DRIFT and ENDOR Study of the Carotenoid Bixin Attached to

Subscriber access provided by UNIV OF DURHAM

C: Surfaces, Interfaces, Porous Materials, and Catalysis 2

DRIFT and ENDOR Study of the Carotenoid Bixin Attached to Irradiated TiO Sefadzi Tay-Agbozo, Shane Clark Street, and Lowell D Kispert J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b06240 • Publication Date (Web): 26 Jul 2018 Downloaded from http://pubs.acs.org on July 29, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

DRIFT and ENDOR study of the Carotenoid Bixin attached to Irradiated TiO2 Sefadzi Tay-Agbozo, Shane Street, Lowell D. Kispert* Department of Chemistry, The University of Alabama, Box 870336, Tuscaloosa, AL 354870336 USA Email: [email protected], Phone: 205 348-7134

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Diffuse Reflectance Infra-red Fourier Transform (DRIFT) spectroscopic measurement of the carotenoid cis bixin on TiO2, shows adsorbance of bixin by the presence of carboxylate, methine, H2O removal or strongly adsorbed OH, and presence of Hydroxyl at the 1200 – 1800, 2800 – 3000, 3000 – 3600 and 3700 – 3900 cm-1 regions respectively. The EPR/ENDOR spectra of bixin on TiO2 show the formation of bixin radical cation and proton-loss neutral radicals upon light irradiation - similar to the radicals formed and detected by ENDOR for bixin adsorbed on silica alumina. On silica alumina, radicals were generated prior to irradiation, and upon irradiation the spectral signal was reduced by a factor of 4-5. On the other hand, on TiO2, no signal was detected until after irradiation. The nuclear magnetic resonance (NMR) of the purified bixin shows the cis bixin configuration, and the mass spectrometry (MS) analysis in both positive and negative modes, shows no major peak besides peaks consistent with bixin.


Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction When all trans carotenoids (Car) like those given in Figure 1 are adsorbed on silica alumina (SiAl), electron transfer occurs in the absence of light to form ground state (GS) trans radical cations. If the Car has a low oxidation potential (ox. pot.) like for β-carotene (0.6 V) 1 or zeaxanthin (0.6 V),2 with a pKa for the radical cation of 4, it transfers a proton to the matrix and forms a trans proton-loss neutral radical upon light irradiation. For Car with higher ox. pot., like astaxanthin (0.8 V, Figure 1),3 less proton transfer occurs with light irradiation because the trans radical cation is less acidic (pKa is 7-8).

Figure 1: cis structure of bixin and all trans structures of carotenoic acid, astaxanthin, zeaxanthin, canthaxanthin and β-carotene.

However, for bixin (Figure 1) the GS molecular structure is found by density functional calculations (DFT) to be cis.4 The DFT determined GS bixin radical cation is all trans and the 3 ACS Paragon Plus Environment


GS bixin proton-loss neutral radical is 9′-cis bixin.4 On SiAl,4 the cis bixin molecule first transfers an electron to the surface leading to the formation of the GS trans radical cation ( all in the absence of light) and then the trans radical cation is adsorbed. Furthermore, ENDOR analysis4 of the radicals formed by the adsorption of a solution of 80% 9'-cis bixin, 20% trans bixin in the absence of light on SiAl indicated the presence of 23% cis radical cation, 38% trans proton-loss neutral radical and 14% cis proton-loss neutral radical, for a total of 75%, all of which are not GS structures for bixin. These decay upon light irradiation. What remains is the 26% of the trans bixin radical cation which is a GS structure. This loss of radical intensity with light exposure for cis bixin absorbed on SiAl is not observed for any other trans Car which all possess a GS trans radical cation and trans proton-loss neutral radical configuration (examples: β-carotene and zeaxanthin). Silica alumina is known to have the right properties such as pore and particle size for the adsorption of Car. Lewis and Brønsted acid sites (Figure 2) are formed by partially dehydrating the silica alumina surface through calcination. These sites are partially the premise for the industrial use of silica alumina as catalyst,5 and the inherent structure and properties of such sites in metal oxides are an active area in catalysis research.6 Once activated, silica alumina is a convenient surface for the generation and stabilization of Car radicals which can be subsequently characterized spectroscopically.1-2, 7-9


Page 4 of 32

Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2: Brønsted acidic (a) and Lewis acidic (b) sites of silica alumina and (c) hydroxyl surface of TiO2. •+

