Reaction Pathways and Intermediates in Selective Ring Opening of

Sep 8, 2016 - While the catalytic hydrogenolysis of biomass-derived aromatic cyclic compounds to functionalized long chain alcohols and polyols has be...
1 downloads 3 Views 1MB Size
Subscriber access provided by Northern Illinois University

Article

Reaction Pathways and Intermediates in Selective Ring Opening of Biomass-Derived Heterocyclic Compounds by Iridium Glen Richard Jenness, Weiming Wan, Jingguang G Chen, and Dionisios G. Vlachos ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01310 • Publication Date (Web): 08 Sep 2016 Downloaded from http://pubs.acs.org on September 9, 2016

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 free 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 accessible to all readers and 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.

ACS Catalysis 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 22

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

ACS Catalysis

Reaction Pathways and Intermediates in Selective Ring Opening of Biomass-Derived Heterocyclic Compounds by Iridium Glen R. Jenness†§, Weiming Wan†‡§, Jingguang G. Chen†‡*, and Dionisios G. Vlachos†* †Catalysis Center for Energy Innovation (CCEI), University of Delaware, 221 Academy St. Newark DE 19716, USA

‡Department of Chemical Engineering, Columbia University, 500 W. 120th St. New York NY, 10027, USA

Abstract: While the catalytic hydrogenolysis of biomass-derived aromatic cyclic compounds to functionalized long chain alcohols and polyols has been known for decades, the factors that control the selectivity remain either unknown or controversial. Previous reports have hypothesized full ring saturation of the aromatic ring is necessary prior to hydrogenolysis. Contradictorily, recent studies have shown hydrogenolysis occurs prior to the saturation of the conjugated bonds. Furthermore, it has been assumed the functional groups present are fully reduced prior to hydrogenolysis; however, this has not been shown a priori. In order to resolve these controversies, we combine density functional theory and high resolution electron energy loss spectroscopy (HREELS) to probe the catalytic hydrogenolysis of saturated and unsaturated

ACS Paragon Plus Environment

1

ACS Catalysis

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

Page 2 of 22

heterocyclic molecules (furan, furfural, furfuryl alcohol, and tetrahydrofurfuryl alcohol) on iridium. Our results reveal that full saturation of the aromatic ring is not only unnecessary, but leads to slower kinetics and differing selectivities. In contrast to previous studies, we show selective partial ring saturation can enhance the kinetics of the hydrogenolysis process. Reduction/oxidation of the functional group leads to a change in the electronegativity, resulting in a change in selectivity. These results provide important mechanistic insights allowing for further improvement of catalysts for the effective transformations of biomass-derived oxygenates to value-added products.

KEYWORDS: ring opening, furfural, furfuryl alcohol, tetrahydrofurfuryl alcohol, density functional theory, HREELS, Ir, diol Introduction As the chemical economy gradually shifts from fossil fuels to renewable resources, it becomes important to develop technologies to upgrade renewable platform molecules to final products of industrial relevance. It has been shown the renewable heterocyclic molecules 5hydroxymethylfurfural and furfural can be produced in high yields by depolymerizing and dehydrating lignocellulosic biomass.1-2 Hydrogenolysis of these heterocyclics has been shown to be a viable route towards the production of long-chain hydrocarbons, alcohols, and polyols.1-11 Selective hydrogenolysis of furfural at the C–O bond(s) in the aromatic ring with an iridium catalyst has been shown to be effective in the production of 1,5-pentanediol (1,5-PeD), a valuable precursor for the production of polyesters and polyurethanes.6,

12-14

While the

hydrogenolysis of furfural to 1,5-PeD has been known for decades,15 heterocyclic or not the reaction proceeds through the full saturation of the aromatic ring9, 16-21 remains a controversial

ACS Paragon Plus Environment

2

Page 3 of 22

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

ACS Catalysis

topic. Recent experiments22 have indicated that the hydrogenolysis of unsaturated homo- and heterocyclics is more facile over that of their saturated equivalents. Additionally, many aromatic heterocyclics are functionalized and the influence of the reduction/oxidation of the functional groups on the hydrogenolysis process has not been established. In the current study, we apply density functional theory (DFT) and HREELS to elucidate the selectivity determining factors and the role of saturation of the aromatic ring for the hydrogenolysis of heterocyclic molecules on the Ir(111) surface. We demonstrate that full saturation of the aromatic ring kinetically hinders the hydrogenolysis of the heterocyclic ring, and partial saturation at selected positions in the aromatic ring results in a more facile hydrogenolysis process and aids in controlling the selectivity towards terminal diols. The reduction/oxidation of the aromatic ring side group results in a rearrangement of the electronic density of the aromatic π-system, which shifts the hydrogenolysis selectivity towards either terminal or secondary diols. Furthermore, there is a competition between the kinetics and thermodynamics governing the hydrogenolysis, indicating a strong temperature and time effect on the selectivity. These insights will allow for the development of more efficient catalysts and processes that can potentially render renewable chemicals commercial. Results and Discussion

