[djvan]

Hello again!  I'm reviewing alcohols and the likelihood that they participate in reactions as acids.  The book suggests that everything makes more sense if you look at the stability of the conjugate base of an alcohol.  Weaker conjugate bases (cb) correspond to stronger acids.  However, they go on to say that methyl groups are electron donating, hence a cb of a tertiary alcohol is less stable than a cb of a primary alcohol, and nearly all are weaker acids than water (exception phenol).

However, if this is all about stability of the conjugate base, how is hydroxide (conjugate base of water) more stable than a tertiary alcohol?  Carbon is more electronegative than hydrogen, so shouldn't it help stabilize the conjugate base more than a hydrogen atom?   

[rwiew]

Because, as you said, alkyl groups are electron donating. The electronegativity of carbon being only slightly higher than that of H is one thing, but you need to consider hyperconjugative contribution of alkyls as well. There's also other considerations that go into this, for example the stabilization of the base that can be offered by hydrogen bonding - hydroxyls can't be stabilized by H-bonding to other hydroxyls and H₂O molecules, that can't happen between alkoxides. The alkyls also will have an entropic effect on the whole deprotonation equilibrium, increasing the pKa. There's a number of good texts dealing with all of this out there if you want more of a read. Basically, the statement "alkyl groups are electron donating" is a misleading old postulate, which is not always true and has a lot of different thermodynamic contributions going into it.

[orgopete]

Welcome to organic chemistry, the area of chemistry that encounters the paradoxes of electronegativity possibly more than any other area. If we were discussing this in street language, we might call this a flimflam. Usually carbon being referred to as an electron donor a few chapters away from it being an electron withdrawer. Do not try to maintain this chapter one topic in chapters three, four, or five, or anywhere except in chapter one. Electronegativity theory is so good it has to be true! In order to maintain this theory, you will encounter all kinds of rationales to explain why fluoride is the most electron withdrawing and yet the least electron withdrawing of the halogens, or carbon is more electronegative than hydrogen yet more electron donating. Is this sounding familiar?

My best suggestion is just adjust to it. Iodide is the best leaving group and fluoride the most electronegative. You will not overturn the paradoxes of electronegativity. Just use whatever rule or tool your book suggests, it will be on the test.

By the way, the pKa of water is listed as 15.7, methanol 15.5, isopropanol 16.5, and t-butanol 17 (Evans pKa table). I do not find these easy to explain.

The above is just my advice to the poster. If anyone feels the need to argue electronegativity theory, just start a new thread and I'll jump in. If anyone has an example of fluorine being the most electron withdrawing element, please find the discussion under that topic and post it there. I am always looking for an example.

[rwiew]

Well said orgopete! The piece about learning whatever the books says because it will be on the test is a sad reality indeed...

[AromaticAcrobatic]

rweiw nailed it.
I think of it kind of like adding acids to epoxides before nucleophiles.. Because a tertiary carbon has more carbons around to hyper conjugate electron density to it, Oxygen can induce harder in turn making the bond more partial ionic.
Ok, so then thinking about the size of oxygen, although an electronegative atom, is still pretty small. So while he wants to get to the noble state, he can't handle it as well because he's inducing more electron density from that tertiary Carbon.

I also agree with Orgo though.
But if you can get a grasp on it, it becomes a beautiful science..
 
 

[djvan]

So, I think I've found a way, using electronegativity, to make sense of this; hopefully this holds true.

Before, when I stated that carbon was more electronegative than hydrogen, and therefore should be electron withdrawing in comparison, I didn't take into account the fact that carbon atom is not alone as a methyl group!

Since, as a methyl group, carbon is bound to 3 hydrogens, you can expect it to have a partial negative charge.  This is because the carbon holds each of the sigma bond electrons closer to itself than the hydrogen bond.  Since it already has a partial negative charge, it can't easily take on more electrons, and therefore it makes a better electron donating group.

That makes most sense to me.  Does anyone know if it'll hold up?

[orgopete]

@djvan
Re: hoping electronegativity holds true

Electronegativity will not hold true, that is why you will find the many explanations trying to overcome the paradoxes. I probably cannot succeed in explaining electronegativity in a short post, but Pauling was trying to offer an explanation for the energy difference between reactants and products. He proposed bonds had ionic and covalent properties and the energy difference between elements with covalent bonds and their products must be due to the products having ionic character in their bonds. Unfortunately, this premise is not correct, but it is also complicated to realize this.

I see you have another post asking about thermochemistry. We need to use it here as well. Typically, the heats of reactions are measured from their elements and the elements are considered as starting at zero energy. Therefore, the formation of hydrogen fluoride releases energy in its formation, so its bonds stronger than the elements. However, stronger is a relative term. If the reactions are assumed to give products in which the products are more stable and hence at zero energy, then you would conclude fluorine is the most reactive element. This is one part of the puzzle.