The formation of Car occurs by electron transfer to the Lewis acidic site. To stabilize •+ •+ Car , the pH of the matrix (Brønsted acidic site) must be less than pKa of the Car . Light

irradiation of the physisorbed Car promotes proton-loss to form proton-loss neutral radical by transfer of proton to the matrix, which acts as a Brønsted base site. What is different about bixin adsorbed on SiAl? It exhibits the highest oxidation potential (ox. pot., ~0.94 V)10 of all measured Car and should exhibit little or no loss of radical intensity upon light similar to that exhibited by astaxanthin. However, it has a variation in GS structure for the molecule, radical cation and proton-loss neutral radical that does not exist for the other Car which have been studied.1-3, 7-9 All other Car studied possess a GS trans structure for the molecule, radical cation and proton-loss neutral radical species. The interest in using bixin in photovoltaics, requires a support matrix where the bixin molecule does not desorb. Carotenoids have been used in the preparation of photocurrent cells by adsorbing β-carotene or canthaxanthin (Figure 1) on TiO2 nanoparticles (Fig. 2c).11 It was later found12 that, the efficiency of the cell increased by more than an order of magnitude by attaching carotenoic acid (8′-apo-β-caroten-8′oic acid in Figure 1) to a TiO2 nanocrystalline mesoporous electrode.



Page 6 of 32

This paper using ENDOR and DRIFT measurements show that attachment of bixin to the hydroxyl surface of TiO2 (Fig. 2c), results in radical formation upon light irradiation which do not decay upon irradiation, an important feature when bixin is used in the construction of photovoltaic cells. Bixin is readily available since it is used industrially as an FDA approved dye and is extracted and purified from annatto seeds. Experimental Bixin Extraction and Purification Bixin was extracted and purified from annatto seeds as reported by Tay-Agbozo et al.4 Briefly, bixin was extracted from annatto seeds using a Soxhlet apparatus, based on the procedure of Silva et al.13 A Bruker 500 MHz instrument was used for the NMR analysis of the bixin with deuterated chloroform (CDCl3). The NMR assignments correspond with known bixin NMR assignments,14-17 while the chemical shifts and couplings show cis bixin configuration. The mass spectrometry (MS) analysis of the bixin was recorded on a Bruker HCTultra Discovery System using electron spray ionization (ESI) in both positive and negative modes. Methanol/chloroform in a 2:1 ratio with 0.1 % trifluoroacetic acid (TFA) and methanol were used in negative and positive modes respectively. The mass spectrum of the purified bixin -

-

molecule shows 393.3 m/z, the [M-H] and 787.5 m/z ([2[M-H]+H] ) - a proton bridging dimer peak- in the negative mode as the only pronounced peaks. The theoretical peak for the negative mode is 393.5 m/z. Peaks at 395.3 and 417.3 m/z in the positive mode of the mass spectra +

+

represent [M+H] and [M+Na] respectively, with respect to the theoretical values of 395.5 m/z and 417.5 m/z, respectively.


Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Adsorption of Bixin on TiO2 Various solutions of bixin in chloroform were prepared (see Table 1). Between 0.13-0.17 g of Degussa P25 TiO2 powder from Acros Organics previously heated at ~ 450 °C (well below phase transition point) and allowed to cool in dry a box were added to the bixin solutions (table 1). Degussa P25 is largely rutile with about 85% rutile:anatase ratio and anatase-rutile transition above 900 °C. It must be noted that all sample preparations were carried out in vials inside a dry box to limit the exposure to atmospheric oxygen and water vapor. This extra caution was taken to eliminate or limit any interference by atmospheric oxygen. The sample solutions were then transferred from the dry box and set on the lab bench for over 24 hours in the dark while still capped tightly in the vial. During the period, the samples were intermittently agitated by sonication. This was to ensure that the bixin solution was thoroughly mixed with the TiO2 powder to attain a homogenous solution during the process. After 24 hr., the samples were centrifuged, decanted and then washed three times with chloroform in the dry box. In between each wash, the samples were agitated using a vortexer. The samples were allowed to sit and dry on the lab bench in the dark under nitrogen, wrapped in Al foil and kept in a desiccator (connected to the in-house vacuum system) until use. Table 1: Concentration of Bixin solutions and the amount of TiO2 added

Bixin (µM) 0.40 0.40 0.20

TiO2 (mg) 133.7 174.8 174.8

DRIFT Sample Preparation The DRIFT sample preparation and measurement was done in a reduced illumination environment by turning off most of the lights in the laboratory. The KBr granules, kept in an oven, were ground into fine powder for about 10 min with each TiO2/bixin sample in about a 1 7 ACS Paragon Plus Environment