ACS Paragon Plus Environment

3

ACS Catalysis

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

Page 4 of 22

Figure 1 Reactants and final products considered in the current study. Terminal diols (i.e., 1,5PeD) are produced via the selective ring opening of the Cβ–O bond, while secondary diols (i.e., 1,2-PeD) are produced via Cω–O ring opening.

Figure 2 Heterocyclic ring opening reaction pathways on the Ir(111) surface for (a) Cβ–O scission and (b) Cω–O scission (see Figure 1a for labeling). Here, the zero is taken to be the

ACS Paragon Plus Environment

4

Page 5 of 22

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

ACS Catalysis

energy of the Ir(111) slab with the adsorbed heterocyclic species and atomic hydrogen. The NBO point charges of the carbon atom (QCNBO) from the gas-phase species are reported in (c) for the β carbon and (d) for the ω carbon. Numerical values can be found in Table S1 of the Supporting Information (SI). In order to examine the controversy surrounding the role of the aromatic ring saturation, we carried out density functional theory (DFT) calculations on the unsaturated heterocyclics furan, furfural, and furfuryl alcohol (FOL), as well as the fully saturated tetrahydrofurfuryl alcohol (THFA) on the Ir(111) surface (as shown in Figure 1). Previous publications on iridium6, 9, 13-14, 23

have shown that the dominant pathways include reduction of the side group,

hydrogenation of the aromatic C=C bonds, and hydrogenolysis of the ring C-O bonds. In contrast to palladium24-25, decarbonylation has not been shown to occur on iridium23, and it is not considered in the current study. The inclusion of furan into the data set is to provide a “control” in order to better understand the effect of the substituent group. The reaction barriers and energies for the ring opening of these heterocyclics are presented in Figure 2. Among the unsaturated heterocyclics, furan has the highest ring opening barrier. Inclusion of either a carbonyl (–CHO) or alcohol (–CH2OH) substituent group leads to a modest lowering of the ring opening barrier, indicating the inclusion of a functional group results in faster kinetics. In contrast to prior studies, fully saturating the aromatic ring (i.e., THFA) results in an increase in the ring opening barrier, resulting in slower kinetics and indicates the production of THFA can hinder the conversion. Consideration of the ring opening barriers indicates furfural would kinetically favor the formation of the secondary diol 1,2-pentanediol (1,2-PeD), while FOL and THFA would favor the terminal diol 1,5-PeD.

ACS Paragon Plus Environment

5

ACS Catalysis

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

Page 6 of 22

Examination of the reversibility (i.e., Ea – Erxn) of the surface reactions (Figures 2a and 2b) reveals if one type of C–O bond (Figure 1) is favored kinetically, the other C–O bond is favored thermodynamically. Therefore, the ring opening selectivity of heterocyclic molecules is determined by a balance of kinetic and thermodynamic factors. Higher temperatures will be required to overcome the kinetic limitations of furfural in order to shift the selectivity towards 1,5-PeD. In contrast, the production of 1,5-PeD from FOL will require lower temperature. While THFA shows the same thermodynamic trend as FOL, higher temperature (relative to FOL) will be required for THFA due to the higher activation energy (∆Ea>0.5 eV). Consequently, the reaction conditions will need to be carefully optimized.

Figure 3. Crystal orbital overlap population (COOP) analysis for the heterocyclic C-O bonds. (a) Results for the Cβ–O bond. (b) Results for the Cω–O bond. Positive peaks denote a bonding interaction, while negative peaks denote an anti-bonding peak. The Fermi level is set at zero (denoted with a black vertical line); peaks below the Fermi level are occupied molecular orbitals, while peaks above the Fermi level are unoccupied.

