Motional modes in bulk powder and few-molecule clusters of tris(8-hydroxyquinoline aluminum) and their relation to spin dephasing

Please download to get full document.

View again

of 3
0 views
PDF
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Motional modes in bulk powder and few-molecule clusters of tris(8-hydroxyquinoline aluminum) and their relation to spin dephasing
Document Share
Document Tags
Document Transcript
  Motional modes in bulk powder and few-molecule clusters of tris „ 8-hydroxyquinoline aluminum …  and their relation to spin dephasing Lopamudra Das, 1 Jennette Mateo, 1 Saumil Bandyopadhyay, 1 Supriyo Bandyopadhyay, 1,a  Jarrod D. Edwards, 2 and John Anderson 2 1  Department of Electrical and Computer Engineering, Virginia Commonwealth University, Richmond,Virginia 23284, USA 2 US Army Engineer Research and Development Center, Alexandria, Virginia 22315, USA  Received 18 August 2010; accepted 24 January 2011; published online 10 February 2011  The ensemble averaged spin dephasing rate of localized electrons in the organic moleculetris  8-hydroxyquinoline aluminum   or Alq 3  has been found to be significantly larger in bulk powderthan in single- or few-molecule clusters confined within 1–2 nm sized nanocavities   B. Kanchibotla et al. , Phys. Rev. B  78 , 193306   2008  . To understand this observation, we have compared themidinfrared absorption spectra of bulk powder and single- or few-molecule clusters. It appears thatmolecules have additional vibrational modes in bulk powder possibly due to multimerization. Theircoupling with spin may be responsible for the increased spin dephasing rate in bulk powder. ©  2011 American Institute of Physics .   doi:10.1063/1.3554753  There is significant interest in understanding the dephas-ing mechanism of a single spin in an organic molecule 1–5 since the latter can be an excellent host for spin-based qubits.Quantum gates fashioned out of spins in organic moleculespossess remarkable gate fidelity 6 and there exists a uniqueand elegant scheme for qubit readout that has no direct ana-log in inorganic systems. 3 Recently, two interesting results were reported. The en-semble averaged spin dephasing time   T  2    of electrons boundto defect or impurity sites in the organic molecule tris-  8-hydroxyquinoline aluminum   or Alq 3   C 27 H 18 N 3 O 3 Al   wasfound to   1   be   2.5 times shorter in bulk powder than insingle- or few-molecule clusters confined in 1–2 nm sizednanocavities and   2   decrease rapidly with increasing tem-perature in the range of 4.2–100 K, indicating that dephasingis mediated by interaction with time-dependent perturbationssuch as phonons, molecular vibrations, etc. 3 There are many possible reasons why  T  2  in bulk powderand in few-molecule clusters confined in nanocavities couldbe different. Kanchibotla  et al. 3 hypothesized that a nanocav-ity may confine environmental acoustic phonons and dis-cretize their energies. 7 In that case, spin-phonon interaction,which might be a major route for dephasing, will be sup-pressed in a nanocavity if phonons whose energies are reso-nant with spin splitting energies are not allowed within thecavity owing to the specific boundary conditions. One couldthink of this as a nontraditional phonon bottleneck effect,somewhat different from the traditional one. 8 Such an effectwill increase the spin dephasing time in a nanocavity. A sec-ond possibility is that the effect has nothing to do with thenanocavity but is entirely due to intermolecular interactions.Molecules in bulk powder can form multimers because of strong intermolecular interactions and experience enhancedelectron-phonon interaction as a result. 9 In one- or two-molecule clusters, they can only form monomers or dimersand the electron-phonon interaction is weaker. 