TiO2/bixin:100 KBr ratio by weight. The prepared samples were then loaded into the Praying Mantis (IR instrument component used for DRIFT measurement) without modification and spectra were then recorded. Prior to the sample measurement, spectrum of raw TiO2 (previously heated at ~ 450 °C and allowed to cool in dry a box) mixed with KBr (1:100 by weight) was recorded. EPR Sample Preparation The EPR samples were prepared in a way similar to the protocol used in a previous study.4 Briefly, a Degussa P25 TiO2 sample was heated at about 120 ºC overnight and quickly transferred before cooling into a nitrogen dry box in an oxygen and water-free environment. A 3.6 mg bixin sample previously dissolved in chloroform, was dried with a stream of nitrogen to form a film, transferred into the dry box and dissolved in 4 mL of dry chloroform to give about 2.28 mM solution. The chloroform was previously passed through a column of basic alumina and stored over 3Å activated molecular sieves before use. Three samples were prepared by adding 1400, 1400 and 800 µL of the 2.28 mM bixin solution to 28, 58 and 6.4 mg TiO2 samples respectively. The three sample vials were vigorously agitated and then centrifuged. The TiO2 particles took on the deep reddish bixin color. UV vis spectra of the supernatants showed no or very little presence of any UV-active substance. The supernatant was discarded, and the samples washed with chloroform three times. The TiO2 nanoparticles retained the bright reddish color after washing, centrifuging and decanting. It must be noted that the wash and decantation processes were done in the dry box while the agitation and centrifuging were done on the lab bench. Dry chloroform was added to the bixin sample, agitated and then transferred into three separate 3 mm quartz EPR sample tubes in the dry box using a Pasteur pipette. The samples were


Page 8 of 32

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


transferred from the dry box and connected to a vacuum line. The samples were first frozen in liquid nitrogen, pumped on for 10-20 min under vacuum. The vacuum valve was then closed, the sample warmed to room temperature and frozen again in liquid. The repeated freeze-vacuumedthaw-freeze led to the evaporation of all the solvent from the sample. The sample was then evacuated overnight resulting in a very dry powdery sample in the EPR tube. The EPR tubes were then sealed under vacuum while still submerged in liquid nitrogen using a natural gas/oxygen torch. EPR Measurement and DFT Calculation CW EPR The CW EPR spectra were recorded with a Bruker ELEXYS E540 CW X-band EPR spectrometer before and after irradiation at 77 K. Samples were irradiated with a Y1711 Xe arc lamp from ILC Technology, filtered through water and glass to eliminate UV and IR radiation. Mims ENDOR The Mims ENDOR was recorded at τ =200 ns on Bruker ELEXSYS E-680W/X EPR spectrometer in an EN 4118X-MD4-W1 resonator in a Flexline cryostat with an ENI A-500 RF power amplifier using the Mims ENDOR (π/2-τ-π/2-T-π/2-τ-echo) pulse sequence with a 10 µs RF π-pulse applied during the delay time T. DFT Calculations The DFT calculations were performed on the Alabama Super Computer. The calculations were performed as detailed be Tay-Agbozo et al.4 Briefly, DFT calculations were used to optimized the cis and trans bixin radical cations and the hyperfine tensors calculated at the optimum geometry. SiAl surfaces that are known to stabilize Car radicals remain uncharacterized. This is because, no known Car radical has ever shown a quantifiable effect on



its hyperfine couplings or electron spin distribution.1-2, 7-9, 18 It has been demonstrated that,1-2, 7-9, 18

the matrix proton ENDOR line from surface -OH sites of SiAl is featureless and shows no

interaction between the electron spin and any surface atom. Since SiAl surface has not been characterized, it was not included in the DFT calculations. In addition, previous studies did not detect any influence of the surface on the EPR properties of other Car radicals.18 A proton loss from the13(13′) or 9(9′) methyl groups of the cis and trans radical cations result in the formation of the proton-loss neutral radicals. The unpaired spin density distribution for the trans and cis 13(13′) or 9(9′) proton-loss neutral radicals were generated from the optimized geometries. It was determined that, the cis radical cation was less stable than the trans by 1.26 kcal/mol. In contrast, the cis-9′ proton-loss neutral radical was found to be the most stable of the proton-loss neutral radicals while the cis-13 proton-loss neutral radical is the least by 4.54 kcal/mol.4 Results and Discussion DRIFT The spectra in Figure 3 shows the DRIFT spectra of bixin adsorbed on TiO2 and of Degussa P25 TiO2 nanoparticles treated in chloroform. The hydroxyl surface of TiO2 is given in Figure 2c. The resultant difference spectrum (Figure 4) reveals interesting features. There are four distinct regions shown in the difference spectrum which clearly indicate the adsorption of bixin on TiO2. These regions are 1200 – 1800, 2800 – 3000, 3000 – 3600 and 3700 – 3900 cm-1 respectively, as laid out in Table 2.