ACS Paragon Plus Environment

6

Page 7 of 22

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

ACS Catalysis

The trends in the ring opening barriers can be rationalized at a molecular level via the application of the crystal orbital overlap population (COOP) analysis to the gas-phase heterocyclics. Examination of the COOP curves in Figure 3 reveals furan displays a large bonding peak at –7.58 eV which corresponds to a molecular orbital responsible for its high ring opening barrier. Inclusion of either a carbonyl (–CHO) or an alcohol (–CH2OH) group results in this peak decreasing in magnitude for both the Cβ–O and Cω–O bonds, indicating the substituent group withdraws electronic density from this molecular orbital, in agreement with the decrease in ring opening barriers for furfural and FOL relative to furan. While THFA shows a decrease in the electronic density at this molecular orbital (indicating a potential lowering of the ring opening barrier), it should be noted that THFA is weakly bound to the surface. In the absence of strong coupling between THFA and the Ir(111) surface, the ring opening transition state should resemble that of a gas-phase transition state (see Figure S5 of the SI). Consequently, this leads to an increase in the ring opening barrier (relative to furan) that is observed in Figure 2. To quantitatively characterize the electron withdrawing effect trend, the natural bond order (NBO) method is applied to the gas-phase heterocyclics and is subsequently compared to the reaction barriers of the surface bound species, with the results in Figures 2c and 2d. A depletion in the electronic density (corresponding to a positive charge) on the carbon atom results in a lower ring opening barrier for the neighboring C–O bond. Taking the Cβ–O bond as an example, furfural and FOL show a decrease in the electron density on the Cβ atom; this corresponds to the lower Cβ–O ring opening barrier for these species relative to furan. As FOL shows a larger electron density depletion compared to furfural, this provides a molecular level explanation of why 1,5-PeD is more kinetically preferred compared to furfural. Interestingly, comparison of the carbon point charges for furfural shows the β carbon having a higher positive

ACS Paragon Plus Environment

7

ACS Catalysis

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

Page 8 of 22

charge and a higher Cβ–O ring opening barrier relative to the ω carbon and the Cω–O bond. This discrepancy is attributed to the higher availability of unoccupied anti-bonding orbitals as shown in Figure 3. In order to provide physical insight into these results, the following conceptual picture is adopted. As the π-system of the furan ring interacts with the d-band of the Ir(111) surface, electron density from the metal surface can flow into the unoccupied molecular orbitals of the molecule.26 Examination of unoccupied region for the Cω–O bond in both furfural and THFA (Figure 2b) reveals an increased anti-bonding character in the unoccupied region, indicating the degree of back-bonding contributes to the selectivity to either 1,2-PeD or 1,5-PeD. This is in contrast to FOL, where the anti-bonding nature of the unoccupied region is nearly identical for both C–O bonds, indicating that the back-bonding from the metal d-bands would be equivalent, and the selectivity controlling factor arises from the presence of the anti-bonding peaks in the occupied region. As the degree of back-bonding is proportional to the overlap between the heterocyclics molecular orbitals and the d-band, the binding strength of the heterocyclic compound to the metal surface affects the selectivity. As the binding of THFA to the Ir(111) surface is significantly weaker than the unsaturated heterocyclics, the overlap between the σ molecular orbitals of THFA and the d-band of the metal surface is small, resulting in a poor back donation to the unoccupied anti-bonding molecular orbitals of THFA. This weak back-bonding is consistent with the large bonding peak in the unoccupied region of the Cβ–O bond (Figure 3a); as the overlap between this molecular orbital and the metal d-band is small, only the lowest lying anti-bonding orbitals are occupied. Consequently, cleaving the Cβ–O is more favorable over the Cω–O bond as a larger anti-bonding peak is found for this molecular orbital. Thus, the selectivity

ACS Paragon Plus Environment

8

Page 9 of 22

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

ACS Catalysis

towards either 1,5-PeD or 1,2-PeD is influenced by the electronegativity of the side group, which increases the availability of unoccupied anti-bonding orbitals for back-bonding. Table 1 Reaction energetics (in eV) for the reduction of furfural to FOL on the Ir(111) surface. Reaction

Ea

Erxn

Cα-O:+H@O

0.41 –0.17

Cα-OH:+H@Cα

0.70 –0.16

Table 2 Reaction barriers and energies (in eV) for partial ring saturation of furfural and FOL on the Ir(111) surface. Position Furfural

FOL

Ea

Erxn

Ea

α

0.75

+0.52

N.A.

β

N.A.

γ

0.98

δ ω

Erxn

0.67

+0.35

+0.61

0.81

–0.11

1.11

+0.24

0.91

+0.12

0.68

+0.12

0.67

+0.41

On the Ir(111) surface, the fully saturated aromatic ring has weaker binding between the heterocyclic and the catalyst, resulting slower kinetics. However, the effect of the reduction of the aromatic ring side group, as well as the effect of partial aromatic ring saturation, has only been partially explored previously. We consider several hydrogenation reactions, starting with the reduction of furfural to FOL (Table 1), which has been shown to occur on iridium catalysts.23 The conversion proceeds through the initial hydrogenation of the carbonyl oxygen (denoted as

ACS Paragon Plus Environment

9

ACS Catalysis

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

Page 10 of 22

+H@O in Table 1) with the subsequent hydrogenation of the α carbon (+H@Cα); while both steps are exothermic, the low barriers make the conversion reversible. As a consequence, furfural and FOL would co-exist on the Ir(111) surface.

