Chemical shifts of phenolic monomers in solution and implications for addition and self-condensation

Article abstract of DOI:10.1002/mrc.3914

Alkali metal counter-cations alter the electron density of phenolates in solution by electrostatic interactions. This change in electron density affects their reactivity toward formaldehyde, hydroxymethylphenols, and isocyanates during polymerization. The electronic perturbation of phenolic model compounds in the presence of alkali metal hydroxides was investigated with ¹³C and ¹H nuclear magnetic resonance in polar solvents relative to non-ionic controls, altering the chemical shifts of the model compounds, thus indicating changes in electron density using the chemical shift as a proxy. These shifts were attributed to Coulombic electrostatic interactions of the counter-cation with the phenolate anion that correlated to hydrated ionic radius and solvent dielectric constants. The predicted relative reaction rates for formaldehyde addition based on electron density ranking from ¹³C nuclear magnetic resonance of the phenolic models was compared with the literature values. Predictions for condensation reactions of 2- and 4-hydroxymethylphenol from chemical shifts were consistent with published results. The results permit predictions for the reaction of phenolic compounds for the formation of thermosetting polymeric materials. Copyright

Full text of DOI:10.1002/mrc.3914

Research Article
Received: 1 June 2012
Revised: 12 November 2012
Accepted: 20 November 2012
Published online in Wiley Online Library: 20 December 2012
(wileyonlinelibrary.com) DOI 10.1002/mrc.3914
Chemical shifts of phenolic monomers in
solution and implications for addition and
self-condensation
Robert A. Haupta and Scott Renneckara,b*
Alkali metal counter-cations alter the electron density of phenolates in solution by electrostatic interactions. This change in
electron density affects their reactivity toward formaldehyde, hydroxymethylphenols, and isocyanates during polymerization.
The electronic perturbation of phenolic model compounds in the presence of alkali metal hydroxides was investigated with 13C
and 1H nuclear magnetic resonance in polar solvents relative to non-ionic controls, altering the chemical shifts of the model com-
pounds, thus indicating changes in electron density using the chemical shift as a proxy. These shifts were attributed to Coulombic
electrostatic interactions of the counter-cation with the phenolate anion that correlated to hydrated ionic radius and solvent
dielectric constants. The predicted relative reaction rates for formaldehyde addition based on electron density ranking from
13C nuclear magnetic resonance of the phenolic models was compared with the literature values. Predictions for condensation
reactions of 2- and 4-hydroxymethylphenol from chemical shifts were consistent with published results. The results permit
predictions for the reaction of phenolic compounds for the formation of thermosetting polymeric materials. Copyright © 2012
John Wiley & Sons, Ltd.
Keywords: phenol-formaldehyde; alkali metal hydroxides; phenolates; electronic behavior; electron density; hydrated ionic radius
influence of the alkali catalyst on the structure and cure rate of
the PF adhesive without explaining the underlying mechanism
Introduction
Phenolic and phenolic derivatives make up an important class
of monomers for polymeric materials used in engineering and
structural applications. Phenolic-based polymers are known for
their stiffness and durability. Furthermore, nature uses these
monomers via the radical coupled polymerization of phenylpro-
pane units in the example of lignification;[1] both stiffness and
durability are key performance properties for terrestrial plants.
In addition to radical coupling, a number of other synthetic
polymerization chemistries related to condensation pathways
have been developed over the years. The number of reactive
sites on the aromatic groups arising from the resonance
structures of these monomers, combined with the ability of
some of the compounds to form quinone methide (QM) inter-
mediates, create products with a complicated chemical struc-
ture. As a result, there remains uncertainty around the reaction
mechanism for these phenolic monomers, such as phenol-
formaldehyde (PF), in the presence of different catalysts that
alter electronic behavior.
Catalysts are important in phenolic adhesives because the
main drawback of PF technology is the requirement of relatively
high temperatures and times to cross-link. One approach to
enhance reactivity (polymerization) has been to add different
alkali metal hydroxides to the resin as a catalyst. Daisy[2] recog-
nized that PF resin cure was accelerated by using potassium
hydroxide instead of sodium hydroxide as a catalyst. Grenier-
Loustalot et al.[3] found that the use of different alkali metal
hydroxides affected the kinetic rate constants for the addition
of formaldehyde to phenol. Higuchi et al.[4,5] showed that alkali
metal hydroxides affect the composition as well as their kinetic
rate constants of hydroxymethylphenols upon the addition of
formaldehyde to phenol. These experiments suggest a strong
for the changes.