Page 10 of 32

Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: DRIFT spectra - (a) Red spectrum: Bixin adsorbed on TiO2 and (b) Blue spectra: TiO 2 powder

The presence of the carboxylate group is seen in Figure 4 by the adsorption in the 12001800 cm-1 region. The methine group from polyene chain gives an absorption peak in the 28003000 cm-1 region. The broad dip in the 3000-3600 cm-1 region indicate removal of H2O and strongly adsorbed OH. The peaks from 3730-3910 cm-1 indicate the presence of OH group from bixin on TiO2. Table 2: Wavenumber ranges and their corresponding functional groups observed in the DRIFT spectrum of Bixin adsorbed on TiO2.

Range (cm-1)

Functional grp.

1200-1800

Carboxylate

This region shows the presence of carboxylate

2800-3000

Methine

Shows the presence of methine groups from polyene chain

3000-3610

H2O/OH

3730-3910

Hydroxyl

Broad dip is an indication of H2O removal/strongly absorbed OH Indicates the presence of OH group (bixin or TiO2 or air)

Comments



Figure 4: Difference spectrum of bixin adsorbed on TiO 2 and spectrum of TiO 2 from Figure 2 powder alone. The assignments are listed in Table 2.

CW EPR The line shape of the CW spectrum after 1 min irradiation at 77 K (Figure 5: blue) is roughly Gaussian in shape with a line width of 0.96 mT and giso of 2.003. The spectrum is nearly symmetrical in shape except for the region at higher fields beyond 339 mT due to titanium based radicals.19-20 Prior to irradiation, no EPR signal (Figure 5: red) was detected indicating little or no radical activity. To a large extent, the line shape of Car at the X-band comes from the unresolved hyperfine interactions with the protons of the radical.21-22


Page 12 of 32

Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5: CW spectrum of Bixin on TiO2 at 77 K before irradiation (red; R) and after 1 min irradiation (blue; B) with about 0.96 mT line width and giso of 2.003

Mims ENDOR The Mims 1H ENDOR spectrum of bixin adsorbed on TiO2 (Figure 6) after 1 min irradiation is plotted as the absolute ENDOR effect which extended all the way beyond ±13 MHz. The line shape and features of the spectrum are very similar to previous work on silica alumina4 and that of other Car.2, 7-8 The intense midsection of the spectrum is due to the small proton couplings probably from neighboring bixin molecules. Previous experience working with Car on MCM-41 shows that surface protons from the matrix are prominent in this mid area, which are difficult to model.23 An examination of the Mims 1H ENDOR spectrum of bixin on TiO2 compared to that of bixin on silica alumina (Figure 6) clearly shows common features between the two spectra. The features between ±2.5 and ±7.5 MHz are common features with minor differences in spectral intensity. Beyond ±7.5 MHz, it becomes difficult to fit the spectral features and hence impossible



to determine all the possible radicals and their percent contributions to the spectral features. The spectral features for bixin on TiO2 could not be fitted beyond ±7.5 as it was done for bixin on silica alumina.4

Figure 6: Mims ENDOR of Bixin on TiO 2 (red) after irradiation at 9.79 GHz and 40K. The intense Mims ENDOR signal from -2.5 to +2.5 MHz is due to weakly coupled protons, from neighboring attached bixin molecules to the TiO2 matrix. Blue: Mims ENDOR spectra of Bixin on silica alumina after irradiation. Only the weakly coupled protons from • the bixin radical cation or the proton-loss neutral, [Bix-H] , are observed for adsorption on silica alumina

Discussion The spectral features in the 1200–1800 cm-1 region in DRIFT indicates the presence of carboxylate functional groups. Carboxylates are known to bound to surfaces via four distinct coordination modes (Figure 7).24 In view of that, the IR frequency of a carboxylate adsorbed on a surface could be in the region of 1510 – 1650 cm-1 and 1280 – 1400 cm-1 for asymmetric and symmetric stretches, respectively, if the mode of adsorption is ionic. In the case of a unidentate and bidentate bridging coordination, the observed carboxylate band would be ~ 1700 cm-1 and ~1400 cm-1, while a bidentate coordination would have bands at 1550 cm-1 and 1450 cm-1,


Page 14 of 32

Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


respectively.24 All the above stretching and vibrational bands lie in the 1200 – 1800 cm-1 frequency region and are very difficult to isolate individually.

Figure 7: Coordination of carboxylate groups on a monovalent cation.