Figure 4 Effect of ω carbon saturation on the ring opening of furfural on the Ir(111) surface (units in eV). Table 2 summarizes the energetics for the hydrogenation of furfural and FOL on the Ir(111) surface at the C positions (Figure 1a). It is readily apparent that with the exception of the β and δ positions, the hydrogenation barriers for furfural are either comparable to the ring opening barriers or lower, with each reaction being endothermic. The low hydrogenation barriers in conjunction with the endothermicity reported in Table 2 indicate that each reaction has a moderate to a high degree of reversibility (low reverse barriers). Intermediates resulting from hydrogenation at the α and γ positions are unlikely to be present in high coverages due to the low reverse barriers of these reactions (potentially low equilibrium constants). Hydrogenation at the δ position has the highest reaction barrier and is higher than the ring opening barriers and is thus unfavorable. Similarly, we expect species hydrogenated at the ω position to be partially equilibrated. Interestingly, the ω hydrogenation of furfural lowers the Cβ–O ring opening barrier

ACS Paragon Plus Environment

10

Page 11 of 22

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

ACS Catalysis

(Figure 4), and consequently shifts the selectivity to 1,5-PeD. These results point to an effect of the H2 pressure on the selectivity; in the presence of excess H2, furfural may become completely saturated and 1,2-PeD is formed. However, lower H2 pressures would allow for partial ringhydrogenation, and push the selectivity towards 1,5-PeD. Similar to furfural, the hydrogenation reactions of FOL are endothermic, with the exception of the γ position; however, owing to the hydrogenation barrier being higher than the ring opening barriers, hydrogenation of the γ position is unlikely prior to ring opening. Furthermore, hydrogenation at the β and ω positions features low reaction barriers; this, coupled with low reverse barriers, indicate that these reactions are more reversible than the ring opening reactions. Consequently, the ring opened intermediates would ultimately dominate. Finally, hydrogenation of the δ position carries a barrier that is higher than the ring opening barriers for FOL, resulting in ring opening dominating. Thus, in contrast to furfural, FOL is unlikely to undergo hydrogenation prior to ring opening.

ACS Paragon Plus Environment

11

ACS Catalysis

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

Page 12 of 22

Figure 5. HREELS and DFT spectra for the various heterocyclics and reaction intermediates. (a) HREELS of furfural, (b) HREELS of FOL, (c) HREELS of THFA, (d) DFT vibrational spectra for various reaction intermediates derived from furfural, (e) The HREELS and DFT spectra for terminal and secondary diols. Here, “RO” denotes a species where the heterocyclic C-O bond was cleaved, and “furfural+Hω” refers to the reactant in Figure 4.

ACS Paragon Plus Environment

12

Page 13 of 22

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

ACS Catalysis

In order to assess the DFT results experimentally, HREELS spectra of heterocyclics were taken following adsorption on the Ir(111) surface with 0.5 ML of pre-dosed hydrogen (Figure 5) at increasing temperature, and compared to the DFT spectra (Figure 5d and 5e). Furfural and FOL show a dual peak feature at ~2950 cm–1 and ~3050 cm–1, which is red shifted to a single peak feature at ~2900 cm–1 as the temperature is increased from 200 K to 400 K. Analysis of the DFT vibrational modes show the ~3050 cm–1 peak arises from the aromatic sp2 C–H stretching modes and the ~2950 cm–1 peak arises from the C–H stretching mode(s) of the aromatic ring side group. The THFA spectra in Figure 5c shows only a single peak at ~2900 cm–1, indicating the redshift is due to a change in the aromaticity of the ring. Furthermore, a peak at ~1180 cm–1 is present. Examination of the DFT vibrational modes shows this feature to be due to the bending mode of the furan ring, in agreement with prior studies27-28. The disappearance of this peak between 200 and 240 K indicates the ring opening of the unsaturated heterocyclics. The persistence of this peak above 300 K for THFA shows the opening of the saturated ring occurs at least 100 K higher in temperature when compared to the unsaturated heterocyclics, supporting our earlier DFT results that saturation of the aromatic ring hinders the hydrogenolysis kinetics. In addition, the similarity in the HREELS spectra for furfural and FOL indicates the side group is capable of undergoing either reduction or oxidation. Finally, a peak is observed at ~1900 cm-1. Prior HREELS studies on oxygenated hydrocarbons have detected a similar peak and attributed it to the presence of background CO that is adsorbed on the metal surface.29-30 At temperatures above 400 K the adsorbed species begin decomposing, producing additional adsorbed CO on the surface. The observed decrease in the CO peak at 500 K is due to CO desorption from the Ir(111) surface. In contrast to recent studies on Ni(111) and Pt(111), no stable hydrocarbon species are present on Ir(111)31-32.