Other compounds like esters, lactones, and organic carbonates
also accelerate PF cross-linking and these mechanisms are better
understood compared with the change of alkali hydroxides.
Conner et al.[6] studied the mechanism of how these compounds
accelerate the condensation of PF model compounds, conclud-
ing that transesterification of the phenolic hydroxymethyl
functional group was the initial reaction. A subsequent reaction
likely proceeded by displacement by either a bimolecular nucle-
ophilic substitution (SN2 mechanism) of a negatively charged
ortho or para carbon or by conversion to a QM that then reacted
with the negatively charged anion (QM mechanism). Kamo
et al.[7] also concluded that the transesterification/QM mecha-
nism was critical for acceleration. Kamo and co-authors showed
how sodium bicarbonate and sodium hydroxide altered the 13C
nuclear magnetic resonance (NMR) chemical shifts of various
carbon atoms of PF model compounds in studies of organic
carbonate-facilitated PF acceleration. The work suggests that
NMR chemical shift information can be used to help elucidate
the mechanism behind the reactivity of PF.
Other additives or solvents are also known to affect the
reactivity of phenolic adhesives. Haupt[8] has demonstrated the
*
Correspondence to: Scott Renneckar, Department of Sustainable Biomaterials,
Virginia Tech, 230 Cheatham Hall (0323), Blacksburg, VA 24061, USA. E-mail:
srenneck@vt.edu
a Macromolecules and Interfaces Institute, 2 Davidson Hall, Blacksburg, VA
24061, USA
b Department of Sustainable Biomaterials, Virginia Tech, Blacksburg, VA 24061,
USA
Magn. Reson. Chem. 2013, 51, 95–101
Copyright © 2012 John Wiley & Sons, Ltd.

R. A. Haupt and S. Renneckar
accelerated copolymerization and self-condensation of PF with
polymeric methylenediphenyl diisocyanate, in the presence of
a PF accelerator and a polyurethane catalyst. Han et al. found
that a polar solvent environment affects the kinetic rates
constants for co-reaction of 2-hydroxymethylphenol (2-HMP)
and phenyl isocyanate to form urethanes, altering the reactivity
and selectivity of the phenolic hydroxyl versus the hydroxymethyl
group. Furthermore, the reactivity of the phenolic hydroxyl
group increased, whereas the reactivity of the hydroxymethyl group
decreased with increasing solvent polarity.[9] The altered reactiv-
ities of the chemistries described above may be understandable
as changes in the electron density of various functional groups
making them more or less nucleophilic or electrophilic. Electron
density is reflected by NMR chemical shift, with low electron
density deshielding the nucleus, resulting in greater chemical
shift values.[10] Thus, changes in chemical shift may be studied
as a proxy for electron density to better understand the
effect of counterions, accelerators, and solvents on phenolic
reactivity.[10] The objective of this study is to elucidate how PF
oligomers participate in addition, self-condensation, accelera-
tion, and copolymerization reactions by perturbing the elec-
tronic structure of PF model compounds with alkali metal
hydroxides and/or solvent selection. 13C and 1H NMR studies of
phenolic monomers were used to reveal the effect of alkali metal
counterions and solvent selection on the relative electron
density. These data are then compared with the reactivity of
phenolic compounds reported in the literature.
Phenol, 2-HMP, and 4-HMP treated with alkali metal
hydroxides and dissolved in D2O/CD3OD
PF model compounds phenol, 2-HMP, and 4-HMP were combined
with the alkali metal hydroxides of lithium, sodium, potassium,
rubidium, and cesium in separate glass sample vials. The target
molar ratio of alkali metal hydroxide to phenol model compound
was equal to or greater than 1.0. Moisture was removed from the
2-HMP and 4-HMP samples by placing the vials in a vacuum oven
at ambient temperature overnight. Because phenol sublimates
under vacuum, the alkali metal phenolate salts were not vacuum
dried, leaving a trace amount of residual H2O in the rubidium-
and cesium-treated NMR samples. The NMR sample solvent was a
mixture of 9.9482g CD3OD (0.2758 mol) and 20.0504 g D2O
(1.0010mol) for a molar ratio of CD3OD to D2O of 0.2755.