If you look at the energy of a Born-Haber cycle, the ionization of a lithium or sodium is endothermic. By difference, the lattice energy is very large. This would be consistent with a strong attraction of the ions for one another. However, we must remember this is a gas phase reaction. If you add sodium chloride to water (or ice), the heat of solvation is small. Water must also break the bonds in that process. So we might ask whether lattice energy is truly telling us how strong a bond is. If you subtract the solvation energy from the heat of formation, and assume the reaction of sodium and chlorine were performed in water, then by Hess's Law, the redox energy should balance the energy difference. Now, you cannot perform the redox reaction in water because lithium or sodium will react with water in a very exothermic reaction rather than reacting with chlorine. I was not able to find the energetics of the reaction of either metal with water except that it is very high energy. This is another part of the puzzle.

You may find many that mistake homolytic bond energies in trying to estimate heterolytic bond strengths. They are different. However, the assumption that bonds have ionic contributions leaves one with predicting that compounds like sodium chloride or lithium fluoride have strong bonds because the energy differences are the greatest, makes ionic contributions to be the greatest contributor. If so, ions should be more attractive than ions to neutrals compounds. If ions had a strong attraction for each other, NaOH, HCl, and NaCl should have the same properties. They do not. Hydroxide is more strongly attracted to the protons of neutral water than to sodium ions. HCl is more strongly attracted to the electrons of water than to ionic chlorides. I argue acidity is virtually the best measure of heterolytic bond strength. If you dissolve NaCl in water, I don't know the ratio of [NaCl] to [Na+] + [Cl-]. If you dissolve an acid in water, the pH tells you the ratio of ionized to nonionized. HI is a strongly ionic in water and HF is not.

The metal hydrides are notorious because their bonds are weaker than predicted by the energy of the reactants. This in addition to all of the other reasons should have killed Pauling's theory of ionic attraction. I found none of this good science, except as a theory too good to not be true. It is in virtually every chemistry book.

Now that you are in organic chemistry and you go into the lab, electronegativity breaks down. Iodide is a much better leaving group than fluoride, HF is a weak acid, carbon is an electron donor, etc. I taught it too. I used all of the arguments in the later chapters that contradicted electronegativity theory. I did one other thing, I read Pauling's papers. In my classes, I told students that electron withdrawing order was I>Br>Cl>>F, S>O, H>C. Electron withdrawing was what they would find in the lab and electronegativity was what they would find on the tests.

I applaud you for looking at the pKa tables for acids and trying to discover what the data teaches you. This is the correct way to discover chemical properties. I have learned much by looking at reaction rates, products, etc. It is possible that we may draw incorrect conclusions, but we must be careful. We do not want to believe the bird is sleeping in the Monty Python skit. 

[AromaticAcrobatic]

In Organic Chemistry there are 3 important things to consider:
Resonance
Induction
Salvation effects (steric effects)

Resonance has priority over induction. As an example consider Aromatic electrophilic addition of halogenation. The halogens are electron withdrawing and yet they are ortho/para directors, because resonance has priority over induction.
So now applying this way of thinking as to why the further you go down the halogen column, the better the leaving group and in turn the further down the column of the sp hybridized atoms the more nucleophilic they become.  What changes? Well, the further you go down the column the more D character the atom accumulates, and the more D character allows more room for the electrons to move around and as we know resonance is the key to stability and resonance is the ability of a charge or electrons to move around. Therefore that extra "space" the heavier sp hybridize atoms have allows them to hold on to the charge better in turn making them good leaving groups/ strong nucleophiles, instead of a strong nucleophile/strong base as hydroxide and alkoxide ions are.
Not that electronegativity is wrong, there is just more to consider then solely the electronegativity of an atom.

 

[rwiew]

Now that you are in organic chemistry and you go into the lab, electronegativity breaks down. Iodide is a much better leaving group than fluoride, HF is a weak acid, carbon is an electron donor, etc. I taught it too. I used all of the arguments in the later chapters that contradicted electronegativity theory. I did one other thing, I read Pauling's papers. In my classes, I told students that electron withdrawing order was I>Br>Cl>>F, S>O, H>C. Electron withdrawing was what they would find in the lab and electronegativity was what they would find on the tests.

I think you go too far here by implying that electronegativity is used to explain things that I have never seen it used to explain (at least not in my education). Iodide is a much better leaving group and it it less electronegative than fluoride, yes. That's fine, bond strengths, solvation energies etc. have to be involved. You mention those in your post too, but you seem to be saying that people forget about those and just use electronegativity. I don't think that's true, you must always look at the sum of factors and most people know that. How do you justify that electron withdrawing order (let's forget about the C/H case for now)? I mean, there's simple experimental data to show this is simply wrong - NMR shifts for example. It's not a dubious qualitative theory, but quantum calculated phenomena, clearly the notion that iodide is more electron withdrawing than fluoride is wrong. I think you did a very dangerous things of putting a number of electronic and thermodynamic terms under the umbrella of "electron withdrawal" because Pauling was not quite right in his discussions a number of years ago.