9 This will in-crease the spin dephasing time in single- or few-moleculeclusters over that in bulk powders. A third possibility is thatformation of multimers in bulk powder could introduce ad-ditional rotational and vibrational modes in the molecules.Coupling of these additional modes with spins may increasethe spin dephasing rate. This is actually not that differentfrom the second possibility since the molecular vibrationsand rotations can be viewed as “phonons.” The only differ-ence is that instead of increasing the strength of electron-phonon coupling   the matrix element  , this mechanism in-creases the number of available phonon modes, which alsoresults in stronger electron-phonon interaction.In order to investigate the third possibility, we preparedone- or two-molecule clusters of Alq 3  within 1–2 nm sizednanocavities. We then measured their midinfrared   mid-IR  absorption spectra in the wave number range of 2000–4000 cm −1  wavelength of 2.5–5    m   and comparedthem to the absorption spectra of bulk powder. The infraredabsorption spectrum of molecules is a signature of the vibra-tional and rotational modes. Our purpose was to see if theabsorption spectra in this frequency range are different inbulk powder and in one- or two-molecule clusters since thatwould indicate a difference in the rotational and vibrationalmodes.In order to fabricate 1–2 nm sized nanocavities that willhost few-molecule clusters, we followed the method in Refs.3 and 10. We first anodized a 99.999% pure aluminum foil of  0.1 mm thickness in 15% sulfuric acid at 10 V dc   at roomtemperature   for 15 min to produce an   1    m thick porousalumina film on the surface of the foil containing pores withdiameter of 10 nm. 11 There are cracks with diameter of 1–2nm in the pore walls 10,12,13 that end in a small nanovoid of 1–2 nm diameter. Reference 3 contains a transmission elec-tron micrograph of such a crack. Formation of these nanofis-sures and the nanovoids has been explored exhaustively bymany groups in the past. 10,12,13 The nanovoids act as the“nanocavities.” Because their linear dimensions are 1–2 nmand the Alq 3  molecule has a size of 0.8 nm, only one or twomolecules can be accommodated within any nanovoid.Therefore, by trapping Alq 3  within these nanovoids andflushing them out from everywhere else, we can synthesizeone- or two-molecule clusters. a  Electronic mail: sbandy@vcu.edu. APPLIED PHYSICS LETTERS  98 , 063109   2011  0003-6951/2011/98  6   /063109/3/$30.00 © 2011 American Institute of Physics 98 , 063109-1  Since we have to measure infrared absorption of theclusters, we need a substrate that is transparent to infrared.Unfortunately, metallic aluminum, on which the porous alu-mina film is formed, absorbs infrared radiation and isopaque. Therefore, the porous alumina film has to be re-leased from the aluminum substrate before filling the nano-cavities with Alq 3 . For this purpose, an organic varnish isbrushed on the porous film and the sample is soaked over-night in HgCl 2  solution which selectively etches aluminum.The varnish provides mechanical stability to the film   whichis only about 1    m thick    during the release phase. Thereleased sample is carefully inspected to ensure that no alu-minum is present and then transferred to a beaker filled withethanol to dissolve out the varnish. The ultrathin porous filmfloats to the surface of ethanol after the heavy varnish dis-solves, and it is captured on a glass slide that is transparent tomidinfrared.The sample is then soaked in 1,2-dichloroethane  C 2 H 4 Cl 2   solution of Alq 3  overnight to impregnate both the1–2 nm cracks and the 10 nm pores with Alq 3  molecules.Huang  et al. 10 claimed that Alq 3  molecules of size 0.8 nmdiffuse through the cracks and come to rest in the nanovoidswhere the cracks terminate. After many hours of soaking inthe Alq 3  solution, the samples are withdrawn and then rinsedrepeatedly in C 2 H 4 Cl 2  to remove any Alq 3  molecule fromwithin the pores and everywhere else but leaving behind themolecules in the nanovoids since C 2 H 4 Cl 2  cannot diffusethrough 1–2 nm cracks and wash out the molecule fromwithin a nanovoid. 