Part 1

Carbonyl compounds

edited by Roger Hill and Peter Taylor

based on Carbonyl compounds, by Jim Iley

INTRODUCTION: CARBONYL COMPOUNDS IN CONTEXT 1

Organic compounds are conveniently subdivided into classes, based on their functional group; each member within a class reacts in the same way with a particular reagent. This Part of the Book explains the chemistry of the carbonyl group (C=O); you can see in Table 1.1 that the carbonyl group turns up in a number of functional groups. You have met many of these before, and, no doubt, have encountered them in everyday life, probably without realizing it (see Box 1.1). Carbonyl compounds are so widespread that some understanding of their behaviour is an essential requirement for every student of organic chemistry.

We'll start our examination of the topic with a close look at the structure of the carbonyl group.

THE STRUCTURE OF THE CARBONYL GROUP 2

X-ray crystallography shows that the carbonyl group is planar. The carbon and oxygen atoms of the carbonyl group, together with the two carbon atoms attached to the carbonyl carbon, all lie in the same plane, with the three bond angles all close to 120° (2.1). This implies a particular arrangement of molecular orbitals.

* What hybridization of s and p orbitals is implied at the carbon atom?

* Since the central carbon atom is bound to three other atoms, it must form three σ bonds and so it will be sp2-hybridized.

* What about the oxygen atom?

* It forms one σ bond to carbon, but has two non-bonding electron pairs, so is also sp2-hybridized.

* So both the oxygen atom and the central carbon atom contribute one electron to the σ bond between them, which leaves one over for each. How are these deployed?

* They are each contained in a p orbital perpendicular to the plane and overlapping sideways with each other, forming a π bond.

The bond between carbon and oxygen is thus a double bond, with two electrons in a σ bond and two electrons in a π bond (Figure 2.1).

The carbonyl bond is polar; that is, the four shared electrons are closer to one atom than to the other, leaving one atom electron-deficient or slightly positive (δ+), and the other with a surplus or slightly negative (δ-).

* Which atom bears the δ+ and why?

* Oxygen is more electronegative than carbon, and claims more than an equal share of the four electrons, leaving the carbon atom electron deficient. This is shown in Figure 2.2.

This polarization of C=O (2.2) is extremely useful in understanding the chemical behaviour of carbonyl compounds.

* Which of the two atoms would be prone to attack by nucleophiles?

* Nucleophiles have an electron-rich centre (they are 'nucleus-liking') and therefore seek to associate with the electron-deficient (positively charged) carbon atom.

* What is the simplest electrophile, and how would it seek to associate with a carbonyl compound?

* The hydrogen ion (proton) is the simplest electrophile, and would associate with the oxygen atom; that is, the carbonyl oxygen atom can be protonated.

You will see shortly how a large part of the chemistry of carbonyl compounds can be explained by the answers to the last two questions. But the electronic structure of a group also determines its spectroscopic properties, so this is a good place for a brief look in that direction.

* Why should you want to know about spectroscopic properties?

* The spectra (e.g. infrared and nuclear magnetic resonance) of a class of organic compounds have features common to that class. Conversely, some features of the spectra are diagnostic for that class of compound, so organic chemists use spectroscopy to tell whether or not a compound belongs to that class — that is, whether it contains the relevant group within its molecular structure.

Each of our exemplar compounds in Table 1.1 fall within the range of IR and 13C NMR values usually quoted for carbonyl groups, namely 1 650–1 850 cm-1 and 160–220 p.p.m., respectively. Details for each carbonyl functional group are given in Table 23.1 of the Data Book, which is available from the CD-ROM. We quote a range because, although the spectroscopic values differ from one functional group to another, there is a specific range for each functional group depending on the nature of the R group(s). When we use these ranges to determine the structure of an unknown compound, there is some overlap; in such cases we need to bring together all the information we have about a compound before we can be sure of the nature of the functional group.

2.1 Summary of Sections 1 and 2

1 The carbonyl group is found in the following classes of compound: aldehydes, ketones, carboxylic acids and their salts, esters, amides, acid halides and acid anhydrides.