[Babcock_Hall]

I think it is a mistake to analyze the pK_a values in aqueous solution, without taking into account the fact that the acidity or basicity of a given series of compounds will be affected by how easily a positive or negative charge is solvated.  For example the gas phase proton affinities ammonium ions methylated to varying degrees can be explained in terms of the electron-releasing effect of methyl groups.  It would be helpful to examine the proton affinities of tert-butoxide, methoxide, etc in the gas phase.

[orgopete]

I think you go too far here by implying that electronegativity is used to explain things that I have never seen it used to explain (at least not in my education). Iodide is a much better leaving group and it it less electronegative than fluoride, yes. That's fine, bond strengths, solvation energies etc. have to be involved. You mention those in your post too, but you seem to be saying that people forget about those and just use electronegativity. I don't think that's true, you must always look at the sum of factors and most people know that. How do you justify that electron withdrawing order (let's forget about the C/H case for now)? I mean, there's simple experimental data to show this is simply wrong - NMR shifts for example. It's not a dubious qualitative theory, but quantum calculated phenomena, clearly the notion that iodide is more electron withdrawing than fluoride is wrong. I think you did a very dangerous things of putting a number of electronic and thermodynamic terms under the umbrella of "electron withdrawal" because Pauling was not quite right in his discussions a number of years ago.

Re: electron withdrawing order

Compare the acidity of H2O and HF. Since fluorine is more electron withdrawing, it is the stronger acid. Now compare HI and HF. Which is more electron withdrawing?

When you compared H2O and HF, you compared atoms with the same numbers of protons and electrons. It was reasonable that as the nuclear charge was increased, the electrons were pulled closer to the nucleus and away from the proton, so it could be a stronger acid. This also increased the nucleus to proton repulsive force that complemented the increase in acidity.

I argue this is instructive. The acidity is the result of these subatomic forces and subject to the inverse square law. Let's see what happens if we apply the same reasoning to HF and HI. The bond length of HF is 92 pm and it has a nuclear charge of +9. The bond length of HI is 161 pm and it has a nuclear charge of +53. At the same distance, the repelling force would be 53/9 times greater, but the bond length is greater. If the bond length of HF were twice its current value, the repelling force would be 1/4. At what bond length would a proton experience the same repelling force from the nucleus in HF and HI?

You don't actually have to do the calculation to predict the result. The charge is more than four times greater and the bond length is less than twice as large. Iodine is going to pull its electrons in with a greater force and repel its proton with a greater force. It should not be a surprise that HI is a stronger acid. Not only that, but all of the additional properties coincide with this calculation, solvolysis rates, the haloform acidities, the Hammett sigma values, etc. I welcome anyone giving examples of fluorine being more electron withdrawing because I'm having a difficult time finding them.

Re: theory of ionic attraction
I had pointed out this theory suffers in comparing the properties of HCl in water. The theory should predict that the strong Coulombic attraction should be between Cl(-) and H3O(+) or why might ammonia be more basic than fluoride? The bond lengths of NH4(+) and HF can be compared. You can perform the same calculations as I have suggested. If you do this, you will discover fluoride, with a greater nuclear charge and shorter bond length, has a circa 1.3 times greater repelling force. Even though ammonia has more protons, they are at a greater distance.

[orgopete]

I think it is a mistake to analyze the pK_a values in aqueous solution, without taking into account the fact that the acidity or basicity of a given series of compounds will be affected by how easily a positive or negative charge is solvated.  For example the gas phase proton affinities ammonium ions methylated to varying degrees can be explained in terms of the electron-releasing effect of methyl groups.  It would be helpful to examine the proton affinities of tert-butoxide, methoxide, etc in the gas phase.

Although these arguments may be made, I am not obligated to agree with them. As I had indicated, it is difficult to measure heterolytic bond strengths. I argue aqueous pKa values may be the single best way to do so. I can imagine that homolyzing a bond in the gas phase is simply different than solution. One cannot use bond dissociation energy to estimate the heterolytic bond strength. I don't agree that it actually measures the heterolytic force.

If you are arguing the electronegativity scale is inherently a measure of homolytic bond strength, I might agree with you, but the poster was trying to understand why the pKa of water and simple alcohols differed as they did.