10 The C 2 H 4 Cl 2  molecules evaporatequickly in air, leaving behind the one- or two-molecule clus-ters trapped in the nanovoids. This completes sample prepa-ration.It is not possible to ascertain with high resolution mi-croscopy that the clusters are actually present in the nano-voids. Therefore, we carried out room-temperature photolu-minescence   PL   measurements on these samples to confirmthe presence of the clusters. Since Alq 3  is optically active, itluminesces in the wavelength range of 400–600 nm, unlikeits host   porous alumina produced in sulfuric acid  . This al-lows us to sense the presence of the clusters. The PL spec-trum of porous films containing clusters is shown in Fig. 1.In each case, the PL of blank samples   containing no Alq 3  was subtracted from the PL of Alq 3  bearing samples in orderto ascertain that the resulting PL is due to the Alq 3  clustersalone.The PL spectrum has three peaks at 497, 516, and 536nm, corresponding to photon energies of 2.48, 2.39, and 2.30eV. There is also a broad shoulder at 480 nm. The peak at FIG. 1.   Color online   Room temperature photoluminescence spectrum of Alq 3  molecules trapped within 1–2 nm sized nanocavities in alumina. Thepeak at 497 nm is due to direct transition across the HOMO-LUMO gap,while the other peaks are either due to interference of waves reflected fromfront and back of the sample or due to impurities introduced during theprocessing steps.FIG. 2.   Color online   Midinfrared absorption spectrum of    a   one- or two-molecule clusters of Alq 3  hosted in the nanocavities and   b   bulk powder of Alq 3 . 063109-2 Das  et al.  Appl. Phys. Lett.  98 , 063109   2011   497 nm is due to direct transition across the highest occupiedmolecular orbital   HOMO  –lowest unoccupied molecular or-bital   LUMO   gap of Alq 3  and was also observed in Refs. 9and 10 which measured the PL spectrum of Alq 3  moleculestrapped in porous alumina prepared in sulfuric acid. Thebroad shoulder and the other peaks could be due to interfer-ence of light reflected from the front and back of the porousalumina film or could be due to impurity states introducedduring the unique processing steps. Since they have no bear-ing on the subject of this investigation, we did not probethem further. Overall, our PL is very similar to those in Refs.9 and 10. This gives us confidence that the organic molecules are present in the nanovoids and have not been removed byrinsing in C 2 H 4 Cl 2 . Transmission electron microscopy inRef. 3 had ascertained that there are no Alq 3  clumps leftoutside the nanocavities either. Therefore, we can be reason-ably certain that the only places occupied by the Alq 3  mol-ecules are the nanocavities. Furthermore, because of theknown size of the nanocavities, we can be confident thatmost clusters contain one or two molecules and not manymore.Finally, mid-IR absorption spectra of Alq 3  moleculesconfined within nanocavities were measured using Fouriertransform infrared spectroscopy in a Nicolet spectrometer.Since a single sample did not produce enough absorption tobe detected by our equipment with adequate signal-to-noiseratio, we had to stack three samples to measure absorption.In all cases, the absorption spectra were obtained by subtract-ing the spectrum of blank samples   that do not contain anyorganics but are otherwise nominally identical with organicbearing samples   from the latter. The spectrum was alwaysaveraged over a large number of scans to ensure that it wasindependent of the number of scans.The absorption spectrum of Alq 3  clusters in nanocavitiesis shown in Fig. 2  a   in the wave number range of 2200–4000 cm −1 . For comparison, we have plotted the ab-sorption of bulk Alq 3  powder in this range in Fig. 2  b  . Notethat the bulk powder’s absorption increases monotonicallywith wave