6th International Conference on Gas Hydrates

UV-VISIBLE AND RESONANCE RAMAN SPECTROSCOPY OF HALOGEN MOLECULES IN CLATHRATE-HYDRATES Janda, Kenneth C.; Kerenskaya, Galina; Goldscheleger, Ilya U.; Apkarian, V. Ara; Fleischer, Everly B. 2008-07-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


5393.pdf [ 922.05kB ]
JSON: 1.0041100.json
JSON-LD: 1.0041100+ld.json
RDF/XML (Pretty): 1.0041100.xml
RDF/JSON: 1.0041100+rdf.json
Turtle: 1.0041100+rdf-turtle.txt
N-Triples: 1.0041100+rdf-ntriples.txt
Original Record: 1.0041100 +original-record.json
Full Text

Full Text

   1    UV-VISIBLE AND RESONANCE RAMAN SPECTROSCOPY OF HALOGEN MOLECULES IN CLATHRATE-HYDRATES    Kenneth C. Janda*, Galina Kerenskaya, Ilya U. Goldscheleger,  V. Ara Apkarian and Everly B. Fleischer   Department of Chemistry University of California, Irvine Irvine, CA 92697-2025 U.S.A.   ABSTRACT Ultraviolet-visible spectra are presented for a polycrystalline sample of chlorine clathrate hydrate and two single crystal samples of bromine clathrate hydrate. The data shows that the UV-visible spectroscopy  is  a  sensitive  probe  for  studying  the  interactions  between  the  halogen  guest molecule and the host water lattice. The spectrum for chlorine hydrate shows a surprisingly strong temperature  dependence.  The  spectra  reported  for  bromine  clathrate  hydrate  single  crystals reinforce  our  previous  conclusion  that  there  is  a  stable  cubic  type  II  structure  as  well  as  the tetragonal  structure.  There  is  also  a  metastable  cubic  type  I  structure.  The  new  results  are discussed  in  the  context  of  previous  results,  resonance  Raman  spectroscopy,  and  how  the molecules fit into the host cages.  Keywords: halogen clathrate hydrates, UV-visible, resonance Raman, polymorphism, spectrum shift versus cage size, water, ice                                                           * Corresponding author: Phone: 1-949-824-5266 FAX: 1-949-824-3168 Email: kcjanda@uci.edu NOMENCLATURE CS-I: The cubic clathrate hydrate structure, type I CS-II The cubic clathrate hydrate structure, type II HOMO: Highest occupied molecular orbital LUMO: Lowest unoccupied molecular orbital TS-I: The tetragonal structure of bromine hydrate X, B and C: The ground and two valence excited states of bromine that are most important for understanding its UV-visible spectroscopy ? and ?*: bonding and antibonding orbitals with ? symmetry ? and ?*: bonding and antibonding orbitals with ? symmetry ?e: harmonic vibrational frequency ?exe: the anharmonicity constant for the vibration ?max: the position of the peak intensity of an absorption band ??max: the shift of the band peak from that of the molecule in the gas phase 5x6y: cage naming notation. x = number of pentagonal faces, y = number of hexagonal faces  INTRODUCTION Although the halogen clathrate hydrate solids were first  discovered  almost  two  hundred  years  ago, [1,2],  and  although  they  have  been  extensively studied  since  then,  there  is  surprisingly  little spectroscopic  data  available  to  characterize  the interaction  between  the  halogen  guest  molecule and the water host lattice. Prior to the recent work in  Irvine[3,4],  the  only  spectroscopy  studies  for Proceedings of the 6th International Conference on Gas Hydrates (ICGH 2008), Vancouver, British Columbia, CANADA, July 6-10, 2008.     2 halogen  clathrates  was  a  Raman  study[5]  of  Cl2, Br2  and  BrCl.  Here,  we  report  new  UV-visible spectra  for  chlorine  and  bromine  hydrates  and compare  them  to  previously  reported  spectra  for bromine  and  iodine  trapped  in  clathrate  cages  as well as to the halogen spectra in various solvents. This new work is then discussed in the context of the  resonance  Raman  spectra  for  the  halogens  in clathrate  hydrate  cages.  Among  the  findings revealed by the spectra reported in this paper is a clear  confirmation  that  two  stable  and  one metastable  crystal  structures  exist  for  bromine clathrate  hydrate  under  specific  conditions  of temperature and bromine concentration.  Characterization  of  the  halogen  hydrates  has  been complicated  and  subject  to  error  from  the beginning.  Chlorine  hydrate  was  synthesized  by accident when Davy was trying to liquefy the gas. [1]  Seven  years  later  liquid  chlorine  was  finally observed  under  pressure.[6]  The  accidental production  of  chlorine  clathrate  hydrate  by  Davy was a preview of one of the reasons that clathrate-hydrate  research  regained  popularity  in  the  last century:  the  plugging  of  natural  gas  pipelines  by clathrate  deposits  due  to  residual  humidity  in  the natural gas.   Determining  the  stoichiometry  of  clathrates  was also difficult. The reason for this was clear after the determination  of  the  crystal  structure  of  chlorine hydrate by x-ray diffraction.[7] In a perfect crystal with complete cage filling the stoichiometry would be  fixed.  However,  in  chlorine  hydrate  the  larger 51262 cages are mostly occupied, but the occupation of  the  smaller  512  cages  depends  on  the  pressure maintained during formation.[8]  The variable cage occupancy, and thus variable stoichiometry, of gas clathrate  hydrates  still  presents  problems  for understanding their stability and thermodynamics in detail.  Determining the structure of bromine hydrate has been especially problematic. A tetragonal unit cell was determined  in  1963  by  Allen  and  Jeffrey.[9] However,  the  variable  stoichiometry  and  crystal shapes  persuaded  many  investigators  that  several crystalline  forms  could  co-exist.[10]  In  1997 Udachin et al.[11] determined that a wide variety of  bromine  hydrate  crystal  shapes  and stoichiometry  all  correspond  to  the  Allen  and Jeffrey  tetragonal  unit  cell  and  a  cage  structure made up of 512, 51262, and 51263 cages in a 10:16:4 ratio. We  will refer to this structure as tetragonal structure  I, or  TS-I. If  all of  the  larger  cages are occupied, then the limiting unit cell stoichiometry is (Br2)20(H2O)172; one bromine molecule for every 8.6 water molecules. Incomplete cage filling could account  for  H2O:Br2  ratios  as  high  as  10.7.  [11]  Here we report that three structures can readily be observed for bromine hydrate: the TS-I structure, the  cubic  type-I, CS-I  hydrate, and cubic  type-II, CS-II, hydrate. The CS-I and CS-II structures are the most common clathrate hydrate forms. CS-I is meta-stable  for  bromine,  but  is  kinetically preferred  under  certain  formation  conditions, while  the  CS-II  structure  is the  stable  form  for  a range  of  temperatures  and  compositions..  We conclude that at least three bromine hydrate crystal structures exist, and that more are possible.[12]  Although it might seem that the halogen hydrates are  not  so  important  for  understanding  the hydrocarbon hydrates that are currently of intense interest,  note  that  the  thermodynamics  data  for hydrates in general are referenced to an early study of bromine hydrate.[13] Since it now appears that this study assumed an incorrect crystal structure, a reexamination  of  the  bromine  hydrate  data  is appropriate. Also, a growing number of studies of the basic interactions of water molecules with the halogens will allow for a  more complete  analysis of  the  important  host  guest  interactions  in  the halogen  hydrates  than  may  be  possible  for  other systems.   Since  few  members  of  the  hydrate  research community  are  familiar  with  the  optical spectroscopy  of  the  halogens,  we  present  a  brief review  here.  Figure  1  shows  the  UV-visible spectrum  of  Br2  and  offers  a  pictorial interpretation.  For  the  ground  state,  called  the  X state, the orbitals are filled up to the ?* level; often referred  to  as  the  highest  occupied  molecular orbitals or HOMOs. The lowest energy electronic excitation involves the promotion of one electron from  the  HOMO  to  the  ?*  lowest  unoccupied molecular orbital, or LUMO.   Several  excited  electronic  states  can  result  from this  electronic  excitation,  but  the  absorption spectroscopy of  Cl2, Br2 and I2 are dominated by only two of these states. One, commonly called the C state, is repulsive so that the X ? C absorption spectrum  is  smooth  and  continuous,  even  in  the    3 gas  phase.  The  second  important  excited  state, commonly  called  the  B  state,  occurs  to  slightly lower energy of the C state because it contains a bonding  well.  In  the  gas  phase  the  X  ?  B  band contains a structured portion due to transitions to bound  states  and  also  a  continuum  due  to transitions  to  dissociative  states.  These  B  and  C states are close in energy and result in overlapping spectra. Formally, the B state is a triplet state and the X ? B transition is only allowed due to spin orbit coupling. Thus, this transition is quite weak for Cl2, and grows in intensity for Br2 and I2.  The spectrum  for  Br2,  shown  in  Fig.  1,  the  X  ?  B transition forms a shoulder on the low energy side of the stronger X ? C transition.     Figure 1 Ground and valence excited state potential curves of Br2 and the two most intense electronic transitions that contribute to the UV-visible spectrum. The inset shows a molecular orbital diagram. See text for more detail. _____________________________________________  The peak of the bromine absorption is in the near UV range,  resulting  in  its red color.   For  Cl2 the spectra  are  dominated  by  the  stronger  X  ?  C transition,  mostly  in  the  ultra  violet,  resulting  in the  pale  yellow  color  of  chlorine  gas.  For  I2  the two  transitions  blend  together  over  much  of  the visible  spectral  range,  resulting  in  the  purple, almost black, color of solid iodine.   EXPERIMENTAL In  our  previous  studies  of  the  optical  spectra  of bromine  and  iodine  hydrates  we  worked  with polycrystalline  samples.  The  new  data  reported here for chlorine hydrate was obtained by similar techniques.  Polycrystalline  samples  of  chlorine hydrate  were  grown  from  water  and  gaseous chlorine  on  a  surface  of  a  conventional  10  ?  10 mm  quartz  cell.  The  cell  used  a  stopcock  to contain  the  volatile  chlorine.  The  crystallization process  was  initiated  on  a  cell  wall  by  applying dry ice  above  the  water  level.  After  chlorine  and water  condensed  on  the  inner  surface  forming  a thin hydrate film of a pale green color, the cell was transferred  to  a  quartz  Dewar  for  spectroscopic measurements. The spectra were recorded at ~200 K and at ~77 K.    For  bromine  hydrate,  we  have  succeeded  in recording spectra for single crystal samples. First, a  polycrystalline  sample  was  prepared  using nanopure  water  and  99%  pure  bromine.  Upon melting,  this  sample  formed  a  supersaturated emulsion,  which  did  not  phase  separate  for  a considerable  time,  allowing  us  to  grow  crystals from the liquid. The emulsion stayed liquid down to  ?17  0C,  so  that  crystal  growth  could  be  con-trolled by varying the temperature. Crystal growth and spectroscopy were  performed using a  10 ?m thick cell that consisted of two windows separated by non-reactive fluorinated grease (Krytox?). The top  window  was  a  quartz  microscope  slide.  The bottom window was 0.5 mm thick sapphire placed on top of the two Peltier elements (separated by 3 mm  to  allow  the  light  to  pass  through  the  cell), which  were,  in  turn,  cooled  by  a  water-cooled copper block. The apparatus enabled us to control the  temperature  from  +20  0C  to  ?30  0C  with  a precision of 0.1 0C.   After  loading  the  emulsion  formed  by  melting  a polycrystalline sample into the cell, it was quickly cooled  to  ?20  0C  to  form  seed  crystals.  Upon heating  above  0  0C  the  ice  melted,  revealing  the small bromine clathrate crystals that were formed. These crystals were stable to +5.8 0C, the melting point  of  bromine  hydrate  in  the  tetragonal crystalline  form,  TS-I.  Careful  cycling  of  the temperature  around  the  hydrate  melting  point resulted in a single seed crystal in the optical path. This  crystal  was  then  allowed  to  grow  for  four days  at  +4.8  0C,  and  formed  an  optical  quality    4 crystal  that  filled  the  10  ?m  space  between  the windows and had a diameter of ? 0.5 mm. Under these  conditions  the  crystal  is  expected  to  be nearly  stoichiometric[14].  