ACS Paragon Plus Environment

13

ACS Catalysis

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

Page 14 of 22

Examination of the vibrational modes reveals the hydrogenolysis of the aromatic ring results in a red shift of the dual peak, in agreement with the HREELS. In Figure 5d the Cω–O ring opened intermediate shows a vibrational feature at ~1600 cm–1, with the HREELS spectra in Figures 5a and 5b showing a small peak at this frequency starting at 400 K, indicating the possible formation of 1,2-PeD. For the Cβ–O ring opening intermediate, the 3000 cm–1 peak is broader in nature than the Cω–O intermediate, with an increase in the H–C–C–H stretching modes at ~1400 cm–1, in agreement with the spectra in Figure 5a and 5b. In addition, the Cβ–O ring opened intermediate shows a peak at ~1285 cm-1 that is present in the HREELS. Owing to the lack of an easily acquired partially saturated species as an experimental control, the DFT spectrum of the partially saturated species were computed (Figure 5d) and compared to the HREELS spectra. Figure 5d reveals the partial saturation of the aromatic ring in furfural at the ω position results in an increase in intensity of the vibrational mode at ~1400 cm–1; however, the two-peak feature centered at ~3000 cm–1 still exists. Subsequent opening of the furan ring towards 1,5-PeD, results in the characteristic singular ~2900 cm–1 peak and three broad sub-1400 cm–1 peaks, in line with observations of the HREELS spectra at 300 K–400 K. Combining this vibrational analysis with the overall irreversibility of the hydrogenolysis of furfural following partial ring saturation (see Figure 4) indicates that 1,5-PeD is being formed on the Ir(111) surface.

ACS Paragon Plus Environment

14

Page 15 of 22

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

ACS Catalysis

Figure 6 DFT spectra for 1,5-PeD and its alkoxy and dioxy forms, and the HREELS spectra for furfural and 1,5-PeD. In order to confirm the presence of 1,5-PeD, the vibrational modes are calculated and compared to the HREELS spectra (Figure 5e). Overall, good agreement is achieved between the DFT and the HREELS spectrum at 200 K, with the sp3 C–H stretching modes being red-shifted due to the presence of neighboring diols. At 400 K the HREELS spectrum obtained for 1,5-PeD qualitatively matches the spectra of furfural and FOL, which further supports the production of 1,5-PeD on the Ir(111) surface. However, DFT predicts the existence of an O–H stretching mode at ~3600 cm–1 that is absent in the HREELS at 400 K. Examination of the 1,5-PeD spectrum at 120 K reveals the existence of the ν(O-H) peak at ~3200 cm-1; however this peak is red-shifted from the predicted DFT value, which is ascribed to the hydrogen bonding network33 formed due to the –OH groups of neighboring 1,5-PeD present at this temperature. Increasing the temperature to 200 K (Figure 5e) results in this peak disappearing, suggesting the formation of a dioxy intermediate (Figure 6). In agreement with the HREELS spectrum of 1,5-PeD at 400 K, the DFT spectrum of the dioxy form of 1,5-PeD shows a decrease in the CH2 vibrational mode.

ACS Paragon Plus Environment

15

ACS Catalysis

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

Page 16 of 22

While the de-protonation of the alcohol groups is predicted to be endothermic (Erxn=~+0.35 eV), enough thermal energy is present at these temperatures to drive the reaction equilibrium towards the formation of a dioxy species. Conclusions In the current study, we combined DFT and HREELS in order to understand the hydrogenolysis of biomass-derived heterocyclic compounds (furfural, FOL, and THFA) on the Ir(111) surface. Our DFT results reveal the absence of the saturation of the aromatic ring allows the ring opening of the heterocyclics to be energetically favorable. The degree of selectivity control towards terminal or secondary diols is modest, with the selectivity being determined via a balance of the kinetic and thermodynamic factors. In addition, the degree of aromatic ring saturation is critical, as the full saturation results in slower kinetics. However, partial saturation results in an increase of the selectivity towards terminal diols (i.e., 1,5-PeD). Furthermore, the reduction of the aromatic ring side group was found to be facile and easily reversible; as a consequence, aromatics with different side groups result. As the selectivity can be shifted via the electronegativity of the aromatic ring side group, in addition to the degree of ring saturation, the hydrogen coverage and temperature would play a key role in determining if the hydrogenolysis of the heterocyclic molecules results in terminal or secondary diols. Analysis of the HREELS and DFT vibrational features at 3000 cm–1 and below 1400 cm– 1