"Without going into detail, I am not surprised that gas phase (unimolecular) bond cleavages can be different than bimolecular reactions or reactions performed in solvents. Heating a molecule in the gas phase will not generate an unbalanced force to overcome the Coulombic attraction holding the molecule together. A collision with another molecule (solvent or otherwise) can tip the balance sufficiently to break a bond. If you were to stand a coin on end and balance a weight it, the force of the weight will not indicate the amount of force needed to tip the coin over. A very small amount of energy can create a whiff of air sufficient to topple the coin. A weight is not creating a force orthogonal to the axis of the coin while a whiff of air does, so only a small force can tip the coin. Although this example is unable to compare the molecular forces, it does represent how heat is different than the force of a collision."©

[Babcock_Hall]

orgopete,

You wrote, "I can imagine that homolyzing a bond in the gas phase is simply different than solution."  That is not what I meant.  I am strictly talking about heterolytic cleavages, in this case of ROH  RO^- + H⁺.  Here is some data on ΔH_i from Table 3.5 in Lowry and Richardson's "Mechanism and Theory" textbook, 2nd edition.
(CH₃)₃COH 373.3  strongest acid
CH₃OH                      379.2
H₂O                          390.8  weakest acid

The authors wrote, "solvation is entirely responsible for the observed order in solution."

Interestingly, the more alkyl groups there are on an amine/ammonoium ion, the more stable is the cation in the gas phase.

[AromaticAcrobatic]

^^ Yeah I just can't agree with that. Could you post a picture of the paragraph/page that this came from?

[orgopete]

The poster was suggesting he found a way in which electronegativity would predict the acidities. At this point, I feel we are nearing the, "When you are up to your neck in alligators, it is easy to forget your objective was to drain the swamp."

If you dissolve methanol and NaCl in water, very little ionization of methanol takes place and I assume virtually all of the NaCl dissociates. In thinking about this problem, it occurred to me the pKa tables could be considered as a tool to estimate dissociation. If one wished to determine the amount of ionic bonds (like NaCl), pKa was a very good method. If HF is indeed ionic, its dissociation could be found in the pH of its solutions. So I was not only baffled, but rather disbelieving that the converse should not be true, HI should have a very low pKa, yet, "It is perhaps desirable to point out that the bond type has no direct connection with ease of electrolytic dissociation in aqueous solution." (Linus Pauling)

orgopete,

You wrote, "I can imagine that homolyzing a bond in the gas phase is simply different than solution."  That is not what I meant.  I am strictly talking about heterolytic cleavages, in this case of ROH  RO^- + H⁺.  Here is some data on ΔH_i from Table 3.5 in Lowry and Richardson's "Mechanism and Theory" textbook, 2nd edition.
(CH₃)₃COH 373.3  strongest acid
CH₃OH                      379.2
H₂O                          390.8  weakest acid

The authors wrote, "solvation is entirely responsible for the observed order in solution."

Interestingly, the more alkyl groups there are on an amine/ammonoium ion, the more stable is the cation in the gas phase.

Anslyn and Dougherty, p 273
"How can we measure the acidity of an HA bond in the gas phase (it would require the heterolysis of the bond to create naked H+ and A-)? Such a reaction is quite unreasonable in the gas phase." They go on to describe how gas phase acidities are measured. I understand that some think of gas phase acidities as the gold standard. I do not. I perceive there is a parallel in how the thermochemical data is understood and handled. It is not unusual for textbooks to report the bond of HF is stronger than that of HI, but refer to homolytic bond strength. As I noted earlier, the Born-Haber cycle suggests the lattice energy (which should be similar to heterolytic bond strength) for sodium chloride is very high, yet it dissolves readily and its solvation energy is low. Here we have a gas phase reaction being used to predict solution effects. I believe the reality is the Born-Haber cycle is a gas phase redox reaction, but in solution, you would get a vastly different result. This is precisely the problem that Anslyn and Dougherty are referring to. I believe the reality is the Bober cycle is a gas phase redox reaction, but in solution, you would get a vastly different result.

I grant solvent effects are present. If we had to determine the pKa in water, DMSO, DMF, MeOH, etc, we would have different values for each. Personally, this is beyond my ability to rationalize all of these differences. Many compounds have been measured in water. It is a pretty good standard as it can interact with cations and anions. I felt that even though I may underestimate or err completely in its interactions, it still would not be a grave error as similar interactions may be thought to be present in all of the measurements.

If you look at my first post (#3), I point out the pKa's do not follow a simple path. Carbon is a better electron donor than hydrogen. This is true whether you use hyperconjugation or not. Because carbon holds its electrons more loosely, it enables a hyperconjugation effect. There is a simple continuum in which as you move across the row, the electrons are being held more and more tightly. Lithium, beryllium, and boron are electron (hydride) donors, that is, their electrons remain with their hydrides. Carbon can do either, and nitrogen, oxygen, and fluorine are acids. The electrons remain with the nucleus. You can learn this from the lab and from data.