The  sample  was  then quickly  cooled  to  ?20  0C  in  order  to  freeze  the surrounding water and the spectrum reported in the next section of this paper was recorded.  After obtaining the spectrum of the TS-I bromine hydrate  as  described  in  the  preceding  paragraph, the  sample  was  warmed  above  0  0C  to  melt  the liquid  water  around  the  crystal  and  then  the temperature was dropped to  ?9 0C. As previously reported,[12]  when the  TS-I crystal is exposed to excess  water,  CS-II  crystals  grow  on  the  TS-I surfaces  at  this  temperature.  After  24  hours,  the CS-II hydrate crystals were large enough and were of optical quality. The spectra were recorded at ?9 0C.  RESULTS The  spectra  for  chlorine  hydrate,  the  two  single crystal samples of bromine hydrate, and iodine in hydrate cages, are shown in Fig. 2. Also shown in Fig.  2  are  spectra  of  halogens  in  other environments  for  comparison.  Fig.  2a  gives  the chlorine  data.  Four  curves  are  shown.  The  green curve is the gas phase spectrum; the blue curve is the spectrum of chlorine hydrate at 200 K; the gray curve is the hydrate spectrum at 77 K and the red curve  is the  spectrum in  aqueous solution with a small  amount  of  NaCl  added  to  eliminate  HOCl from  the  solution.  The  possible  contamination  of aqueous chlorine  spectra  by  HOCl  was discussed previously in detail.[15]  The intensity scales, optical density, for the spectra are separately adjusted so that each has the same peak intensity. To the precision of the experiment, the electronic spectrum of chlorine in the clathrate hydrate at 200 K is the same as that of gas phase chlorine, while that of chlorine in aqueous solution is  shifted  500  cm-1  to  higher  energy.  When  the chlorine hydrate sample temperature  was lowered to 77 K, there was a 500 cm-1 shift of the spectrum so that it was  much closer to that of the aqueous solution in position but retained the narrow width of  the  200  K  spectrum.  This  effect  was  repro-ducible  and  observed  with  several  different samples. As will be discussed below, the width of a spectrum is an important clue to the environment of  the chromophore. In  this case, the  fact that the Figure 2 Spectra of a) chlorine, b) bromine, and c) iodine in various environments. In each case the green curve is the gas phase spectrum and the red curve is for an aqueous solution. For chlorine, the blue curve was recorded as described above, and is mainly due to chlorine in 51262 cages at 200 K, the gray curve was recorded at 77K. For bromine, the blue curve is for the  TS-I  single  crystal  sample,  the  magenta  curve  is  for  the CS-II crystal, and the dark blue dashed line is the previously recorded spectrum for bromine in 51264 cages of THF clathrate hydrate. For iodine the blue curve is the previously reported spectrum  for  iodine  substituted  into  the  51264  cages  of  THF clathrate hydrate.                      _____________________________________________  width of the 77 K hydrate spectrum is the same as that of the 200 K sample indicates that the chlorine in each sample is in a clathrate hydrate cage.   The  spectra  for  the  two  bromine  hydrate  single crystals show distinct shifts from that of gas phase bromine. The peak of the CS-II spectrum is shifted 440 cm-1 to the blue of the peak of the gas phase spectrum,  and  the  peak  of  the  TS-I  spectrum  is shifted  880  cm-1  to  the  blue  of  the  peak  of  gas phase  spectrum.  These  spectra  are  similar  to,  but measurably different than those  of  the previously reported spectra, as  will be discussed below.  We note  that TS-I and CS-II  single  crystals were  the same thickness, and that the optical density of the TS-I  crystal  was  nearly  twice  that  of  the  CS-II    5 crystal. This reflects the relative stoichiometry of the  two  crystals,  1:8.6  vs.  1:17,  and  further supports  the  assignment  of  the  second  crystal  to the CS-II structure.   For bromine in aqueous solution, the shift from the gas  phase  spectrum  is  1750  cm-1.  The  bromine spectra  are  considerably  more  sensitive  to  the specific  local  environment  than  are  the  chlorine spectra.  The  previously  reported  shifts  for  iodine are  even more dramatic.  The hydrate spectrum is shifted  1440  cm-1  from  the  gas  phase  and  the aqueous  spectrum  is  shifted  2820  cm-1  from  the gas phase.   Also,  notice  that  halogens  in  aqueous  solution exhibit strong UV charge transfer transitions to the blue of the valence bands (mostly off the scale of Fig.  2).  In  addition,  X3-  (X=Cl,  Br,  I)  bands  can also  be  observed  at  sufficient  halogen concentration  midway  between  the  valence  and charge  transfer  bands.    Caging  the  halogen molecule  in  a  clathrate  hydrate  eliminated  the ionic  bands  and  significantly  reduces  the contribution of the charge transfer bands.  DISCUSSION  In  the  section  above,  several  new  results  were presented.  For  chlorine  clathrate  hydrate  the  first UV-visible  spectra  were  reported  for polycrystalline  samples,  and  found  to  have  a remarkable temperature dependence. For bromine clathrate  hydrate  spectra  are  reported  for  single crystals  of  both  the  TS-I  and  the  CS-II  hydrate structures. Although the existence of a stable CS-II crystal  structure  was  reported  previously  on  the basis  of  Raman  spectroscopy,[12]  the  new  data reported  here  provides  an  unambiguous,  inde-pendent  confirmation  of  this  result.  We  will discuss these new results in the context of previous observations and the spectra of halogen molecules in other environments. Results for the shifts of the UV-visible spectra of halogens in various environ-ments are summarized in  Table  I.  Results for the halogen stretching frequencies are summarized in Table II.  Perturbations of halogen by nearby molecules We start with a general discussion of how halogen molecules  interact  with  other  species.  A particularly illuminating example is the interaction of  bromine  with  a  single  argon  atom.  Most chemists  who  do  not  already  know  the  result would  be  surprised  to  learn  that  there  are  two isomers of the Ar-Br2 van der Waals molecule. In T-Ar-Br2 the argon atom attaches to the side of the bromine molecule and interacts equally with each bromine atom via van der Waals forces.[16] In L-Ar-Br2 the  argon atom  attaches to the  end of  the bromine molecule.[17] The linear isomer is stable because  the  bromine  LUMO,  a  ?*  orbital,  is located  on  the  end  of  the  bromine  molecule  and creates a slight ?dimple? in the electron density so that the argon atom can get closer to the bromine atom  in  this  configuration.  The  net  result  is  that the bond energies of the two isomers are the nearly equal,  and  that  the  isomerization  barrier  between them is surprisingly high.  In addition to having different geometries, the two Ar-Br2  isomers  have  very  different  UV-visible spectra.  The  spectrum  of  the  T-Ar-Br2  isomer  is sharp  and  very  slightly  shifted  from  that  of  the bromine molecule. This is because the argon atom on the side of the bromine molecule mostly acts as a  spectator,  hardly  interacting  with  the  bromine orbitals.  The  spectrum  of  the  L-Ar-Br2  isomer  is much  broader  and  is  significantly  blue  shifted from  that  of  the  free  bromine  molecule.  This  is because the argon atom on the end of the bromine molecule  interacts  strongly  with  the  bromine LUMO  orbital,  stabilizing  the  ground  state  and destabilizing the excited state.   Recent calculations show that the analogous result for  the  H2O-Br2  dimer  molecule  should  be  even more  dramatic.[18]  The  lone  electron  pair  of  the water  oxygen  atom  interacts  strongly  with  the bromine  LUMO  orbital  so  that  a  large  blue  shift and broadening is predicted by the calculations.    From  the  above  discussion,  we  note  two  general observations. To the extent that a second molecule is  allowed  to  interact  directly  with  a  halogen LUMO orbital, the valence excitation spectra will tend to be strongly blue shifted and broadened. To the extent that the interacting molecule is a good electron  donor,  the  shift  and  broadening  will  be larger.  Bonding  with  water  in  the  linear configuration  also  results  in  a  considerable lowering of the halogen vibrational frequency.[18]  These  effects  are  apparent  in  the  solution  phase spectra  summarized  in  Table  I.  For  instance,  the spectrum  of  bromine  in  cyclohexane  is  not  blue shifted relative to the gas phase since cyclohexane     6 Table I.  Band maxima of the valence electronic bands of halogens in different environments compared                 to the gas phase. Environment  Chlorine a  Bromine b  Iodine c   ?maxd  (cm-1) ??maxe (cm-1) ?max(C-X ) (cm-1) ??max (cm-1) ?max(B-X) (cm-1) ??max (cm-1) ?max (cm-1) ??max (cm-1) Gas phase  30300  0  24270  0  20830  0  18870  0 C6H12 solution  -  -  24160  -120  20760  -70  19120  250 CCl4 solution  30180  -120  24390  120  20920  90  19360  490 CH2Cl2 solution  30830  530  24720  450  21300  470  19800  930 51262 cage  30300  0  25150  880  21720  890     51264 cage (doped  THF hydrate)  -  -  24630  360  21190  360  20310  1440 51264  cage (single crystal)   -  -  24710  440  21270  440  -  - aqueous solution (T=293 K)  30850  550  26000  1730  22590  1760  21690  2820 amorphous ice (T=120 K)  -  -  25980  1710  22670  1640  21870  3000 a. The chlorine spectrum is dominated by the X?C transition. b. The bromine spectrum has been deconvoluted into the X?C and X?B transitions. See ref [3] for details. c. The iodine spectrum is dominated by the X?B transition. See ref [4] for details. d. ?max is the position of the absorption maximum, ? 50 cm-1. e. ??max is the shift of the absorption maximum from the gas phase value.   Table II. Vibrational frequencies and anharmonicity constants for Chorine, Bromine and Iodine in clathrate hydrate cages and in aqueous solution. Environment  Chlorine  Bromine  Iodine   ?e b  ?exe c  ?e b  ?exe c  ?e d  ?exe Gas phase  559.7  2.67  323.3  1.06  214.5  0.61 51262 cage  -  -  321.2  0.82  -   51264 cage  -  -  317.5  0.7  214  ?0.61 Aqueous solution  538 a  -  306  -  -  - a.  Ref. [19] b. ?e  is the harmonic frequency constant, ? 0.1 cm-1, 35Cl2 and 78,81Br2. c. ?exe  is the anharmonicity constant, ? 0.05 cm-1 d. ? 1 cm-1.    7 has no unpaired electrons available to interact with the  bromine  LUMO  orbital.  Instead,  the  shift  is  slightly  toward  the  red,  most  likely  due  to dielectric  effects.  The  blue  shift  becomes gradually  larger  going  to  CCl4,  then  to  CH2Cl2 then  to  aqueous  solution:  the  order  of  electron donating  propensity.  Similar  shifts  are  observed for  iodine  solutions.  The  shifts  for  chlorine  are analogous,  but  substantially  less.  Also,  note  that the  shifts  for  halogen  molecules  frozen  into amorphous ice are very similar to those in aqueous solution. Finally, although not reported in Table I, the  spectra  in  both  aqueous  solution  and amorphous ice are considerably broader than those in the gas phase. For more detail, see refs. 3 and 4.  The  information  above  provides  a  context  for discussing the spectra of halogens in the clathrate hydrate cages. First, we note that the widths of the halogen clathrate spectra are very similar to those of the gas phase spectra and considerably less than those of aqueous solutions. From this we infer that the  halogen  molecules  in  the  cages  are  not interacting strongly with the lone electron pairs on the surrounding water molecules. To some extent this  is  not  too  surprising  since  the  lone  electron pairs  are  all  involved  in  hydrogen  bonding. However,  the  blue  shifts  for  the  bromine  and iodine  clathrate  spectra  are  still  substantial, especially  when the  fit into the  cage  is  relatively tight.  Chlorine clathrate hydrate spectra Since  the  chlorine  hydrate  samples  studied  here were formed at less than 0.5 atm chlorine pressure, most of the chlorine was located in the larger 51262 cages of  the  CS-I  crystal structure.[8]  At  200  K, the  spectrum  of  the  chlorine  clathrate  hydrate  is found  to  be  very  similar  to  that  of  gas  phase chlorine, indicating that the guest host interaction is  quite  weak  at  this  temperature.  The  narrow width  is  further  evidence  that  the  spectrum  we observe is due to a single cage type. To the extent that  the  chlorine  is  in  the  center  of  the  cage,  on average,  the  chlorine  atoms  will  be  3.45  ?  from the surrounding oxygen atoms. This is far enough that  even  the  chlorine  excited  state  may  not  be perturbed by the cage walls. At a high temperature , 200 K, this assumption might be valid due to the thermal motion, which randomizes the position of the chlorine molecule within the cage. So there is no  direct  interaction  between  the  cage  walls  and the chlorine LUMO orbital.  When  the  spectrum  is  recorded  for  chlorine hydrate  at  77  K,  it  shifts  nearly  as  much  as  for aqueous  chlorine  solutions,  but  is  narrower  than the  aqueous  spectrum.  The  narrow  width  of  the spectrum  indicates  that  chlorine  is  still  in  the clathrate cage. We speculate that the blue shift at 77 K is due to the chlorine molecules settling into a  fixed  position  near  the  cage  walls  so  that  the walls  perturb  the  excited  state.  Careful measurements for single  crystals as a function of composition  and  temperature  would  provide extremely  valuable  data  for  to  aid  the interpretation of the spectra for chlorine hydrate.   Schofield  and  Jordan  performed  calculations  of the  chlorine  ground and valance  excited states in both the 512 and 51262 cages.[20] The calculations predict a spectrum shift of 700 cm-1 for a distorted 512 cage. Few details regarding the calculation for the 51262 cage were presented and they did not test whether  the  observed  spectrum  should  change with the location of the  chlorine molecule  within the cage.  Bromine clathrate hydrate spectra Next, we shift to the interpretation of the bromine clathrate hydrate spectra. Here  we  report the first spectra  measured  for  halogen  clathrate  single crystals. The TS-I crystal spectrum is similar, but slightly  different  from  the  previously  reported bromine clathrate film spectrum; the shift from the gas phase  is 880 cm-1. As discussed below, from the  differences  between  the  two  spectra  we conclude  that  the  previously  studied  bromine hydrate film did not consist of TS-I micro-crystals but rather CS-I micro-crystals.    Similarly,  the  spectrum  reported  for  the  CS-II single  crystals  is  very  similar  to  that  reported previously for bromine in substitutional sites in a THF-CS-II sample. Each spectrum exhibits a 400 ?  40  cm-1  shift  from  the  gas  phase.  This  proves unambiguously that the bromine molecules in the CS-II  single  crystal  are  in  51264  cages.  These crystals  were  previous  attributed  to  the  CS-II structure  on  the  basis  of  their  morphology  and their  Raman  spectra.[12]  The  new  data  reported here confirms that assignment.  The difference between the new spectra for single crystal  TS-I  and  that  of  the  previously  reported samples  that  were  thought  to  be  TS-I  poly-crystalline films are shown in Fig. 3. Although the    8 differences  between  the  two  spectra  are  not dramatic, this is because the width of the figure is 8000 cm-1. In each case, the peak of the spectrum is  shifted  880  cm-1  compared  to  that  of  the  gas phase  molecule.  Upon  close  inspection  the differences  between  the  two  spectra  are  both important and instructive.    Figure  3  Spectra  of  the  TS-I  single  crystal  (3.a)  and  the previously  reported  bromine  hydrate  (3.b).  As  discussed  in the  text,  we  conclude  that  the  TS-I  single  crystal  spectra include contributions from bromine in 51263 cages, while the previously reported sample does not. We further conclude that the  previously  reported  sample  was  a  CS-I  microcrystalline bromine hydrate film.  ________________________________________  The single crystal spectrum, Fig. 3.a, shows a less pronounced shoulder and slightly higher intensity on the low frequency side of the peak compared to the polycrystalline sample. We attribute this extra intensity  to  the  contribution  of  bromine  in  the 51263  cages  in  the  single  crystal  spectrum.  If  so, why  didn?t  the  51263  cages  contribute  to  the previously  observed  spectra?  The  observations reported here, as well as other observations to be reported  in  a  future  paper,  convince  us  that  the previously reported poly-crystalline film consisted of  CS-I  microcrystals  rather  than  TS-I microcrystals.  Thus  in  the  previous  study  only 51262  cages  were  occupied.  Although  the  CS-I bromine  hydrate  crystal  structure  is thermodynamically  unstable  with  respect  to  the TS-I  structure,  the  CS-I  structure  is  kinetically favored since its structure is much easier to grow.  Note  that  the  blue  shift  for  bromine  in  the  51262 cage  is  880  cm-1,  while  that  in  the  51264  cage  is only  400  cm-1.  Clearly,  the  blue  shift  is  very sensitive to cage size. It would be very surprising if bromine in the 51262 and the 51263 cages had the same  spectrum  since  the  cage  sizes  are  quite different. The longest ?free diameter? of the 51262 cage is 5.9 ?, while that of the 51263 cage is 6.59 ?,  close  to  the  value  for  the  51264  cage,  6.56  ?. The  fact  that  the  previous  microcrystalline  film spectrum  shows  a  single,  narrow  peak  is  strong evidence  that  only  a  single  type  of  cage  is occupied.  To  follow  up  the  reasoning  in  the  above paragraph, we perform the following simulation of the TS-I single crystal spectrum. First, we assume that the spectrum of bromine in any given cage has a  cage  independent  shape,  but  a  cage  dependent shift. Next, we assume that the TS-I single crystal spectrum  (Fig.  3.a)  contains  contributions  from both the 51263 and the 51262 cages in a 1:4 optical density  ratio  while  the  polycrystalline  spectrum (Fig.  3.b)  is  due  only  to  bromine  in  51262  cages. This  assumption  has  built  into  it  a  second assumption  that  the  oscillator  strength  does  not vary  between  cages.  This  is  consistent  with  the intensities of the TS-I and the CS-II spectra.   Figure 4 The blue curve is the TS-I single crystal spectrum. The  dashed  black  curve  is  the  CS-I  spectrum  previously obtained  from  a  polycrystalline  film  and  dominated  by bromine  in  51262  cages.  Its  peak  intensity  is  adjusted  to  be 80%  of  the  TS-I  spectrum.  