, show as the temperature increases from 200 K to 500 K, the vibrational signatures of the

heterocyclics considered resemble those of the terminal diol 1,5-PeD, indicating its formation. The disappearance of the vibrational mode associated with the ring indicate that the unsaturated heterocyclics undergo hydrogenolysis between 240 K and 300 K; however, this feature persists

ACS Paragon Plus Environment

16

Page 17 of 22

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

ACS Catalysis

above 300 K for THFA, supporting our earlier claim of ring saturation resulting in slower kinetics. Similarities in the HREELS spectra for our unsaturated heterocyclics indicate the side group is capable of being either reduced or oxidized. In conclusion, our detailed study resolves the apparent contradiction of the effect of ring saturation on the hydrogenolysis of aromatic molecules. Consistent with recent reactor experiments22, complete ring saturation hinders the hydrogenolysis; however it is facilitated via partial saturation. Our findings elucidate the key selectivity controlling factors for hydrogenolysis of heterocyclic molecules to functionalized long-chain molecules, and may help in developing better catalysts for the establishment of renewable chemicals from abundant lignocellulosic biomass. Materials and Methods Computational: All DFT calculations were performed with the VASP software package.34-37 The core electrons were represented with the PAW formalism38-39, while the valence region was represented with the PBE exchange-correlation functional40. Total energies were minimized selfconsistently using a convergence threshold of 10–6 eV. The nuclear degrees of freedom were optimized to a force convergence threshold of 0.05 eV Å–1. The reaction barriers were calculated using the dimer method41-44, with a 5 image NEB45 calculation as an initial guess. The frequency calculations utilized a tighter SCF convergence of 10–8 eV and a finer (by 30%) integration grid. The DFT spectra reported in Figures 5 and 6 were generated by smearing the vibrational energies utilizing a Gaussian distribution with a width of kB(120 K). Crystal orbital overlap population (COOP)46 analysis, which weighs the density of states with a Mullikan factor,

ACS Paragon Plus Environment

17

ACS Catalysis

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

COOP(ε ) = 2∑∑ ci*c j S ij δ (ε − ε n (k )) n

k

Page 18 of 22

(1) ,

was employed. Here ci/j are the coefficients of atom i/j, Sij is the overlap matrix between atoms i and j, and δ(ε– ε n(k)) is the delta function responsible for the density of states. The factor of

ci*c j Sij

is negative for an anti-bonding, positive for a bonding, and zero for a non-bonding

interaction46. The coefficients and overlap matrix required for Equation 1 were taken from the GPAW code47 utilizing a double zeta plus polarization (DZP) linear combination of atomic orbitals (LCAO) basis set48. The natural bond order (NBO) calculations were performed with Gaussian0949, utilizing a cc-pVDZ basis set50 and the PBE exchange correlation functional. The surface contains four layers of metal atoms and is repeated in the (x, y) directions four times (for a total of 64 iridium atoms per unit cell). To minimize interactions with the mirror in the direction normal to the surface, 10 Å of vacuum was applied. The bottom two layers of the slab were held fixed in their bulk positions, whilst the top two layers were allowed to relax. As hydrogen is present in the experiments, two hydrogen atoms per unit cell are added in the DFT calculations. Experimental: The Ir(111) single crystal was purchased from Princeton Scientific Corp. with a purity of 99.99%, and is 2 mm thick with a diameter of 10 mm. Two tantalum posts were spotwelded onto the back of the crystal which served for resistant heating and liquid nitrogen cooling. A K-type thermocouple was spot-welded onto the back of the crystal for measuring the surface temperature. Cycles of Ne+ sputtering and annealing at 1000 K were used to clean the Ir(111) surface. Oxygen treatment was used for removing extra surface carbon, wherein the crystal was heated to 900 K and dosed with 10 L (1 L=1x10-6 Torr second) of O2. The crystal

ACS Paragon Plus Environment

18

Page 19 of 22

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

ACS Catalysis

was subsequently annealed at 1000K for one minute. Auger electron spectroscopy (AES) was employed and detected no surface oxygen species. Furfural, FOL, THFA, and 1,5-PeD were purchased from Sigma Aldrich with purities higher than 99% and were purified through freeze-pump-thaw cycles before use. The gas samples (H2 and Ne) were purchased from Airgas, Inc. with research purity (99.999%) and were used subsequently without further purification. The purity of all samples was checked using mass spectrometry. The HREELS experiments were performed in a UHV chamber with a base pressure of ~8x10–10 Torr. The Ir(111) surface was annealed prior to raising the temperature and was subsequently cooled to 120 K. Prior to dosing the surface with the heterocyclic samples, the Ir(111) crystal was pre-dosed with hydrogen to achieve a surface coverage of 0.5 ML of atomic hydrogen. AUTHOR INFORMATION Corresponding Author *Email: [email protected]; [email protected] Author Contributions §These

authors contributed equally.

