Goldilocks Catalysts: Computational Insights into the Role of the 3,3

Subscriber access provided by West Virginia University | Libraries

Article

Goldilocks Catalysts: Computational Insights Into The Role Of The 3, 3’ Substituents On The Selectivity of BINOL-derived Phosphoric Acid Catalysts Jolene P Reid, and Jonathan M. Goodman J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b02825 • Publication Date (Web): 26 May 2016 Downloaded from http://pubs.acs.org on June 5, 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.

Journal of the American Chemical Society 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

Journal of the American Chemical Society

Goldilocks Catalysts: Computational Insights Into The Role Of The 3, 3’ Substituents On The Selectivity of BINOLderived Phosphoric Acid Catalysts Jolene P. Reid and Jonathan M. Goodman* Centre for Molecular Informatics, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, United Kingdom [email protected]

ABSTRACT

BINOL-derived phosphoric acids provide effective asymmetric catalysis for many organic reactions. Catalysts based on this scaffold show a large structural diversity, especially in the 3, 3’ substituents, and little is known about the molecular requirements for high selectivity. As a result, selection of the best catalyst for a particular transformation requires a trial and error screening process, as the size of the 3, 3’ substituents is not simply related to their efficacy: the right choice is neither too large nor too small. We have developed an approach to identify and quantify structural features on the catalyst that determine selectivity. We show that the application of quantitative steric parameters (a new measure, AREA(θ), and rotation barrier) to an imine hydrogenation reaction allows the identification of catalyst features necessary for efficient stereoinduction, validated by QM/MM hybrid calculations.

ACS Paragon Plus Environment


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 The discovery of catalysts that efficiently facilitate organic transformations in a stereocontrolled fashion is central to synthetic organic chemistry. 1 Currently this process is usually dominated by empirical observations. However, if we can identify and understand general design principles, it should be possible to devise improved or novel catalysts with a high level of confidence. 2 Recent years have seen the emergence of BINOL-derived phosphoric acids as powerful catalysts for asymmetric transformations.3-5 Despite their popularity, most literature reports of reactions do not analyse the detailed origin of stereoinduction. Why are some catalysts more selective than others? Calculation of the transition states reveals key catalyst-substrate interactions that allow efficient transfer of chiral information from catalyst to substrate, assisting catalyst design.2 However, this approach is intrinsically time consuming, as subtle structural changes can have a major impact on enantioselectivity and separate calculations are required for each catalyst. An alternative approach is to discover relationships between catalyst descriptors and enantioselectivity. Such relationships should give rapid insights into the important structural features that are necessary for efficient stereoinduction. 6 - 8 Unfortunately, however, it seems that stereoinduction does not simply follow steric bulk or any other readily comprehensible measure of the catalyst. The selectivity of chiral phosphoric acid catalysed reactions is dependent on the choice of substituents in the 3, 3’ positions of the binaphthol rings: they must be neither too large, nor too small, but just right. In this paper, we present a model, validated by DFT studies, which provides insight into the role of 3, 3’ groups.


Page 2 of 22

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


RESULTS AND DISCUSSION 1. Trends and Steric Descriptors The structure of the substituents at the 3, 3’ positions have a considerable effect on the stereochemical outcome of many reactions, and, usually, large steric bulk is required for high enantioselectivity. However, excessive steric demands in these positions may stop reactions altogether, 9 or, in some cases, reverse the sense of stereoinduction.10 As a result, the optimal choice of substituent and the stereochemical outcome are extremely difficult to predict (Figure 1).

HN Ph

PMP CO2Et

85% ee

5 mol% (S)-PA-1

N

Et2O rt, 24h Ph

PMP CO2Et

EtO 2C Me

CO2Et N H

Me

5 mol% (S)-PA-2 Et2O rt, 24h

HN Ph

i-Pr

PMP CO2Et

87% ee

i-Pr

SiPh 3 i-Pr O O P O OH i-Pr

O O P O OH SiPh 3

(S)-PA-1

i-Pr

i-Pr

(S)-PA-2

Figure 1. Example reported by You et al.10b demonstrating reaction sensitivity to 3, 3’ group.

