Philosophy to Chemistry to Elucidation

Logic Puzzle #11: Pairwise Correlation Confidence … Solution

Long range 2D NMR experiments do not necessarily provide information about all the connectivities. The following structure elucidation problem set is one such example.

Based on the 1H-13C HMBC shown below, there is no evident correlation between the 1H at 5.31 ppm and the 13C at 21.1 ppm. Note the green box describes the region of interest.

The correlation between the methyl 1H at 2.15 ppm and 13C at 84.7 ppm (indicated by the purple arrow and the red bonds below) is a weak 4J coupling (also denoted as a W-coupling or M-coupling). The coupling arises from the individual methyl protons rotating and interacting through a W-relationship (or M-)with the carbon. This coupling is not evident for the methine 1H at 5.31 ppm and the 13C at 21.1 ppm.

 
Thank you Serge for your comment.

Logic Puzzle #11: Pairwise Correlation Confidence

Typical for long range 2D NMR experiments, spectral data may exhibit more than one correlation for two coupled nuclei (e.g. A to B and B to A). The pairwise correlations offer an extra degree of confidence in the interpretation.  

For the following fragment, an 1H-13C HMBC correlation exists for the 1H 2.15 ppm to 13C 84.7 ppm (represented by the purple arrow in the diagram below).

 
Is the pairwise correlation for the 1H 5.31 ppm to 13C 21.1 ppm evident?

 
A special thanks goes to Gene M. for pointing me to the data.

Logic Puzzle #10: Deciphering the Fragment Pattern using 2D NMR Data … Solution

Like any new process, it takes some practice to extract, understand and convert the information presented from a set of experimental NMR datasets into a fragment.

The only fragment that can accommodate the set of restrictions from a 1H-13C HSQC and HMBC is 2,3-dimethylbutane-1,1-diyl. The green arrows illustrate the 2-3JCH coupling responses extracted from an HMBC experiment.

 

Logic Puzzle #10: Deciphering the Fragment Pattern using 2D NMR Data

The goal of this puzzle is to conceptualize a fragment(s) from the given information.

In the following example, a set of protonated sp3 carbons were extracted from an HSQC experiment (not shown). The green arrows represent the 2-3JCH coupling responses extracted from an HMBC experiment. Based on these restrictions, what fragment(s) supports the data?

Note there is an open valence off one of the carbon atoms.

Logic Puzzle #7: Almost Missed It … Solution 2

With intense solvent signals present on a spectrum, a smaller signal(s) can easily be missed. If 2D NMR data is available, then this extra information can assist in clarifying whether a small signal(s) is obscured by larger signals.

On the 1H-13C HMBC below, the correlations for CDCl3/CHCl3 (due to 1J coupling responses and more) are more intense in comparison to the weak correlation at approximately 6.9 and 77.3 ppm. In this case, the weak correlation is attributed to a quaternary carbon obscured by a set of intense solvent signals.

 

Logic Puzzle #2: How to link 3 Fragments

The goal of this puzzle is to logically combine a set of fragments using valence and NMR information.

In this puzzle, three fragments are correlated through 2-3J coupling responses (represented by a green arrow) that were extracted from a 1H-13C HMBC data (spectrum not shown). The carbon atoms with the 13C chemical shifts displayed in blue indicate the presence of an adjacent heteroatom. Based on these criteria, what 'complete' fragment(s) supports the data and is there anything missing?

In order to accommodate these restrictions, a logical fit is to consider a trivalent atom, e.g. nitrogen.

Logic Puzzle #1: The Missing Link

A great skill to master is the capability to conceptualize a fragment or structure directly off a spectrum without resorting to paper-and-pen work. This skill is learnt through lots of practice. Whenever partial information is available, an elucidator can conjure up a mental image of possibilities and should it be required instinctively hunt for any missing data.

In the following example, a set of fragments including 13C and 1H chemical shifts and long-range coupling information were extracted from an HMBC experiment (not shown). The green arrows represent the 2-3J coupling responses between the 3 equivalent methyl groups and the carbonyl’s quaternary carbon. Based on these restrictions, what fragment(s) support the data and is there anything missing?

To accommodate these restrictions, three potential fragments, assigned A, B and C, are shown below. Fragment A can be disregarded on the basis of the carbon valence. Fragment B is not a good candidate because the CH3 chemical shifts do not support the presence of an adjacent heteroatom. Fragment C seems to be the most logical choice. However, there is a missing quaternary carbon. The next step is to re-evaluate the NMR data in search of a weak 13C signal at ~40 ppm.

Will the correct structure please stand up? … Part 2

Part 1 presented a challenge to determine an experiment to distinguish two very similar products from each other, namely 3-methyl-5-(pyridin-2-yloxy)pyridine and 5'-methyl-2H-1,3'-bipyridin-2-one. The products have identical formula weights and the LC/MS and 1H NMR are too similar to draw any conclusion from.

 

The first step is to determine what is different between the two products and then identify an experiment specifically designed to focus on that difference. The obvious difference between the two products is the position of the oxygen atom—an ester group verse a carbonyl group. An FT-IR experiment, as commented by the reader Felipe A., can be used to sort out the products.

 

Other experiments can include the use of reducing agents, 15N NMR, 1H -13C HMBC, 1D NOE, 1H-1H TOCSY, MS2, etc. Note free water, acids and sample concentration can inhibit the use of some of these experiments.

 

A 13C NMR experiment may appear to be another good choice when trying to identify a carbonyl group. However, the carbonyl is part of a conjugated system and so the 13C chemical shift is expected around 160 ppm, which also happens to be expected for the 13C chemical shift of the O-C=N group on the other product.

Correlating 1D NMR data on an HMBC

There are many challenges present in deciphering correlations on an HMBC experiment. Barring any data collection issues, most of the challenges will arise from signals that are overlapping, thus, adding an element of ambiguity.   

The 1H -13C HMBC spectrum below shows a correlation between the 13C at 71.2 ppm and the overlapping 1H signals at 2.08 (d) and 2.12 (m) ppm. There is nothing usually here, right?

A nearby correlation on the same 1H -13C HMBC spectrum (see below) shows a correlation between the 13C at 75.01 ppm and the 1H at 2.28 ppm. The splitting pattern for this correlation suggests that the previous correlation might also be splitting. If this is the case, then the correlation is centered on 71.26 and 2.10 (m) ppm, and thus, there is no correlation to the doublet at 2.08 ppm. In the end, further analysis of the overlapping multiplets between 2.00-2.15 ppm is needed to resolve the ambiguity.

Distinguishing Impurities … Part 6

Although a 1H-13C HMBC experiment may take a long time to collect with adequate signal-to-noise, it can offer an idea as to which signals belong 'together'. Generally, signals belonging to the same structure leave a trail of connectivity information via long range correlations. If the atoms belong to the same structure, then a typical HMBC pattern is as follows: proton A is correlated to carbon B, carbon B is correlated to proton C, proton C is correlated to carbon D, etc.

The 1H-13C HMBC experiment below shows a lack of long range connectivity information for the singlets at 1.68 and 2.14 ppm to any of the other signals. This leaves two possibilities: either the singlets do not pertain to the main unknown structure or the protons are separated far enough from any carbon atoms to show any correlations.