Also, the 2800 – 3000 cm-1 spectral range shows symmetric and asymmetric stretching features indicative of methine groups.25 It must be noted however that, vibrational modes of terminal methyl groups also show up in the 2800 – 3000 cm-1 region.26 Thus, the presences of spectral features within the 2800 – 3000 cm-1 region is an indication of the presence of a polyene chain with methyl groups. Meanwhile, the features of the spectral region from 3000–3600 cm-1 show a broad dip typical of H2O removal in an IR spectrum. In addition, the broad band in this spectral region is characteristic of strongly adsorbed OH group.25 The dip shows the displacement of H2O upon chemisorption of the bixin chromophore and or a strongly adsorbed hydroxyl.25 Maximum information could only be obtained from any of the spectral overlapping regions by using separate prisms of lithium fluoride, calcium fluoride and sodium chloride in the respective regions to help isolate the bands.26 Finally, the band in the region of 3700 – 3900 cm-1 indicates the presence a hydroxyl group.27 All the above spectral features are strong indication of the chemisorption of bixin on TiO2 for which the radical formation of bixin on TiO2 is worth studying. 15 ACS Paragon Plus Environment


Vittadini et al.28 reported that carboxylate headgroups preferred a dissociated bridging bidentate and a molecular monodentate configurations, Figure 8, as way of anchoring onto a TiO2 substrate with the bidentate bridging configuration providing enhanced stability compared to the monodentate.29 The DRIFT data is indicative of bixin grafted on TiO2.

Figure 8: Monodentate and bidentate configurations preferred by a carboxylate headgroup on binding on a TiO2 surface.

As reported by Park et al.30, the density of hydroxyl groups present on P25 TiO2 is ca. 250 µmol/g. Based on the figure reported by Park et al., any carboxylate headgroup grafting by way of molecular monodentate and/or bidentate bridging configuration would result in an estimated monolayer coverage of roughly ~125-250 µmol/g on P25 TiO2 surface. Meanwhile, the Brunauer-Emmett-Teller (BET) surface area of the P25 TiO2 used to prepare the samples is 28.38 m2/g. Hence by rough estimation, bixin adsorption at monolayer coverage would be 5.3 × 1018 bixin molecules /m2 or 1 bixin molecule/19 Å2. The CW EPR spectrum (Figure 5) of bixin on TiO2 before irradiation shows no signal and is relatively flat unlike that of bixin on silica alumina as seen in Figure 9,4 which is gaussian in shape, very intense and with a line width of 1.62 mT. However, after irradiation of the TiO2 sample, an intense signal was detected which was roughly gaussian in shape with a line width of 0.96 mT (Figure 5). In the case of silica alumina, after light irradiation, the intensity of the signal


Page 16 of 32

Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


was reduced 4-5 fold and the spectrum still maintained the gaussian shape. It is obvious from Figures 5 and 9 that bixin radicals were trapped on silica alumina before irradiation, while on TiO2 the radicals were only generated after irradiation. This indicates that on TiO2, the Car requires light to be excited and thus transfer electrons, a desirable function if the Car is to serve as a sensitizer.

Figure 9: CW EPR spectrum at 77 K of bixin on SiAl before irradiation (red) with a line width of 1.62 mT and giso = 2.002 and after 10 min visible light irradiation (blue) with a 4-5 fold decrease in intensity.4

The Mims ENDOR spectral intensity between -2.5 MHz < and +7.5 and < -7.5



MHz indicate the presence of species on TiO2 that were not present on silica alumina. In addition, the intense midsection spectral features for TiO2 are not due to radical cations from bixin adsorbed on the surface; but due to matrix protons. A look at the features and intensities of the proton-loss neutral radical signals between -7.5 MHz < and < -2.5 MHz and between +2.5 MHz > and > +7.5 MHz almost mirrors that for silica alumina but for the peak at about -3.6 MHz and +6.1 MHz. It can be seen from Figure 6 that there are no spectral features for the silica alumina spectrum below -10.0 MHz and above +10.0 MHz, respectively, as the spectral line is relatively flat. But in the same range for TiO2, the spectral line shows multiple peaks with reasonable intensities. Judging from the possible radicals from the silica alumina spectrum, new radicals formed by bixin adsorbed on TiO2 that are present below -10.0 MHz and above +10.0 MHz are not likely to be cis and trans 9/9′ nor 13/13′, which contributed about half of the spectral features of silica alumina. A quick look at the spectral features below and above ±7.5 MHz (Figure 6) shows that there is not enough resolution in those regions to specifically identify the radicals for bixin adsorbed TiO2. It should be noted that in the sample preparation and irradiation, UV and IR radiation were cut off by passing the beam of light through a water filter. This was to reduce the excitation of any TiO2 by UV and or IR rays but not the Car bixin. Carotenoids, like any other chromophore, absorb mostly in the visible region spectrum unlike TiO2. With band gap energy of 3.2 and 3.0 eV for anatase and rutile respectively, TiO2 only absorbs below the wavelength band edge of the visible region (below 400 nm).31-32 With the filtering of the UV and IR radiation during the irradiation process, the EPR signals from interfering Ti radicals were reduced relative to the EPR signal around g = 2.0023 due to the bixin radicals as shown in Figure 5.