Phenol in dissolved in DMSO-d6 and THF-d8
Additional samples of phenol were dissolved in DMSO-d6 and
THF-d8 for NMR analysis. Samples of phenol in DMSO should be
handled carefully in a hood with proper personal protective
equipment for the skin due to the combination of the penetrating
characteristic of DMSO and the toxicity of phenol.
1H and 13C NMR
NMR spectra were acquired using a Varian Inova 400 model at
1
400 MHz proton frequency. Both H and 13C spectra were taken
for all materials studied. Chemical shift data were analyzed with
MestReNova software by Mestrelab Research S.L. All spectra,
except for those deuterium oxide 13C spectra, were adjusted to
the solvent peak reference chemical shift; the 13C spectra for
model compounds in deuterium oxide were used as is without
adjustment.
Experimental
Materials
Phenolic model compounds were phenol, 2-hydroxymethyl alcohol
(2-HMP), and 4-hydroxymethyl alcohol (4-HMP), purchased from
Sigma-Aldrich. Alkali metal hydroxides were of lithium (100%
anhydrous; Fisher), sodium (98%; Sigma-Aldrich), potassium (85%;
Acros), rubidium (50% solution in H2O; Sigma-Aldrich), and cesium
(50% solution in H2O; Sigma-Aldrich). The NMR solvents were D2O
(Sigma-Aldrich), CD3OD (Sigma-Aldrich and Cambridge Isotope
Labs), DMSO-d6 (Cambridge Isotope Labs), and THF-d8 (Cambridge
Isotope Labs). The materials were used as received.
Results
Phenol treated with alkali metal hydroxides and dissolved in
D2O or CD3OD
The phenolic model compounds and their corresponding
numbering schemes are shown in Fig. 1. The effect of ionization
to the alkali metal phenolate salts on the chemical shifts of different
carbons of the phenolic model compounds are shown in Fig. 2. For
the different carbon atoms of the alkali metal phenolate salts in
D2O or CD3OD solution in comparison with unperturbed phenol,
strong deshielding occurred at C-1 as indicated by the increase in
chemical shift, with moderate shielding at C-2 and C-6. For all the
phenolates against their phenol reference models, weak shielding
was observed at C-3 and C-5, whereas moderate shielding was
observed at C-4 (Fig. 2). The overall pattern of chemical shifts
followed principles well known from resonance structures for
aromatic rings with electronegative substituents.[10] From the shift
position, it is noted that positions 2, 4, and 6 had high electron
Preparation of NMR sample solutions
All NMR samples were dissolved in approximately 0.80 mL of
solvent in a glass vial then transferred to a 5-mm glass NMR tube.
The samples were prepared to provide at least 25 mg of model
compound for strong NMR signals
Phenol treated with alkali metal hydroxides and dissolved in
D2O or CD3OD
Phenol was accurately weighed and dissolved in D2O or CD3OD.
The phenol solutions were combined with the alkali metal
hydroxides of lithium, sodium, potassium, rubidium, and cesium
in separate glass sample vials. The target molar ratio of alkali
metal hydroxide to phenol was equal to or greater than 1.0.
Sample masses were measured to 0.1 mg accuracy to calculate
the formulated molar ratios. Two replicates were made for each
sample–solvent combination.
7
OH
7
OH
7
OH
1
8
9
6
5
2
3
1
1
CH2OH
6
5
2
3
6
5
2
3
4
CH2OH
8
4
4
9
Figure 1. NMR numbering scheme for phenol, 2-HMP, and 4-HMP.
Copyright © 2012 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2013, 51, 95–101
wileyonlinelibrary.com/journal/mrc

Electronic behavior of phenolic monomers in solution and implications for addition and self-condensation
Figure 3. 13C NMR of phenol in THF-d8, DMSO-d6, CD3OD, and D2O.
whereas the R2 correlation coefficients are progressively lower
for the other ring positions in the order C-1 > C-2,6> C-3,5> C-4.