2 The carbonyl group is planar, and the three bond angles around the central carbon atom are all close to 120°.

3 The bonding in the carbonyl group involves σ bond formation by overlap of an sp2 hybrid orbital on carbon with an sp2 hybrid orbital on oxygen, and π bond formation by overlap of a p orbital on carbon with a p orbital on oxygen.

4 The carbon–oxygen double bond is polarized such that the carbon atom is slightly positively charged, and the oxygen atom slightly negatively charged.

5 Each carbonyl functional group has a distinctive range of IR and 13C NMR parameters that can be used to help identify the nature of an unknown carbonyl compound.

QUESTION 2.1

Which of the following six compounds contain one or more carbonyl groups (see Box 2.1 for information about camphor)? Identify each carbonyl group as ester, amide, etc., and, based on Table 1.1, indicate the wavenumber that you might expect the IR stretching vibration of the group to have.

QUESTION 2.2

Complete Table 2.2 by inserting in the inference column for each of the compounds A–G either the likely class of carbonyl compound involved or 'carbonyl compound unlikely'.

STUDY NOTE

A set of interactive self-assessment questions are provided on the Mechanism and Synthesis CD-ROM. The questions are stored, and you can come back to the questions as many or as few times as you wish in order to improve your score on some or all of them. This is a good way of reinforcing the knowledge you have gained while studying this Book.

NUCLEOPHILIC ATTACK AT THE CARBONYL GROUP 3

3.1 An overview: organization through mechanism

You saw in the previous Section that we expect nucleophiles to attack the carbon atom of the carbonyl group. The outcome would be the formation of a new carbon–nucleophile bond:

[FORMULA NOT REPRODUCIBLE IN ASCII] (3.1)

In Reaction 3.1 we assume the nucleophile is a negatively charged reagent like HO-, and show it forming a new bond by donating one of its electron pairs to the carbon atom of the carbonyl group. In order to maintain an eight-electron noble gas outer shell, however, the carbon atom must simultaneously give up its hold on another electron pair. One of the two electron pairs that make up the double bond now becomes exclusively associated with the oxygen atom, and thus the oxygen becomes negatively charged.

Reaction 3.1 expresses the type of chemistry most frequently encountered with carbonyl compounds. So how does it fit with the orbital picture of the C=O bond you saw in Section 2 (Figure 2.1)? Which of the electron pairs, σ or π, ends up on the oxygen atom?

The product retains a single bond between carbon and oxygen, and as single bonds are always σ bonds, with the electron pair distributed about the internuclear axis, intuition suggests that it's the π electron pair that moves. This is correct, and the reason is that π electrons are more polarizable than a electrons; that is, being further away from the atomic nuclei they are more easily influenced by external agents — more easily 'pushed around', if you like. Important consequences follow from this involvement of the π bond. For a start, we can now begin to envisage nucleophilic attack in three dimensions as in Structure 3.1, where the broken green line indicates the trajectory of the nucleophile as it approaches the carbonyl compound.

The nucleophile will approach from above (or below) the plane of the carbonyl group and, as its pair of electrons is heading for the slightly electron-deficient carbon atom, and pushing the π electrons towards oxygen, one might expect the nucleophile to try to keep away from the oxygen and its developing negative charge. This is all borne out by experimental and theoretical studies.

Careful examination of the crystal structures of molecules that contain both a carbonyl group and a nucleophilic group (such as an amino group) shows that the nucleophile approaches the carbonyl group at an angle of about 105° to the plane containing the R1, R2 groups and the C and O atoms. In practice, this angle is seen to vary between 100° and 115° (Structure 3.2a). This is not very surprising if you consider that in the tetrahedral product this angle must be about 109°.

Not only does the nucleophile approach the carbonyl group at this angle, but it does so from a direction that lies between the R1 and R2 groups (see Structure 3.2b). Again, this is not too surprising, since this is the direction along which the Nu group lies in the tetrahedral product.