The complete steric description of a system is a demanding task. Computationally screening a large number of substrate-catalyst complexes is a computer-intensive process for establishing which properties at an atomic scale determine the enantioselectivity. Alternatively, it may also be possible to develop correlations



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

between catalyst descriptors and enantioselectivity. This method allows exploration of the molecular features affecting the transition state without making any assumptions about the mechanism. A key aspect in developing correlations is identifying appropriate parameters to connect changes in structure with selectivity. The identification of descriptors is facilitated by the observation that large steric bulk is required for high selectivity. On the basis of this we began to examine methods in which steric effects could be described numerically. We tried three different measures of steric requirements: (i) A-values – a widely used steric parameter,11 (ii) rotation barrier for a phenyl group, 12-13 (iii) A Remote Environment Angle: AREA(θ) – a measure of space less close to the phosphoric acid group (details in Figure 2 and Figure 4). (B) C-C rotation (kinetic barrier)

(A) A-value (thermodynamic difference)

(C) AREA(θ) (a measure of remote steric effects) p c

θ

p c

θ

P OOH O O

O OH P O O

Remote Sterics Nearby Sterics

p is a vector oriented from the midpoint of the naphthol oxygen atoms along the centre of phosphorous atom. c is the vector from the phosphorous atom to an atom on the 3,3' substituent. AREA(θ) is the smallest of the angles between p and all the possible c vectors.

Figure 2. Catalyst steric parameters evaluated in this study. Steric parameters (A) and (B) measure nearby bulk, (C), AREA(θ), measures steric effects distant from the phosphoric acid moiety. 3D structure of (S)-TRIP is shown in a wire frame model as an example. We used the value for AREA(θ) from the global minimum, as the value


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


did not vary by more than a degree when conformations within 10 kJ mol-1 were considered.

A-values are derived from a conformational study on the equilibrium position of ring flipping in mono-substituted cyclohexane rings. The destabilisation of the axial conformation is due to 1,3 diaxial interactions. Rotational barriers for a phenyl group are derived from the energy required for rotation around the central C-C bond. The interaction between the R groups and the hydrogens on the opposing aryl ring are responsible for the destabilisation of the eclipsed conformation. 3, 3’ Group H CHPh2 SiPh3 Ph 1-naphth 2-naphth 9-anthryl 9-phenanthryl 4-PhC6H4 4-tBuC6H4 3,5-(Ph)2C6H3 3,5-(CF3)2C6H3 2,4,6-(Me)3C6H2 2,4,6-(iPr)3C6H2

Computed Α-Value (A) 0.00 1.76 4.85 4.39 4.07 4.35 14.40 4.09 4.35 4.37 4.21 4.03 15.97 26.33

Rotation Barrier (B) 0.00 1.13 1.35 2.05 13.63 2.13 28.31 14.45 2.01 2.02 2.21 2.02 21.58 28.40

AREA(θ) (C) 107 47 29 70 62 49 61 48 50 49 36 62 61 51

Table 1. Steric parameters calculated using OPLS-2005 force field,

14 - 16

in

MacroModel (details in SI).17 All energies in kcal mol-1. Although both A-Values and rotation barriers can be derived experimentally, we have calculated them to ensure they are available and consistent for all groups of interest. Table 1 gives the approximate values of the energy difference between the equatorial position and the axial, which were calculated using the local Boltzmann-weighted energetic minimum of both the axial and equatorial conformations generated from a