In other words, the C1 carbon is most strongly affected by the
solvent and possibly the C-2,6 carbons to a lesser extent due to
their proximity to C-1. D2O and CD3OD interacting by hydrogen
bonding with the phenol would likely alter the electron density
of the hydroxyl oxygen; these results suggest that solvent
interactions added to the electron density, resulting in increased
shielding at the adjacent C-1 and a lower observed d. The DMSO-
d6 may have some hydrogen bonding interaction at the
sulfoxide oxygen, whereas THF-d8 has relatively little interaction.
The low correlation coefficients for C-3,5 and C-4 indicate that
the solvent interaction has little to no effect at these positions
more distant from C-1. Overall, the data (Fig. 2) show the
sensitivity of the chemical shift of the C1 position induced by
the intermolecular interactions of the different alkali metals
and the medium.
Figure 2. 13C chemical shifts of alkali metal phenolates in CD3OD and D2O.
density; position 1 had low electron density, whereas the positions
3 and 5 had intermediate density. The alkali metal cation element
affected the overall chemical shift (Fig. 2). At C-1 and C2/C6, the
series from Li to Cs exhibited increased deshielding from Li to K.
There was minimum effect to the C-3 and C-5 positions, whereas
the cations caused the C-4 chemical shift to increase shielding from
Li to Cs.
Furthermore, there were consistent differences for the average
chemical shifts between the alkali metal phenolates in D2O and
CD3OD solvent (Fig. 2). Overall, the CD3OD allowed a greater
deshielding effect than the D2O, being on average 0.93 ppm
greater. Among the ring positions, the differences from the
largest to the smallest were C-1 (1.63) > C-2,6 (0.98) > C-4
(0.70) > C-3,5 (0.24). The minimum difference occurred with
Na and the maximum difference with Cs. Moreover, the effect
of solvents is attributed to dielectric effects, which will be consid-
ered further in the discussion on electrostatic effects.
The above mentioned data were compared with Kremer and
Schleyer’s[10] theoretical values for the 13C and 1H NMR chemical
shifts of the alkali metal phenolates. The major difference is
that the magnitude of the theoretical chemical shift is larger for
the C-1 carbon and smaller for the C-4 carbon. One possible
explanation of this discrepancy between theory and practice is
that their theoretical model did not take into account the
distance between the solvated cation and anion. Their model
may have assumed a very short distance between the anion–
cation pair. Kremer and Schleyer[10] provided one set of
calculations showing natural charges as a function of lithium
cation–phenolate anion distance, illustrating the effect of the
distance between the anion–cation pair on their effective charge.
Hence, a lower chemical shift can be interpreted as more of the
electron density remaining on the phenolate oxygen because
the effective field strength of the cation is stronger due to the
shorter distance between the ions.
Phenol, 2-HMP, and 4-HMP treated with alkali metal
hydroxides and dissolved in D2O/CD3OD mixture
The solvent mix of D2O and CD3OD was used to assure the
dissolution of all the model compound–alkali metal combinations.
The three alkali metal phenolates (phenol, 2-hydroxymethylphenol,
and 4-hydroxymethylphenol) exhibited the same general trends of
shielding and deshielding in reference to neutral phenol as for D2O
and CD3OD alone, although some specific variations occurred in
the data. Table 1 provides an overview of results for the effects of
alkali metal hydroxides on the chemical shifts of phenol, 2-HMP,
and 4-HMP. Similar to phenol, the effect at C-1 is strong deshielding
for 2-HMP and 4-HMP when exposed to an alkali metal hydroxide.
For the substituted phenols, the direction and magnitude of the
chemical shifts were affected by the counter-cation at different
substitution positions. The alkali metal counter-cation caused
deshielding at C-2 and C-6 for 4-HMP whereas the hydroxy-
methyl substituted C-2 was deshielded and the C-6 position
was shielded. In further contrast to phenol and 4-HMP, 2-HMP
demonstrated shielding at C-3 and C-5 positions. Another signif-
icant difference for the phenolic occurred at the C-4 position.
For phenol and 2-HMP, the chemical shift demonstrated
shielding with the different counter-cations, whereas 4-HMP
was deshielded due to the electronegative oxygen of the hydro-
xymethyl group attached to the C-4 carbon.