Funding Sources This work was funded through the Catalysis Center for Energy Innovation (CCEI), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Award No. DE-SC0001004. Notes The authors declare no competing interests.

ACS Paragon Plus Environment

19

ACS Catalysis

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

Page 20 of 22

Supporting Information Available ASE database (ASE-DB) file containing all the initial, final, transition, and adsorption state geometries, instructions on accessing this data, numerical data for Figure 2, and pictorial representations of the geometries in the ASE-DB file. This material is available free of charge via the Internet at http://pubs.acs.org. ACKNOWLEDGMENTS We acknowledge support from the Catalysis Center for Energy Innovation, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Award number DE-SC0001004. We would like to acknowledge Geun Ho Gu and Prof. Bingjun Xu for useful conversations. The authors also acknowledge the resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC0205CH11231. This research was supported in part through the use of Information Technologies resources at the University of Delaware, specifically the high-performance computing resources. REFERENCES 1. Dutta, S.; De, S.; Saha, B.; Alam, M. I., Catal. Sci. Tech. 2012, 2, 2025-2036. 2. Lange, J.-P.; van der Heide, E.; van Buijtenen, J.; Price, R., ChemSusChem 2012, 5, 150166. 3. Vennestrøm, P. N. R.; Osmundsen, C. M.; Christensen, C. H.; Taarning, E., Angew. Chem. 2011, 50, 10502-10509. 4. Gilkey, M. J.; Xu, B., ACS Catal. 2016, 6, 1420-1436. 5. Schlaf, M., Dalton Trans. 2006, 4645-4653. 6. Chen, K.; Mori, K.; Watanabe, H.; Nakagawa, Y.; Tomishige, K., J. Catal. 2012, 294, 171-183. 7. Ziaei-Azad, H.; Semagina, N., ChemCatChem 2014, 6, 885-894. 8. Shen, J.; Semagina, N., ACS Catal. 2014, 4, 268-279. 9. Nakagawa, Y.; Tamura, M.; Tomishige, K., ACS Catal. 2013, 3 (12), 2655-2668. 10. Mizugaki, T.; Yamakawa, T.; Nagatsu, Y.; Maeno, Z.; Mitsudome, T.; Jitsukawa, K.; Kaneda, K., ACS Sus. Chem. Eng. 2014, 2, 2243-2247.