As the nucleophile gets closer to the carbonyl carbon atom, the geometry around the carbonyl group resembles more and more that of the product. So, the carbon–oxygen bond gets longer (the C=O bond length in a ketone is 120 pm, compared to a C — O bond length of 141 pm in an alcohol), and the R1 and R2 groups move out of the plane containing the carbonyl C and O atoms, to take up positions nearer to those seen in the tetrahedral arrangement in the product. (The R1 and R2 groups are shown in black in the carbonyl, but in magenta as they move to their positions in the tetrahedral product.)

[FORMULA NOT REPRODUCIBLE IN ASCII] (3.2)

This anion, however, is not the end product, so you need to consider how this may react further. 3.3 is simply the anion of an alcohol, so one possible process available to it is protonation, to form the corresponding alcohol:

[FORMULA NOT REPRODUCIBLE IN ASCII] (3.3)

This reaction preserves the tetrahedral centre in the product. However, if instead of an alkyl group, R2, one of the groups attached to the carbonyl is a good leaving group, X, an alternative process (leading to the formation of 3.4) occurs — decomposition with loss of X;

[FORMULA NOT REPRODUCIBLE IN ASCII] (3.4)

This is analogous to a β-elimination reaction to form an alkene. For a β-elimination reaction, the pair of electrons used to form the carbon–carbon π bond comes from a C — H bond, and the C — X bond breaks heterolytically; that is, the two electrons in the C — X bond both end up on the leaving group X:

[FORMULA NOT REPRODUCIBLE IN ASCII] (3.5)

In Reaction 3.4, the electrons used to form the carbon-oxygen π bond originate from a non-bonded electron pair on oxygen. Notice that this decomposition of 3.5 regenerates the planar carbonyl group, and that the nucleophile, Nu, has replaced the group X. Looking at the reaction overall, substitution of Nu for X. has occurred, but this has taken place by a mechanism involving addition of Nu followed by elimination of X.

Now, a major question arises: 'Do all carbonyl compounds undergo each of these reactions?'. Certainly, all carbonyl compounds undergo nucleophilic attack to form.3.5. However, not all of them undergo decomposition with loss of X-. You can see why with the following argument:

* For each of the following classes of carbonyl groups, identify X in the general formula RCOX: aldehyde, ketone, carboxylic acid, ester, amide, acid chloride and acid anhydride.

* aldehyde, RCHO: X = H

ketone, R1R2CO: X = R2

carboxylic acid, RCOOH: X = OH

ester, R1COOR2: X = OR2

amide, R1CONR2R3: X = NR 2R3

acid chloride, RCOC1: X = Cl

acid anhydride, R1COOCOR2: X = OCOR2

* Which of these groups, X, correspond to good leaving groups, and which are poor leaving groups?

* The acids H — H and R2 — H are very weak acids, so H- and (R2)- are very poor leaving groups. Of the rest, Cl- is a good leaving group, and R2CO2- is a reasonable leaving group. R2R3N-, R2O- and HO- are the anions of the weak acids R2R3NH, ROH and H2O, so they are poor leaving groups. However, protonation will convert 3.6 to 3.7, 3.8 to 3.9 and 3.10 to 3.11, generating the good leaving groups R2R3NH, R2OH and H2O, respectively.

The outcome of this reasoning is as follows. Aldehydes and ketones do not possess good leaving groups, whereas, under appropriate conditions, carboxylic acids and their derivatives do.

Because of this, we shall discuss the chemistry of these two groups of compounds separately. You should always bear in mind, however, that the reaction of all carbonyl compounds with nucleophiles involves the same first step, namely nucleo-philic attack at the carbonyl carbon atom to form a tetrahedral intermediate. The different scenarios are summarized in Scheme

Take a few minutes to consolidate this important distinction between aldehyde/ketone chemistry and carboxyl chemistry by attempting Question 3.1, which anticipates some reactions we'll discuss in more detail later.

QUESTION 3.1

Predict the structures of the final products when each of the carbonyl compounds CH3CHO, CH3COCH3 and CH3CO2CH3 is allowed to react with a strong nucleophile Y-, when the initial concentration of Y- is at least twice that of the carbonyl compound. Assume that the reactions are completed by the addition of H+.