Page 18 of 32

Page 19 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Conclusion The MS analysis in both positive and negative modes show peaks consistent with bixin molecule, and the NMR of the purified bixin shows the cis configuration. DRIFT measurements have been used to characterize the adsorption of cis bixin on TiO2, and Mims-ENDOR and DFT calculations have used to identify the bixin radicals stabilized on TiO2. In the sample preparation process, the TiO2 with cis bixin was washed three times with chloroform, the solvent used in the sample preparation, yet the white TiO2 powder still maintained the reddish coloration of cis bixin. On silica alumina, radicals were generated prior to irradiation, and upon irradiation the spectral signal was reduced by a factor of 4-5. On the other hand, on TiO2, no signal was detected until after irradiation. The adsorption of cis bixin on TiO2 results in the formation of radicals only upon irradiation with a Xe arc lamp. But in the case of cis bixin adsorbed on silica alumina, radical cations were produced upon adsorption prior to the sample being irradiated (a common occurrence for most trans Car studied). After the samples were irradiated, the radical intensity reduced by about 4-5 fold, which is contrary to all the trans Car that have been studied on silica alumina. The center (-2.5 to +2.5 MHz) part of the Mims ENDOR spectrum for adsorption on TiO2 is 10 times the intensity of the remaining spectrum, indicating a significant concentration of protons from bixin molecules surrounding the bixin proton-loss neutral radicals. The percent of radical cations on TiO2 is less than it is on silica alumina with about 1:2 TiO2/silica alumina ratio. The nature of the spectrum below -10.0 MHz and above +10.0 MHz shows that, besides the radicals cis and trans 9/9′ and 13/13′, there are some unidentified protonloss neutral radicals which contribute to the TiO2 spectral features, but these do not occur on silica alumina. Lack of good spectral resolution makes the identification of the spectral signals below and beyond ±7.5 MHz impossible. Even though spectrum could not be fit, the area under



the TiO2 spectrum is larger than that under the silica alumina spectrum. However, due to poor resolution of the ENDOR spectral features, the percent distribution of the trapped bixin radicals could not be quantified, although bixin radical cations and proton-loss neutral radicals are shown to have been formed upon light irradiation. It a promising sign for bixin on TiO2 as a sensitizer. The Brunauer-Emmett-Teller (BET) surface area of the P25 TiO2 was determined to be 28.37 m2/g. On the assumption that the amount of OH groups present on P25 TiO2 is ~ 250 µmol/g,30 by simple calculation, the number of molecules of OH head group per m2 of the TiO2 used, is ~ 5.31 × 1018. Presuming a monolayer coverage, the number of molecules of bixin required on the surface would be ~ 5.31 × 1018 and ~ 2.65 × 1018, for a mono- and bidentate bridging respectively. On the other hand, for an average of 1.2 mL of 2.28 µM concentration of bixin, the number of bixin molecules per m2 of the P25 TiO2 theoretically would be, ~ 5.81 × 1013 and ~ 2.90 × 1013 m2, assuming a trans bixin mono and bidentate bridging coverage, respectively. It therefore shows that the number of bixin molecules presumed to be on the P25 TiO2 for a monolayer coverage falls short by order of ~105. Bixin possesses properties that could be an advantage as a photovoltaic material. It transfers electrons when irradiated, it attaches to TiO2, and is readily available since it is used industrially as an FDA approved dye in food, cosmetics and drugs. It exhibits a high extinction coefficient of 105 and possesses the highest measured oxidation potential (0.94 vs SCE) of all Car, reducing the chance of oxidation by other species. AKNOWLEDGEMENTS This work was supported in part by the Chemical Sciences, Geosciences, and Biosciences Division, Office of Basic Sciences, U.S. Department of Energy, Grant DE-FG02-86ER-13465 which provided RA support for Sefadzi Tay-Agbozo, supplies and support for the P.I. L.D