ACS Paragon Plus Environment

20

Page 21 of 22

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

ACS Catalysis

11. Corma, A.; de la Torre, O.; Renz, M.; Villandier, N., Angew. Chem. 2011, 50, 2375-2378. 12. Werle, P.; Morawietz, M.; Lundmark, S.; Sörensen, K.; Karvinen, E.; Lehtonen, J., Alcohol's, Polyhydric. In Ullman's Fine Chemicals, Elvers, B., Ed. Wiley-VCH: Weinheim, Germany, 2014; Vol. 1, pp 37-58. 13. Liu, S.; Amada, Y.; Tamura, M.; Nakagawa, Y.; Tomishige, K., Green Chem. 2014, 16, 617-626. 14. Nakagawa, Y.; Tomishige, K., Catal. Today 2012, 195, 136-143. 15. Hambrock, K. O.; Robertson, J. A., Production of 1, 5-pentanediol from furfural. US Patent 2,768,979: 1956. 16. Kaufman, D.; Reeve, W., Org. Synth. 1946, 26, 83-84. 17. Corma, A.; Torre, O. d. l.; Renz, M., ChemSusChem 2011, 4 (11), 1574–1577. 18. Corma, A.; de la Torre, O.; Renz, M., Energy & Env. Sci. 2012, 5 (4), 6328-6344. 19. Corma, A.; de la Torre, O.; Renz, M.; Villandier, N., Angew. Chem. Int. Ed. 2011, 50 (10), 2375-2378. 20. Li, G.; Li, N.; Wang, Z.; Li, C.; Wang, A.; Wang, X.; Cong, Y.; Zhang, T., ChemSusChem 2012, 5 (10), 1958–1966. 21. Subrahmanyam, A. V.; Thayumanavan, S.; Huber, G. W., ChemSusChem 2010, 3 (10), 1158-1161. 22. Sutton, A. D.; Waldie, F. D.; Wu, R. L.; Schlaf, M.; Silks, L. A.; Gordon, J. C., Nat. Chem. 2013, 5 (5), 428-432. 23. Nakagawa, Y.; Takada, K.; Tamura, M.; Tomishige, K., ACS Catal. 2014, 4, 2718-2726. 24. Wang, S.; Vorotnikov, V.; Vlachos, D. G., Green Chem. 2014, 16, 736-747. 25. Pang, S. H.; Medlin, J. W., ACS Catal. 2011, 1, 1272-1283. 26. Blyholder, G., J. Phys. Chem. 1964, 68, 2772-2777. 27. Pallassana, V.; Neurock, M.; Coulston, G., Catal.Today 1999, 50 (3–4), 589-601. 28. Yan, F. Q.; Qiao, M. H.; Wei, X. M.; Liu, Q. P.; Deng, J. F.; Xu, G. Q., J. Chem. Phys. 1999, 111, 8068-8076. 29. Menning, C. A.; Hwu, H. H.; Chen, J., J. Phys. Chem. B 2006, 15471-15477. 30. Skoplyak, O.; Menning, C. A.; Barteau, M. A.; Chen, J. G., J. Chem. Phys. 2007, 127, 114707-1 - 114707-11. 31. Shi, D.; Vohs, J. M., ACS Catal. 2015, 5, 2177-2183. 32. Xiong, K.; Wan, W.; Chen, J. G., Surf. Sci. 2016. 33. Skoplyak, O.; Barteau, M. A.; Chen, J. G., Surf. Sci. 2008, 602, 3578-3587. 34. Kresse, G.; Hafner, J., Phys. Rev. B 1993, 47, 558-561. 35. Kresse, G.; Hafner, J., Phys. Rev. B 1994, 49, 14251-14269. 36. Kresse, G.; Furthmüller, J., Comput. Mat. Sci. 1996, 6, 15-50. 37. Kresse, G.; Furthmüller, J., Phys. Rev. B 1996, 54, 11169-11186. 38. Blöchl, P. E., Phys. Rev. B 1994, 50, 17953-17979. 39. Kresse, G., Phys. Rev. B 1999, 59, 1758-1775. 40. Perdew, J. P.; Burke, K.; Ernzerhof, M., Phys. Rev. Lett. 1996, 77, 3865-3868. 41. Henkelman, G.; Jónsson, H.; Jónsson, H., J. Chem. Phys. 1999, 111, 7010-7022. 42. Olsen, R. a.; Kroes, G. J.; Henkelman, G.; Arnaldsson, A.; Jónsson, H., J. Chem. Phys. 2004, 121, 9776-9792. 43. Heyden, A.; Bell, A. T.; Keil, F. J., J. Chem. Phys. 2005, 123, 224101-1 - 224101-14. 44. Kästner, J.; Sherwood, P., J. Chem. Phys. 2008, 128, 014106-1 - 014106-6. 45. Henkelman, G.; Jónsson, H., J. Chem. Phys. 2000, 113, 9978-9985.

ACS Paragon Plus Environment

21

ACS Catalysis

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

Page 22 of 22

46. Hughbanks, T.; Hoffmann, R., J. Am. Chem. Soc. 1983, 105, 3528-3537. 47. Enkovaara, J.; Rostgaard, C.; Mortensen, J. J.; Chen, J.; Dułak, M.; Ferrighi, L.; Gavnholt, J.; Glinsvad, C.; Haikola, V.; Hansen, H. a.; Kristoffersen, H. H.; Kuisma, M.; Larsen, a. H.; Lehtovaara, L.; Ljungberg, M.; Lopez-Acevedo, O.; Moses, P. G.; Ojanen, J.; Olsen, T.; Petzold, V.; Romero, N. a.; Stausholm-Møller, J.; Strange, M.; Tritsaris, G. a.; Vanin, M.; Walter, M.; Hammer, B.; Häkkinen, H.; Madsen, G. K. H.; Nieminen, R. M.; Nørskov, J. K.; Puska, M.; Rantala, T. T.; Schiøtz, J.; Thygesen, K. S.; Jacobsen, K. W., J. Phys. 2010, 22, 253202-253226. 48. Larsen, a. H.; Vanin, M.; Mortensen, J. J.; Thygesen, K. S.; Jacobsen, K. W., Phys. Rev. B 2009, 80, 195112-1 - 195112-10. 49. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J., Gaussian 09 Revision D.01. Gaussian Inc. Wallingford CT 2009. 50. Thom H. Dunning, J., J. Chem. Phys. 1989, 90, 1007-1023.

Table of Contents Graphic

ACS Paragon Plus Environment

22