number, but the absorption of clusters is non-monotonic and shows a very broad peak centered at  3400 cm −1 . In this range, pure Alq 3  should have a peak at3050 cm −1 due to stretching motion of the C v H bond andanother at 3400 cm −1 due to stretching motion of the hy-droxyl bond. 14 They may have merged to form the broadpeak. Clearly, the absorption in clusters decreases rapidly athigh wave numbers above 3600 cm −1 , but the absorption inbulk powder does not. The obvious conclusion is that mo-tional modes that contribute to the high frequency absorptionin powders are  absent   in the clusters. These additional modesthat occur in powders but not in clusters are most likely dueto creation of multimers which form when many moleculesare present within range of intermolecular interactions. Mul-timerization can obviously occur in bulk powders but cannotoccur in clusters since they contain only one or two mol-ecules.The Alq 3  monomer has 52 atoms and 300 degrees of freedom for rotational and vibrational motions. 15 Dimers andmultimers have many more atoms than monomers and cor-respondingly more degrees of freedom giving rise to manymore possible vibrational and rotational modes. Therefore, itis natural to find additional IR absorption in powders wheremultimers can form. It is noted that the additional modescrowd around the high frequency end of the IR spectrum, butits cause is not well understood. Nonetheless, the rotationaland vibrational motions associated with these additionalmodes would couple to electron spin and could be respon-sible for the increased spin dephasing rate in powders.In conclusion, we have shown that the previously re-ported increase in the spin dephasing rate in bulk powdersof Alq 3  compared to one- or two-molecule clusters 3 couldbe due to multimerization in the powder. This does notrule out other possible mechanisms for the increase butcertainly establishes multimerization as one possible cause.This could be an important consideration for all of “organicspintronics” since organic molecules—unlike inorganicsemiconductors—are susceptible to multimerization. 1 S. Pramanik, C.-G. Stefanita, S. Patibandla, S. Bandyopadhyay, K. Garre,N. Harth, and M. Cahay, Nat. Nanotechnol.  2 , 216   2007  . 2 S. Bandyopadhyay, Phys. Rev. B  81 , 153202   2010  . 3 B. Kanchibotla, S. Pramanik, S. Bandyopadhyay, and M. Cahay, Phys.Rev. B  78 , 193306   2008  . 4 P. A. Bobbert, W. Wagemans, F. W. A. van Hoost, B. Koopman, and M.Wohlgenannt, Phys. Rev. Lett.  102 , 156604   2009  . 5 M. N. Grecu, A. Mirea, C. Ghica, M. Colle, and M. Schwoerer, J. Phys.:Condens. Matter  17 , 6271   2005  , and references therein. 6 J. Lehmann, A. Arino-Gaita, E. Coronado, and D. Loss, Nat. Nanotechnol. 2 , 312   2007  . 7 M. A. Stroscio, K. W. Kim, S. G. Yu, and A. Ballato, J. Appl. Phys.  76 ,4670   1994  . 8 H. Benisty, M. Sotomayor-Torres, and C. Weisbuch, Phys. Rev. B  44 ,10945   1991  . 9 C. Xu, Q. Xue, Y. Zhing, Y. Cui, L. Ba, B. Zhao, and N. Gu, Nanotech-nology  13 , 47   2002  . 10 G. S. Huang, X. L. Wu, Y. Xie, F. Kong, Z. Y. Zhang, G. G. Siu, and P. K.Chu, Appl. Phys. Lett.  87 , 151910   2005  , and references therein. 11 S. Bandyopadhyay, A. E. Miller, H.-C. Chang, G. Banerjee, V. Yuzhakov,D.-F. Yue, R. E. Ricker, S. Jones, J. A. Eastman, E. Baugher, and M.Chandrasekhar, Nanotechnology  7 , 360   1996  . 12 D. D. Macdonald, J. Electrochem. Soc.  140 , L27   1993  . 13 S. Ono, H. Ichinose, and N. Masuko, J. Electrochem. Soc.  138 , 3705  1991  . 14 N. B. Colthup, L. H. Daly, and S. E. Wiberly,  Introduction to Infrared and  Raman Spectroscopy , 3rd ed.   Academic, New York, 1990  . 15 T. Gavrilko, R. Federovich, G. Dovbeshko, A. Marchenko, A. Naumovets,V. Nechytaylo, G. Puchokovska, L. Viduta, J. Baran, and H. Ratajczak, J.Mol. Struct.  704 , 163   2004  . 063109-3 Das  et al.  Appl. Phys. Lett.  98 , 063109   2011 
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks