.  Study  how  the  BSSE  behaves  with  basis  sets  of  increasing  size.  Do  the  diffuse  

functions  in  aug-­‐TZVPP  improve  the  results?    

• Calculate  the  interaction  energy  by  means  of  the  supermolecular  approach.  

• Calculate  the  counterpoise  corrected  interaction  energies    

• Compare   the   results   (interaction   energy,   BSSE   and   geometry)   which   you   have  

obtained   with   the   different   methods   among   each   other   and   compare   them   to  

experimental  result  (interaction  energy  =  5.0±0.7  kcal/mol  =  20.9±2.9  kJ/mol)2  

1.2.2 Exercise  2:  Noble  gas  dimers  In  this  section  we  study  the  potential  energy  surface  (PES)  of  the  He  and  the  Ne  dimer.  

For  this  purpose  we  calculate  the  energy  of  the  He  and  Ne  dimer  at  different  interatomic  

                                                                                                               1  Note  that  you  obtain  the  RHF  results  as  a  byproduct  of  the  MP2  calculation.  2  Quoted   from  Mas,   E.M.;   Bukowski,   R.;   Szalewicz,   K.;   Groenenboom,  G.C.;  Wormer,   P.E.S.;   van   der   Avoird,   A.   (2000),   J.  Chem.  Phys.,  113,  6687.  

Page 9: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     9  

distances  (RHe-­‐He  or  RNe-­‐Ne).  The  calculations  of  this  exercise  should  be  performed  both  

with  MP2  and  B3LYP  each  with  a  TZVPP  basis  (for  higher  accuracy  aug-­‐TZVPP).  

• Begin  the  calculation  with  an  interatomic  distance  (RHe-­‐He  /  RNe-­‐Ne)  which  is  twice  the  

van  der  Waals  radius  of  the  treated  atoms.  

• Vary  (increase  and  decrease)  this  distance  in  steps  of  10  pm.  You  should  consider  at  

least  10  points.  

• Calculate  for  each  point  the  interaction  energy  (EI  and  EICC).  

• Nearby  the  minimum  you  can  choose  a  closer  meshed  net  for  your  calculation.  

• Compare   the   results   of   the   MP2,   CCSD(T)   and   B3LYP   calculation   and   discuss   the  

importance  of  the  BSSE  in  this  case.  Compare  your  results  to  accurate  computational  

data  (Basis-­‐set  limit  MP2  values:3  He-­‐dimer:  0.06  kJ/mol,  R(He-­‐He)=3.061  Angström;  

Ne-­‐dimer:  0.23  kJ/mol,  R(Ne-­‐Ne):  3.202  Angström;  CCSD(T)-­‐values:4  He-­‐dimer:  0.09  

kJ/mol,   R(He-­‐He):   2.977   Angström;   Ne-­‐dimer:   0.34   kJ/mol,   R(Ne-­‐Ne)=   3.101  

Angström).  

Literatur:  [1]     L.  Pauling,  The  Nature  of  the  Chemical  Bond,  Cornell  University  Press,  Ithaca,  NY,  

1939  [2]     E.  Arunan,  R.  Klein,  IUPAC  workshop  “Hydrogen  Bonding  and  Other  Molecular  

Interactions”,  Pisa,  2005  [3]     H.  Umeyama,  K.  Morokuma,  JACS,  1977,  99,  1316  [4]     a)  S.  F.  Boys,  F.  Bernardi,  Mol.  Phys.  1970,  19,  553               b)  F.  B.  van  Duijneveldt,  J.  G.M.  van  Duijneveldt-­‐van  de  Rijdt,  J.  H.  van  Lenthe,  Chem.  

Rev.  1994,  94,  1873  [5]     Th.  Steiner,  Angew.  Chem.  2002,  114,  50  [6]     G.  A.  Jeffrey,  W.  Saenger,  Hydrogen  Bonding  in  Biological  Structures,  Springer,  

Berlin,  1991  [7]     G.  A.  Jeffrey,  An  Introduction  to  Hydrogen  Bonding,  Oxford  University  Press,  Oxford,  

1997  [8]     S.  Scheiner,  Hydrogen  Bonding.  A  Theoretical  Perspective,  Oxford  University  Press,  

Oxford,  1997  [9]     G.  R.  Desirju,  Th.  Steiner,  The  Weak  Hydrogen  Bond,  Oxford  University  Press,  

Oxford,  2001  [10]    W.  Koch,  M.  C.  Holthausen,  A  Chemist’s  Guide  to  Density  Functional  Theory,  Wiley-­‐

VCH  Weinheim,  2000  

                                                                                                               3  Tew,  D.P.;  Klopper,  W.  (2006)  J.  Chem.  Phys.,  125,  094302  4  Van  Mourik,  T.;  Wilson,  A.;  Dunning,  T.H.(1999)  Molec.  Phys.,  96,  529  

Page 10: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     10  

2 Computer  Experiment  11:  Ions  and  Solvation  Effects  

2.1  Background  

2.1.1 Introduction  Solvent  effects  play  an  important  role  in  all  areas  of  chemistry.  In  it's  broadest  sense  

solvent  effects  can  be  defined  as  the  change  of  a  solute's  properties  in  the  presence  

of  a  solvent,  compared  to  the  gas  phase.    This  can  relate  to  changes  in  geometry,  as  

well  as  in  electric  and  magnetic  properties.  

In  the  field  of  quantum  chemistry  people  looked  for  ways  to  overcome  the  necessity  

of  adding  explicit  solvent  molecules,  when  trying  to  include  solvent  effects  into  their  

systems.  These  implicit  solvent  models    usually  treat  the  molecule's  surroundings  as  

a  dielectric  continuum,  which  reacts  to  the  solute's  charge  distribution  and  in  turn  

influences  the  solutes  charge  distribution.    Two  popular  approaches  of  these  

dielectric  models  are  the  Polarizable  Continuum  Model  (PCM),  and  Conductor-­‐like  

Screening  Models  (COSMO).  In  these  models  the  solute's  cavity  is  either  represented  

by  a  series  of  atom-­‐centered  spheres  or  by  a  solvent-­‐accessible  /  solvent-­‐excluding  

surface.  The  solute/solvent  boundary  is  divided  into  small  surface  elements  and  the  

Poisson  Equation  is  solved  on  each  of  them  to  yield  an  apparent  surface  charge  

(ASC).    The  COSMO  method  treats  the  surrounding  continuum  as  conductor  which  

greatly  facilitates  the  solution  of  the  electrostatic  equations.  The  ASCs  calculated  

this  way  are  later  modified  to  yield  a  result  for  a  dielectric  medium.  These  models  

can  generally  reproduce  the  experimental  solvation  energies  for  small  charged  

molecules  and  are  precise  up  to  1  kcal/mol  for  small  uncharged  molecules.    

2.1.2 Theory  The  most  versatile  approach  to  implement  a  continuum  solvent  model  are  apparent  

surface  charges  (ASC).    These  can  be  derived  by  noting  that  the  charge  associated  

with  each  surface  part,  can  be  expressed  as  the  difference  between  the  polarization  

vectors  at  regions  where  different  dielectric  media  come  into  contact:5  

                                                                                                               5  For  an  introduction  into  the  treatment  of  dielectric  media,  see  for  example  Feynman,  “Lectures  in  Physics”,  Vol.  2  

Page 11: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     11  

!

12= ! P

2!P

1( )n12          

 (6)  here  σ  is  a  charge  distribution  on  the  cavity  surface,  and   n12

the  normal  vector  from  

one  region  to  the  next.  The  polarization  vector  is  related  to  the  potential,  which  is  

generated  by  its  region  by    

P

i= !!!14!"#          

 (7)  

In  this  case,  only  two  regions  exist,  namely   !1 = 1; !2= !  for  the  inside  of  the  cavity  

and  the  outside,  so  the  boundary  charges  can  be  calculated  by    

!=

1-"4!"!"

in

!n             (8)  

where   !in  is  generated  by  the  solute's  charge  distribution  and  the  generated  

surface  charges,  or  written  in  terms  of  the  electric  field  

!="!14!"

En            

 (9)  

These  surface  charges  can  then  be  used  in  quantum  chemical  calculations  to  

influence  the  wavefunction  which  in  turn  gives  rise  to  an  new  potential  on  the  

surface  charges.  This  process  is  continued  until  self  consistensy  is  reached  (self-­‐

consistent  reaction  field,  SCRF).  

In  the  framework  of  the  COSMO  model  the  surrounding  is  treated  as  a  conductor,  

such  that  the  potential  vanishes  at  the  surface.  Due  to  this  condition  the  equations  

become  much  easier  to  solve  and  also  much  easier  to  implement  into  quantum  

chemical  codes.  

2.2 Description  of  the  Experiment  

2.2.1  Calculation  of  Solvation  Energies  of  Small  Molecules  Calculate  the  solvation  energy  of  the  following  molecules:  

● Water  

Page 12: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     12  

● Acetone  

● n-­‐Octane  

● Benzene  

● OH-­‐  

● Cl-­‐  

The  calculation  includes  several  steps:  

1. Optimization   of   the   molecule's   geometry   in   the   gas   phase   using   Density  

Functional   Theory   (B3LYP   functional)   with   the   SVP   basis   set:  ! RKS B3LYP SVP TightSCF TightOpt

2. Optimization  of  the  structure  in  solution  using  the  COSMO  solvation  model  at  

B3LYP/SVP   level:  ! RKS B3LYP SVP COSMO(water) TightSCF TightOpt

Using  the  final  energies  resulting  from  these  calculations  the  electrostatic  energy  of  

solvation  can  be  calculated.  What  do  you  observe  if  you  look  at  the  absolute  

numbers?  Do  the  results  for  Benzene  look  odd  to  you?  Which  contributions  do  you  

think  are  missing  to  compute  a  Free  Energy  of  solvation?  

2.2.2 Solvent  Shifts  on  Electronic  Spectra  Some  molecules  show  a  different  absorption  spectrum  when  immersed  in  liquids  of  

different  polarity.  One  of  these  molecules  is    

 One  of  the  fastest  ways  to  calculate  a  π  →  π*  transition  is  the  TD-­‐DFT  method.  An  

ORCA  input  for  this  calculation  may  be  constructed  as  follows:  

Page 13: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     13  

 The  transitition  that  we  are  studying  is  essentially  the  HOMO-­‐LUMO  transition  and  

is  the  first  excited  state  of  the  system.  The  solvatochromic  shift  is  calculated  as  the  

energy  difference  between  the  excitation  energy  in  hexane  and  water  (just  put  in  

COSMO(hexane)).    

To  produce  reasonable  results  it  is  absolutely  necessary  to  optimize  the  geometry  of  

the  molecule  in  the  solvent  under  consideration.  Thus,  the  geometric  relaxation  in  a  

different  solvents  provides  an  important  contribution  to  the  observed  solvent  

shifts.6  

The  experimental  band  maxima  are  observed  at  20830  wavenumbers  in  hexane  and  

18830  wavenumbers  in  water.  Do  the  calculation  results  differ  from  the  

experimental  values?  If  so,  how  do  you  explain  the  discrepancies?    

2.2.3 Calculation  of  Glycine  in  Gas-­‐  and  Liquid  Phase  As  one  of  the  twenty  naturally  occurring  amino  acids  glycine  is  a  molecule  of  high  

biological  significance.  The  aminoacids  are  composed  -­‐  in  their  monomeric  forms  –  

of  a  central  carbon  atom  to  which  an  amino  group,  a  carboxyl  group,  a  hydrogen  and  

a  variable  sidechain  are  attached.  The  peptide  bond  between  these  monomers  is  

created  as  an  amide  condensation  between  the  carboxyl  group  of  one  amino  acid  

and  the  amino  group  of  another  one.  The  smallest  amino-­‐acid  is  glycine,  in  which  the  

side  chain  is  only  made  up  by  an  additional  hydrogen  atom.  In  its  natural  

surrounding,  that  is  -­‐  in  solution  -­‐  glycine  takes  the  form  of  a  'zwitter'  ion.  This  

means,  that  both,  the  amino  group  and  the  carboxyl  group  are  present  in  their  ionic  

form,  although  the  molecule  as  a  whole  remains  neutral.  

                                                                                                               6  For  experimental  results  see  Reichardt,  “Solvents  and  Solvent  Effects  in  Organic  Chemistry”,  p.  335  

! RKS BP RI TZVPP TZVPP/C COSMP(water) TightSCF %tddft Nroots 5 Maxdim 50 End *xyz 0 1 . . *

Page 14: ORCA labs 10 weak interactions solvationscienide2.uwaterloo.ca/.../ORCA_labs_10_weak_interactions_solvatio… · IntermolecularandWeakInteractions, , 1, 1 ComputerExperiment*10:IntermolecularandWeakInteractions*

Intermolecular  and  Weak  Interactions     14  

 Figure  1:  The  neutral  and  zwitterionic  forms  of  glycine.  

Thus  this  molecule  presents  a  demanding  test  for  an  implicit  solvent  model,  that  is,  

can  it  stabilize  the  glycine  molecule  in  its  zwitterionic  form  and  therefore  predict  

the  correct  energetic  preference?  To  answer  this  question,  calculate  (optimize)  the  

structure  of  the  glycine  molecule  in  the  gas  phase  and  in  solution  with  different  

dielectric  constants.  What  is  the  critical  value  of  the  dielectric  constant  where  the  

preference  changes  from  the  neutral  to  the  zitterionic  form?