Page 20 of 32

Page 21 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Kispert. The EPR measurements were made possible by the National Science Foundation for EPR Instrument Grants CHE-0342921 and CHE-0079498 to UA. Professor Michael Bowman is thanked for helpful comments. References 1. Jeevarajan, A. S.; Kispert, L. D.; Piekara-Sady, L., An ENDOR study of carotenoid cation radicals on silica—alumina solid supports. Chem. Phys. Lett. 1993, 209, 269-274. 2. Focsan, A. L.; Bowman, M. K.; Konovalova, T. A.; Molnár, P.; Deli, J.; Dixon, D. A.; Kispert, L. D., Pulsed EPR and DFT characterization of radicals produced by photo-oxidation of zeaxanthin and violaxanthin on silica-alumina. J. Phys. Chem. B 2008, 112, 1806-1819. 3. Focsan, A. L.; Pan, S.; Kispert, L. D., Electrochemical Study of Astaxanthin and Astaxanthin n-Octanoic Monoester and Diester: Tendency to Form Radicals. J. Phys. Chem. B 2014, 118, 2331-2339. 4. Tay-Agbozo, S. S.; Krzyaniak, M. D.; Bowman, M. K.; Street, S.; Kispert, L. D., DFT and ENDOR Study of Bixin Radical Cations and Neutral Radicals on Silica–Alumina. J. Phys. Chem. B 2014, 119, 7170-7179. 5. Crépeau, G.; Montouillout, V.; Vimont, A.; Mariey, L.; Cseri, T.; Maugé, F., Nature, structure and strength of the acidic sites of amorphous silica alumina: an IR and NMR study. J. Phys. Chem. B 2006, 110, 15172-15185. 6. Bhaduri, S.; Mukesh, D., In Chemical Industry and Homogeneous Catalysis: Mechanisms and Industrial Applications; Homogeneous Catalysis. In Homogeneous Catalysis, John Wiley & Sons, Inc: New York, USA, 2000; pp 1-12. 7. Polyakov, N. E.; Focsan, A. L.; Bowman, M. K.; Kispert, L. D., Free radical formation in novel carotenoid metal ion complexes of astaxanthin. J. Phys. Chem. B 2010, 114, 16968-16977. 8. Focsan, A. L.; Bowman, M. K.; Shamshina, J.; Krzyaniak, M. D.; Magyar, A.; Polyakov, N. E.; Kispert, L. D., EPR Study of the Astaxanthin n-Octanoic Acid Monoester and Diester Radicals on Silica–Alumina. J. Phys. Chem. B 2012, 116, 13200-13210. 9. Konovalova, T. A.; Dikanov, S. A.; Bowman, M. K.; Kispert, L. D., Detection of anisotropic hyperfine components of chemically prepared carotenoid radical cations: 1D and 2D ESEEM and pulsed ENDOR study. J. Phys. Chem. B 2001, 105, 8361-8368. 10. Tay-Agbozo, S.; Street, S.; Kispert, L. D., The Carotenoid Bixin found to exhibit the highest measured carotenoid Oxidation Potential to date consistent with its practical protective use in Cosmetics, Drugs and Food. J. Photochem. Photobiol., B 2018, 186, 1-8. 11. Gao, G.; Kispert, L., Photovoltaic response of carotenoid-sensitized electrode in aqueous solution: ITO coated with a mixture of TiO2 nanoparticles, carotenoid, and polyvinylcarbazole. J. Chem. Soc., Perkin Trans. 2 1999, 0, 1225-1230. 12. Gao, F. G.; Bard, A. J.; Kispert, L. D., Photocurrent Generated on a CarotenoidSensitized TiO2 Nanocrystalline Mesoporous Electrode. J. Photochem. Photobiol., A 2000, 130, 49-56. 13. Silva, M. C. D.; Botelho, J. R.; Conceiçao, M. M.; Lira, B. F.; Coutinho, M. A.; Dias, A. F.; Souza, A. G.; Filho, P. F. A., Thermogravimetric investigations on the thermal degradation of bixin, derived from the seeds of annatto (Bixa orellana L.). J. Therm. Anal. Calorim. 2005, 79, 277-281. 21 ACS Paragon Plus Environment