The  dashed  green  curve  has  the same shape as the black curve, but is shifted to the red by 700 cm-1, and adjusted to be 20% as intense as CS-I spectrum. The sum of the black and green curves yields the red curve, which closely fits the observed single crystal spectrum except in the UV wing. The purple curve is the residual of the fit.  ________________________________________  Following the above assumptions, the contribution of  the  51263  cages  to  the  TS-I  spectrum  has  the    9 same  shape  as  the  CS-I  spectrum,  but  shifted  to the red and only 20% as intense. The contribution of the 51262 cages is the same as the CS-I spectrum but with the optical density reduced by 20%. This simulation is shown in Fig. 4. The blue line is the TS-I spectrum.  The dashed black line is the CS-I spectrum with the intensity reduced by 20%. The green  dashed  line  is  the  same  spectrum,  shifted ~700 cm-1 to the  red  and adjusted to 20% of  the peak  optical density.  The  sum of  the  two  dashed lines yields the red line, which is then compared to the  TS-I  spectrum,  the  blue  line.  The  residual between  the  two-component  model  and  the  TS-I spectrum is shown as a purple line at the bottom of the figure.   The two-component model yields an excellent fit to  the  portion of  the  TS-I  spectrum to  the  red of the peak. There is a slight residual to the blue of the peak that we attribute to surface effects. Since bromine  in  the  51263  cages  contribute,  at  most, about  20%  of  the  total  intensity,  the  analysis presented  above  cannot  be  taken  to  be quantitative.  We  are  confident,  however,  in  the conclusion  that  the  polycrystalline  sample consisted of CS-I microcrystals.[14]   Comparison of the three halogens:  Table I. summarizes the halogen clathrate spectral shifts  from  the  gas  phase  observed  in  this  and previous  studies.  Qualitatively,  the  shift  of chlorine in 51262  cages is negligible, except at 70 K; that of bromine in 51264 cages is small, ? 400 cm-1; that of bromine in 51262 cages is large, ? 900 cm-1;  and  that  of  iodine  in  51264  cages  is  very large,  ?  1400  cm-1.  For  comparison,  the  spectral shifts between the gas phase and aqueous solution are 500 cm-1 for chlorine, 1750 cm-1 for bromine and 3000 cm-1 for iodine. It is clear that the spectra of halogens are very sensitive to the proximity of water  molecules  and,  once  analyzed  with  a sufficiently  detailed  model,  will  reveal considerable  insight  into the  details  of  the  guest-host interaction in halogen clathrates.  Although the  valence  transition shifts  in aqueous solutions  are  rather  large,  recent  calculations suggest  that  those  of  isolated  H2O-X2  (X  = halogen) dimers would be even larger.[18]  In the isolated  dimers,  a  rather  strong  bond  is  formed between the two molecules: 982 cm-1 for H2O-Cl2 and  1273  cm-1  for  H2O-Br2.  This  is  due  to  the propensity of the oxygen atom on water to donate its  non-bonding  electrons  to  the  ?*  antibonding orbital of the halogen. Thus the ground state of the H2O-X2 dimer is considerably lower in energy that that  of  the  free  molecules.  Valence  electron excitation  of  the  halogen  promotes  an  electron from the ?* HOMO, to the ?* LUMO. However, in  the  H2O-X2  dimer,  the  oxygen  lone  pair electrons are already occupying some of the space of the LUMO and the valence excited states of the dimer  at  the  geometry  of  the  ground  state  are shifted  to  higher  energy.  The  lowering  of  the ground state energy and the raising of the excited state  energy  account  for  the  large  predicted  blue shift of the spectrum: 1600 cm-1 for H2O-Cl2 and 2000 cm-1 for H2O-Br2.  Why is the blue shift for an aqueous solution less than  for  an  isolated  H2O-X2  dimer?  There  are probably  two  important  effects.  First,  the  water-water  hydrogen  bonding  is  stronger  than  the water-halogen bonding and will serve to displace the water molecules away from the ideal geometry for  bonding with the  halogen molecules. Second, the  other  water  molecules  in  the  liquid environment  may  serve  to  stabilize  the  halogen excited state via dielectric effects.   These caveats are especially true for chlorine, for which  the  electron-accepting  propensity  is considerably weaker than for bromine.  Why are the shifts for the halogen clathrates less than  those  for  aqueous  solution?  To  form  a clathrate  crystal,  all  of  the  oxygen  atom  lone electron  pairs  must  participate  in  hydrogen bonding.  Thus,  to  a  first  approximation,  the HOMO-LUMO  effect  is  not  expected  to  be important in the clathrate environment. This is as observed for chlorine in 51262 cages and bromine in 51263 and 51264 cages. However, for bromine in 51262 cages and iodine in 51264 cages the observed shifts are  still substantial. Although this data  has yet to be interpreted quantitatively, it seems clear that close proximity to the cage walls will raise the energy  of  the  halogen  molecule  valence  excited states.  Resonance Raman spectroscopy Another set of data that helps reveal the host guest interactions  for  halogen  hydrates  is  Raman spectroscopy,  the  results  for  which  are  given  in Table  II.  To  date,  we  have  collected  stretching    10 vibration frequencies and first anharmonicities for bromine in TS-I and CS-II single crystals, 321.2 ? 0.1  cm-1,  ?exe  =  0.82  ?  0.05  cm-1  in  the  former, 317.5 ? 0.1 cm-1 ?exe = 0.7 ? 0.05 cm-1 in the later using 532 nm excitation laser source resonant with the  X?B  transition.[12]  These  values  can  be compared  to  323.3    cm-1  ?exe  =  1.06  cm-1  for bromine in the gas phase and 306 cm-1 in aqueous solution. Unlike the UV-visible spectral shifts, the vibrational frequencies  are  strictly a  ground state property.  In  this  case,  the  larger  51264  cage produces the larger spectral shift. This is in accord with  the  trend  first  noticed  by  Fleyfel  and Devlin:[21] molecules in larger hydrate cages tend to have lower vibrational frequencies.   As  for  bromine  in  the  51262  cage,  the  stretching vibrational frequency for iodine in the 51264 cage of THF clathrate is close to the gas phase value. In contrast  to  the  electronic  spectra,  a  tight  fitting cage  has  a  smaller  effect  on  the  vibrational frequency than a larger cage. At a first glance this is a counterintuitive result. However, the observed trends  have  a  simple  explanation  when  we consider  the  effect  of  polarization  of  the  guest molecule by the net electric field of the cage water molecules.  The  ground  electronic  state  of  the polarized  molecule  would  have  an  admixture  of the  ion-pair  states,  which  have  much  lower vibrational  frequencies.[22]  In  a  tight  cage  the guest  is  closer  to  the  center,  where  it  feels  a smaller electric field so that and polarization of the halogen will be smaller. In contrast, when the cage is loose, the guest may fall on a side where electric field  causes  stronger  polarization  and  thus  the vibrational frequency of the guest molecule drops.   Note that the anharmonicity constant for bromine in  the  51264  cage  is  significantly  smaller  in  the clathrate cages than in the gas phase. This implies that the cage increases the potential energy as the halogen stretches away from its equilibrium bond length. This indicates again that the halogen in the large cage might move to one side and be in closer contact  to  the  cage  wall.  However,  until  a quantitative model is developed this remains quite speculative.  The combination of the  UV-visible spectrum and the  Raman  results  for  liquid  water  may  be especially  interesting  to  simulate.  Intermolecular bonding, polarization  and  thermal  effects  will all be  quite  important  Calculations  on  the  isolated H2O-Br2 dimer predict that the bromine stretching frequency  is  even  lower  in  the  dimer  than  in aqueous solution.[18]  The  resonance  Raman  spectra  of  enclathrated bromine  are  different  from  those  of  an  aqueous solution in another important respect. Those of the enclathrated bromine exhibit a long progression of vibrational  overtones.  In  aqueous  solution,  only the  fundamental  is  observed.  This  difference indicates  that  electronic  dephasing  time,  which determines  formation  of  the  resonance  Raman progression,  is  extremely  short  in  aqueous solution,  <5-10fs.[23]  Upon  electronic  excitation of  a  halogen  molecule,  vibrational  modes  of  the solvent  are  directly  excited  (this  is  evident  in  a broadening of the molecular absorption spectrum), leading to a very fast electronic dephasing in the halogen  excited  state  and  shortening  of  the resonance Raman progression.   Pictorial views halogens in the clathrate cages Although  development  of  a  quantitative  analysis for  the  data  presented  here  is  yet  to  come,  it  is instructive to visualize the cage environments for the  halogens  in  the  several  cages.  To  create  the following  figures,  the  cages  were  constructed  by taking  the  oxygen  coordinates  from  X-ray diffraction data  and assigning hydrogen positions as  described  in  ref  [4].  Bromine  molecules  were given the  correct bond length and their positions with  the  cage  were  calculated  using  the  MMFF model[24]  in  the  Spartan  program.[25]  Although the MMFF potential is not expected to be accurate enough  for  detailed  predictions,  it  is  useful  for illustrating the relative sizes of the cages and the relative fits of the halogen molecules.  Figure 5 shows two views of bromine relative to a slice  through  the  51262  cage.  The  bromine  is aligned  parallel  to  the  hexagonal  faces,  roughly centered in the cage. The end view, relative to the bromine bond, Fig. 5a, shows that the bromine is in  closer  contact  with  the  top  and  bottom  of  the cage  than  with  the  sides  perpendicular  to  the bromine  bond.  The  side  view  shown  in  Fig.  5b shows  that  ends  of  the  bromine  molecule  are  in close contact with the cage walls, but do not align directly with any of the oxygen atoms.   Figure 6 shows analogous views of bromine in the 51263 cage.  For the 51263 cage the hexagonal faces    11 form  a  ring around  the  bromine  bond, giving the cage  a  prolate  shape.  This  results  in  a  more uniform spacing between the bromine and the cage wall.  In  particular,  the  ends  of  the  bromine molecule are well spaced from the cage walls. The special stability of bromine in this cage, resulting in  the  unique  TS-I  crystal  structure,  is  probably due  to the fact that  the  prolate  shape  of  the  cage allows the van der Waals interaction between the water and bromine molecules to be optimized for a large fraction of the cage walls.   Figure  5.  Bromine  in  the  51262  cage.  The  left  view  (5a)  is along the bromine bond, the right view (5b) is from the side of the bromine bond.      Figure  6.  Bromine  in  the  51263  cage.  The  left  view  (6a)  is along the bromine bond, the right view (6b) is from the side of the bromine bond.     Figure  7.  Bromine  in  the  51264  cage.  The  left  view  (7a)  is along the bromine bond, the right view (7b) is from the side of the bromine bond.             Figure 7 shows views of bromine in the 51264 cage. In this case the spacing between the bromine and the cage walls is quite large except on the ends of the  molecule.  It  would  not  be  surprising  if  the bromine is not located in the center of the cage at low temperature.  Prospects for future work Although a wealth of new data has recently been obtained  for  the  spectroscopy  of  halogen molecules  in  clathrate  hydrate  environments,  the analysis  of  the  data  has  just  begun.  The  first  ab initio study of chlorine in hydrate cages has been performed.[20]  Many  more  results  will  be necessary  before  the  spectra  can  be  interpreted. Water-halogen  model  potentials  currently  in  the literature, such as the MMFF potential used above, do not accurately model the full anisotropy of the intermolecular  forces.  The  success  of  the electrostatic  model  in  interpreting  the  iodine spectrum  suggests  that  perhaps  a  Diatomic-in-Molecule approach might be useful. We hope that the data presented here, along with new data to be collected, will stimulate our theoretical colleagues to investigate these systems in more detail.  Conclusion  The first UV-visible spectra for chlorine clathrate hydrates  are  reported  and  the  first  single  crystal spectra  for  two  crystalline  forms  of  bromine clathrate  hydrate  are  reported.  The  spectrum  for chlorine  clathrate  at 200  K is very similar  to the gas phase spectrum, while the spectrum at 77 K is shifted 500 cm-1 to the blue. We speculate that at 200  K  the  thermally  averaged  position  of  the chlorine is in the center of the cage but that at 77 K the lack of thermal motion allows the  chlorine to  settle  into  a  position  near  the  cage  walls. Clearly,  more  data  will  be  required  to  test  this hypothesis. It will also be valuable to study more highly  doped  chlorine  samples  with  hope  of extracting spectra for chlorine in the 512 cages.  The spectrum of the bromine clathrate CS-II single crystal is quite similar to that previously reported for  bromine  substituted  into  some  of  the  51264 cages  of  CS-II  THF  clathrate.  This  confirms  the previous  assignment  of  the  resonance  Raman spectra  of  this crystal to the  CS-II  structure.  The spectrum  of  the  TS-I  single  crystal  shows  extra intensity on the red side of the previously recorded polycrystalline  spectrum.  We  conclude  that  the    12 previously  studied  films  consisted  of  metastable CS-I microcrystals.  In  general,  the  more  tightly  a  halogen  molecule fits into a clathrate cage, the more blue shifted its UV-visible  absorption  spectrum.  In  contrast,  the resonance  Raman  spectra  reveal  that  the  halogen vibrational  frequencies  are  shifted  less  from  the gas  phase  in  the  tight  fitting  cavities,  while  the frequencies  are  more  red  shifted  in  the  larger cages. These results will provide a rigorous test for quantitative models, which are under development to  better  understand  the  details  of  the  halogen-clathrate interactions.                                                         REFERENCES [1]  Davy  H.  The  Bakerian  Lecture:  On  Some  of  the Combinations  of  Oxymuriatic  gas  and  Oxygen,  and  on  the Chemical  relations  of  These  Principles,  to  Inflammable Bodies.  Philosophical  Transactions  of  the  Royal  Society 1811;101:1-35. [2] Lowig C. Uber eine Bromverbindungen und uber Brom-Darstellung. Ann. Chim. Phys. Ser. 1829;42(2):113-119. [3]  Kerenskaya  G,  Goldschleger IU,  Apkarian VA, Janda KC.  Spectroscopic  signatures  of  halogens  in  clathrate hydrate cages. 1. Bromine. Journal of Physical Chemistry A 2006;110:13792-13798. [4] Kerenskaya G, Goldschleger GI, Apkarian VA, Fleisher E,  Janda  KC.  Spectroscopic  Signatures  of  Halogens  in Cathrate  Hydrate  Cages.2.  Iodine.  Journal  of  Physical Chemistry A 2007;111(43):10969-10976. [5] Anthonsen JW. The  Raman  Spectra of Some Halogen  Gas Hydrates. Acta Chemica Scandinavia A 1975; 29:175. [6]  Davy  H.  On  the  Fallacy  of  the  Experiments  in  Which Water is Said to Have Been formed by the Decomposition of Chlorine.  Philosophical  Transactions  of  the  Royal  Society 1818;108:169.  [7]  Pauling  L,  Marsh  RE.  Structure  of  the  Clorine  Hydrate. Proc. Natl. Acad. Sci. USA 1952;38:112-117. [8]  Cady  GH.  Composition  of  Cathrate  Gas  Hydrates  of CHCIF2,  CCI3F,  CI2,  CIO3F,  H2S,  and  SF6.  Journal  of Physical Chemistry 1981;85:3225-3230. [9]  Allen  KW,  Jeffrey  GA.  On  the  structure  of  bromine hydrate. Journal of Chemical Physics 1963;38:2304. [10]Dyadin  YA,  Belosludov  VR.  Thermal  expansion  and lattice  distortion  of  clathrate  hydrates  of  cubic structures  I  and  II.  Comprehensive  Supramolecular Chemistry. 1996;6:789. [11]Udachin  KA,  Enright  GD,  Ratcliffe  CI,  Ripmeester  JA. Polymorphism  in  Br2  Cathrate  Hydrates.  Journal  of  the American Chemical Society 1997;119:11481. [12]  Goldschleger  IU,  Kerenskaya  G,  Janda  KC,  Apkarian VA.  Polymorphism  in  Br2  Clathrate  Hydrate  Journal  of Physical Chemistry A 2008;112(5):787-789. [13]  Platteeuw  J.  C.  van  der  Waals  J.  H.  Thermodynamic Properties  of  gas  hydrates,  Advances  in  Chemical  Physics 1957;2:1 [14]  Kerenskaya  G,  Goldschleger  I.  U.,  Apkarian  V.A., Janda K.C., Growth and morphology of bromine hydrates, In preparation.                                                                                        [15] Zimmerman G, Strong FC. Equilibria and Spectra of Aqueous  Chlorine  Solutions.  Journal  of  the  American Chemical Society 1957;79:2063-2066. [16] Cabrera J, Bieler CR, McKinney N, van der Veer  WE, Pio  J,  Roncero  O,  Janda  KC,  Time  and  frequency  resolved dynamics  of  ArBr2.  Journal  of  Chemical  Physics 2007;127:164309.  [17] Pio J, van der Veer WE, Bieler CR, Janda KC. Product state resolved excitation spectroscopy of He-, Ne- and Ar-Br2 linear isomers: Experiment and theory.  Journal of Chemical Physics 2008;128:134311. [18]  Hernandez-Lamoneda  R,  Uc-Rosas  VH,  Bernal-Uruchurtu  MI,  Halberstadt  N,  Janda  KC.  Two  Dimensional H2O-Cl2 and H2O-Br2 potential surfaces: an ab initio study of ground  and  valence  excited  electronic  states.  Journal  of Physical Chemistry A 2008;112(1):89-96. [19]  Cherney  DP,  Duirk  SE,  Tarr  JC,  Collette  TW. Monitoring  the  Speciation  of  Aqueous  Free  Chlorine from pH 1 to 12 with Raman Spectroscopy to Determine the  Identity  of  the  Potent  Low-pH  Oxidant.  Applied Spectroscopy 2006;60(7):764-772. [20]  Schofield  DP,  Jordan  KD.  Theoretical  investigation  of the  electronically  excited  states  of  chlorine  hydrate.  Journal of Physical Chemistry A 2007;111(32):7690-7694.  [21]  Fleyfel  F,  Devlin  JP.  FTIR  spectra  of  90  K  films  of simple,  mixed  and  double  clathrate  hydrates.  Journal  of Physical Chemistry 1988;92:631. [22]  Senekerimyan  V,  Goldschleger  I,  and  Apkarian  VA, Vibronic dynamics of I2 trapped in amorphous ice: Coherent following  of  cage  relaxation.  Journal  of  Chemical  Physics 2007;127:214511.  [23] Ovchinnikov M, Apkarian VA, Voth GA.  Semiclassi-cal molecular dynamics computation of spontaneous light emission in the condensed phase: Resonance Raman spectra.  Journal of Chemical Physics 2001;114,7130-7143. [24]  Hehre  WJ.  A  Guide  to  Molecular  Mechanics  and Quantum Chemical Calculations. Wavefunction, 2001, Irvine, CA. [25] Spartan ?04, Wavefunction Inc., Irvine, CA. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
United States 1 0
City Views Downloads
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items