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

conformational search in MacroModel. For the groups, the equatorial preference increases with size, although increasing the steric demands remote from the cyclohexane ring is unimportant – only nearby sterics are assessed by this measure (Figure 2). As we expected, A-values were connected with stereoselectivity, but, as we also expected, it is clear that they do not tell the complete story (Figure 3). The general agreement (Figure 3, circles and squares) between the parameter sets imply that they are both measuring similar properties. The steric bulk distant from the active site, is not taken into account by the A-value or the rotational barrier, is a key structural feature affecting enantioselectivity. For example, aromatic rings with hydrogen in the 2,6 positions all have similar A-values (approximately 4.2) and similar rotation barriers (about 2.0). Changing substituents in the 3, 4 and 5 positions does not affect the proximal steric effect but does change the enantioselectivity. Figure 3, therefore, shows a vertical line of points for A-values, at about 4.2, and for rotation barriers at about 2.0.18


Page 6 of 22

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


Figure 3. Proximal steric parameters evaluated in this study, literature example reported by You et al.10b

We therefore sought to develop an alternative steric parameter to describe the remote steric demands of the substituents. Conformational analysis of the phosphoric acid shows the distinctive feature that the active site is deep inside a chiral pocket. What size cone, with its point on the phosphorous atom, would fit inside this cavity? In order to devise a measure of this which is straightforward to calculate, using the global minimum we define the ligand AREA(θ) as the smallest angle (in degrees) between the centre of an atom located on the 3, 3’ group and the vector from centre of the phosphorus atom to the midpoint of the binapthol oxygens (Figure 2). Substituents which crowd access to the phosphoric acid, such as 4-tert-butyl-benzene, AREA(49), have smaller AREA(θ) values than less sterically demanding substituents, such as phenyl, AREA(70), even though the A-value is identical (Table 1). TRIP has a much larger A-value than Ph3Si (26.33 vs 4.85) but is less sterically demanding away from the phosphorus: AREA(51) vs AREA(29). Computation of this parameter is straightforward and Figure 4 shows a few examples.

c

θ c

P OOH O O

Large

p

p

p

θ P OOH O O

Medium

Figure 4. Variation in AREA(θ) with 3, 3’ substituent.


θ c POOH O O

Small


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

On comparing a series of catalysts for the Mannich reaction,19 it was found that the correlation provides a good fit to the experimental data, suggesting that it reflects an underlying physical phenomenon.20 R1

N Ph

Boc H

2 mol% (R)-PA-3 1.1 equiv acac CH2Cl2 rt, 1h

HN

O O P O OH

Boc

Ac

Ph

R1

Ac

(R)-PA-3

4-(β-Naphth)C6H4 4-PhC6H4

Ph

H

Figure 5. Ligand AREA(θ) parameter provides a linear relationship with enantioselectivity in Terada’s Mannich reaction.19

2. Insight into the stereochemical role of the 3, 3’ groups Not all reactions follow such a straightforward pattern, in which the stereoselectivity increases linearly with decreasing AREA(θ). For some reactions, the increase stops and then changes to a decline. Arguably the most interesting and mechanistically informative plots are those that are non-linear. This is shown in the transfer


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


hydrogenation reaction reported by You.10b Experimentally You et al. found that changing the 3, 3’ groups from 2,4,6-triisopropylphenyl (medium AREA(θ)) to SiPh3 (small AREA(θ)) inverted the sense of stereoinduction (Figure 6). However, it is not clear how the 3, 3’ group induces such a strong reversal. R1 O O P O OH R1

(S)-PA-4 Entry

Catalyst R1

AREA(θ)

1 2 3 4 5 6 7 8 9 10 11

H Ph 4-NO2C6H4 1-naphth 3,5-(CF3)2C6H3 9-anthryl 2,4,6-(iPr)3C6H2 4-PhC6H4 2-naphth 9-phenanthryl SiPh3

107 70 64 62 62 61 51 50 49 48 29

Rotation Barrier (kcal/mol) 0.00 2.05 2.03 13.63 2.02 28.31 28.40 2.01 2.13 14.45 1.35

ee (%)

Mechanism

-22 57 54 82 21 90 87 26

Goldilocks Catalysts: Computational Insights into the Role of the 3,3

Recommend Documents