14. Rehbein, J.; Dietrich, B.; Grynbaum, M. D.; Hentschel, P.; Holtin, K.; Kuehnle, M.; Schuler, P.; Bayer, M.; Albert, K., Characterization of bixin by LC‐MS and LC‐NMR. J. Sep. Sci. 2007, 30, 2382-2390. 15. Mercadante, A. Z.; Steck, A.; Pfander, H., Isolation and identification of new apocarotenoids from annatto (Bixa orellana) seeds. J. Agric. Food. Chem. 1997, 45, 1050-1054. 16. Chiste, R. C.; Yamashita, F.; Gozzo, F. C.; Mercadante, A. Z., Simultaneous extraction and analysis by high performance liquid chromatography coupled to diode array and mass spectrometric detectors of bixin and phenolic compounds from annatto seeds. J. Chromatogr. A 2011, 1218, 57-63. 17. Gomez-Ortiz, N. M.; Vázquez-Maldonado, I. A.; Pérez-Espadas, A. R.; Mena-Rejón, G. J.; Azamar-Barrios, J. A.; Oskam, G., Dye-sensitized solar cells with natural dyes extracted from achiote seeds. Sol. Energy Mater. Sol. Cells 2010, 94, 40-44. 18. Kispert, L. D.; Polyakov, N. E., Carotenoid Radicals: Cryptochemistry of Natural Colorants. Chem. Lett. 2010, 39, 148-155. 19. Hurum, D. C.; Agrios, A. G.; Gray, K. A.; Rajh, T.; Thurnauer, M. C., Explaining the enhanced photocatalytic activity of Degussa P25 mixed-phase TiO2 using EPR. J. Phys. Chem. B 2003, 107, 4545-4549. 20. Macdonald, I. R.; Rhydderch, S.; Holt, E.; Grant, N.; Storey, J. M. D.; Howe, R. F., EPR studies of electron and hole trapping in titania photocatalysts. Catal. Today 2012, 182, 39-45. 21. Konovalova, T. A.; Krzystek, J.; Bratt, P. J.; Van Tol, J.; Brunel, L.-C.; Kispert, L. D., 95− 670 GHz EPR Studies of Canthaxanthin Radical Cation Stabilized on a Silica− Alumina Surface. J. Phys. Chem. B 1999, 103, 5782-5786. 22. Kispert, L. D.; Konovalova, T.; Gao, Y., Carotenoid radical cations and dications: EPR, optical, and electrochemical studies. Arch. Biochem. Biophys. 2004, 430, 49-60. 23. Lawrence, J.; Focsan, A. L.; Konovalova, T. A.; Molnar, P.; Deli, J.; Bowman, M. K.; Kispert, L. D., Pulsed Electron Nuclear Double Resonance Studies of Carotenoid Oxidation in Cu (II)-Substituted MCM-41 Molecular Sieves. J. Phys. Chem. B 2008, 112, 5449-5457. 24. Lu, Y.; Miller, J. D., Carboxyl stretching vibrations of spontaneously adsorbed and LBtransferred calcium carboxylates as determined by FTIR internal reflection spectroscopy. J. Colloid Interface Sci. 2002, 256, 41-52. 25. Sinclair, R. G.; McKay, A. F.; Jones, R. N., The Infrared Absorption Spectra of Saturated Fatty Acids and Esters1. J. Am. Chem. Soc. 1952, 74, 2570-2575. 26. Sinclair, R. G.; McKay, A. F.; Myers, G. S.; Jones, R. N., The Infrared Absorption Spectra of Unsaturated Fatty Acids and Esters1. J. Am. Chem. Soc. 1952, 74, 2578-2585. 27. Yu, Z.; Chuang, S. S. C., Probing methylene blue photocatalytic degradation by adsorbed ethanol with in situ IR. J. Phys. Chem. C 2007, 111, 13813-13820. 28. Vittadini, A.; Selloni, A.; Rotzinger, F. P.; Grätzel, M., Formic acid adsorption on dry and hydrated TiO2 anatase (101) surfaces by DFT calculations. J. Phys. Chem. B 2000, 104, 1300-1306. 29. Nazeeruddin, M. K.; Humphry-Baker, R.; Liska, P.; Grätzel, M., Investigation of sensitizer adsorption and the influence of protons on current and voltage of a dye-sensitized nanocrystalline TiO2 solar cell. J. Phys. Chem. B 2003, 107, 8981-8987. 30. Park, H.; Bae, E.; Lee, J.-J.; Park, J.; Choi, W., Effect of the Anchoring Group in Ru− Bipyridyl Sensitizers on the Photoelectrochemical Behavior of Dye-Sensitized TiO2 Electrodes: Carboxylate versus Phosphonate Linkages. J. Phys. Chem. B 2006, 110, 8740-8749.


Page 22 of 32

Page 23 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


31. Serpone, N., Is the band gap of pristine TiO2 narrowed by anion-and cation-doping of titanium dioxide in second-generation photocatalysts? J. Phys. Chem. B 2006, 110, 2428724293. 32. Lisachenko, A. A.; Mikhailov, R. V., Point defects as the centers of titanium dioxide sensitization in the visible spectral range. Tech. Phys. Lett. 2005, 31, 21-24.

TOC Image



119x151mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 24 of 32

Page 25 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


333x148mm (96 x 96 DPI)



127x101mm (220 x 220 DPI)


Page 26 of 32

Page 27 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5: CW spectrum of Bixin on TiO2 at 77 K before irradiation (red; R) and after 1 min irradiation (blue; B) with about 0.96 mT line width and giso of 2.003 185x147mm (300 x 300 DPI)



116x96mm (220 x 220 DPI)


Page 28 of 32

Page 29 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


174x87mm (96 x 96 DPI)



168x71mm (96 x 96 DPI)


Page 30 of 32

Page 31 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


109x88mm (220 x 220 DPI)



TOC Image 187x155mm (300 x 300 DPI)


Page 32 of 32

DRIFT and ENDOR Study of the Carotenoid Bixin Attached to

Recommend Documents