Phenol dissolved in D2O, DMSO-d6, CD3OD, and THF-d8
Chemical shifts for the hydroxymethyls were studied for both
the carbon and hydrogen groups. The hydroxymethyl methylene
group was deshielded for 13C and 1H NMR, whereas the hydroxy-
methyl hydroxyl proton was shielded for both 2-HMP and 4-HMP
(Table 2). The phenolic hydroxyl proton (H-7) was not detected in
Figure 3 shows the 13C chemical shifts of phenol in D2O, DMSO-
d6, CD3OD, and THF-d8 along with each solvent’s respective
dielectric constant. The C-1 carbon has the highest correlation
of decreasing chemical shift with increasing dielectric constant,
Magn. Reson. Chem. 2013, 51, 95–101
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

R. A. Haupt and S. Renneckar
Table 1. General effects of perturbing phenolic model compounds with alkali metal hydroxides in D2O/CD3OD for H/Li ! Cs
Position
Phenol
2-HMP
4-HMP
d for H/Li ! Cs
eꢁ effect
d for H/Li ! Cs
eꢁ effect
Deshielded
d for H/Li ! Cs
eꢁ effect
C-1
Deshielded
Deshielded
Low change
Shielded
—
156.7/162.5-167
156.7/162.5–167
(2) 127.7/130.1–129.6; (6) 116/115
(3) 129.2/(5) 130.0
121/120
Deshielded
Deshielded
Low change
Deshielded
Deshielded
Deshielded
Shielded
156.5/164–167
116/119–120
130.5
C-2,6
116/118–120
2 deshielded, 6 shielded
Shielded
C-3,5
131
C-4
121.3/118–115
Shielded
133.3/129–127
64.6/65.2
–CH2–
–CH2–
–CH2–OH
—
—
—
Deshielded
Deshielded
Shielded
60.6/62.3
—
4.63/4.90
4.51/4.90
—
4.85/4.63
4.87/4.43
Table 2. Chemical shift (ppm) for hydroxymethyl functional groups and the attached ring carbon
In D2O/CD3OD
2-HMP
4-HMP
Functional group
Neutral phenol
Anionic phenolate
Neutral phenol
Anionic phenolate
–OH
4.85
4.65
4.40
4.65
4.87
4.51
4.44
4.94
–CH2–
–CH2–
C–CH2–
60.59
127.67
62.34
129.89
64.60
130.24
65.21
127.31
1H NMR due to solvent exchange of this acidic proton. Compared
with the respective phenol C-2 and C-4 positions, the equivalent
ring positions on 2-HMP and 4-HMP were deshielded in the range
of 8 to 12 ppm due to the oxygen-containing substituent. Overall,
the chemical shifts of the phenolic model compounds at many
of the positions were significantly affected by the introduction
of the alkali metal hydroxides.
Electrostatic interactions
The Coulombic interaction of the alkali metal cation and
phenolate anion is the electrostatic interaction of two point
charges in a solvent medium at a specified distance.[11] The
important variables for consideration are the separation distance
between the point charges (anion–cation pair) and the dielectric
constant of the medium. In an ionic solution, hydration or
solvation creates a region of solvent molecules associated with
the ions in equilibrium. The solvation spheres of the phenoxide
anion and alkali metal cation will determine the distance “r”
between their charges, and thus the electrical field strength.
The solvent can also screen the electrical charge of the cations,
a property represented by its dielectric constant. The effects of
ionic radius and hydrated ionic radius on ionic volume and the
hydration shell volume are represented in Fig. 4. At the atomic
and molecular level, the dielectric effect is subject to local
fluctuations because the dielectric constant is a bulk property.[12]
A model has been developed for approximating the Coulombic
attraction of the alkali metal cations in solution and relating those
values to the 13C chemical shifts of phenol (døOH) as perturbed by
the alkali metal hydroxides (døOM). In the model, the strength of
the point charge is adjusted for the screening effects of the core
electrons by using the Pauling electronegativity value (w).[11] The
screening effects of the dielectric medium and the charge
separation distance created by the hydration radius are modeled
by the volume of the hydration shell (VHS). The hydration shell is
Figure 4. Ionic volume and hydrated shell volume for alkali metal
hydroxides in water.
the hydrated ionic volume (VHI) calculated from the hydrated
ionic radius (rHI) minus the ionic volume (VI) from the crystalline
ionic radius (rI). The combination of Pauling electronegativity in
the numerator and the hydration shell volume in the denomina-
tor gives an approximation of the Coulombic electric field
strength, with the equation:
w
w
E ꢀ
¼
VHS ðVHI ꢁ VIÞ
The relative chemical shifts are calculated from the equation:
Δ ¼ døOM ꢁ døOH
The plot of Δ versus w/VHS is shown in Fig. 5 for phenol, in
which w/VHS is normalized to the value for lithium with that being
1.00. As indicated by the R2 values, the correlations are good for
electronically sensitive positions at C-1, C-2,6, and C-4, whereas
wileyonlinelibrary.com/journal/mrc
Copyright © 2012 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2013, 51, 95–101

Electronic behavior of phenolic monomers in solution and implications for addition and self-condensation
is approximated by the chemical shift of the respective ring
carbon atoms. For electrophilic aromatic substitution, higher
electron density on unsubstituted ring carbons in positions 2, 4,
and 6 would be expected to be more reactive, thus lower
chemical shift values should indicate the more reactive sites
and reactivity can be ranked accordingly. The NMR chemical
shift data of these experiments were ranked for phenol, 2-HMP,
and 4-HMP at the unsubstituted ring carbons in positions 2, 4,
and 6 and was compared with the literature kinetic rate constants
for formaldehyde addition reactions. On the basis of the data
used in Fig. 2, under alkaline conditions, the C-4 position of
phenol will be more reactive than the C-2 position. The relative
reactivity for the alkali metal hydroxide cations, however, will
vary by species; the rate constants for the positions at C-2 and
C-4 in the presence of lithium should be similar, whereas the
kinetic rates for the C-4 position should be greater than the C-2
position in the presence of sodium and potassium ions. The
addition reaction rate constants for C-2 and C-4 under rubidium
and cesium catalysis should be similar to those for potassium.
From these data, the directing effects for a specific catalyst may
be predicted. The preferential direction of formaldehyde addition
to phenol for the alkali metal hydroxides should be in the order
Li > Na > Rb > K > Cs for the C-2 position and Cs > K > Rb > Na >
Li at the C-4 position.
Figure 5. Correlation of change in chemical shifts of phenol versus the
Coulombic electrical field strength modeled from Pauling electronegativ-
ity (w) and hydrated shell volume (VHS).
little correlation is found at the electronically insensitive position
for C-3,5. The good correlation gives credence to the approach of
using the electrostatic interactions (electric field strength) to pre-
dict the effect on chemical shifts which affect the reactivity of the
different carbon positions.
Discussion
For 2-HMP, the addition rate to the C-6 position should be
faster than to the C-4 position due to the higher overall electron
density on C-6, although this prediction directly contradicts
published literature rate constants.[4,13–16] In principle, the
ranking of direction of formaldehyde addition to 2-HMP for
the alkali metal hydroxides should be Cs > K > Rb > Na > Li for
the C-6 position and Li > Na > Rb > K > Cs for the C-4 position.
The ranking of the direction of formaldehyde addition to 4-HMP
for the alkali metal hydroxides should be equal because C-2
and C-6 are chemically equivalent.
The relative shielding for each of these positions for phenol,
2-HMP, and 4-HMP have been ranked based on the average ion of
the alkali metal hydroxide series to give an expected relative
reaction rank. The comparison rankings are generated from the
relative rate constants averaged from the five separate measure-
ments for the alkali metal hydroxides as the literature rate
constants showed substantial divergence among themselves.[4,13–16]
Although the ranking from the chemical shifts (Table 3) correctly
predicts the average relative reaction rate of four of the five
reactions based on the literature (Table 4), the prediction for form-
aldehyde addition to position 4 of 2-HMP, however, is substantially
off. Freeman and Lewis[14] attributed the faster addition of
formaldehyde to 2-HMP, 2,4-DHMP, and 2,6-DHMP to resonance
Implications for reactivity
The electron density of the unsubstituted phenolic ring carbons
in positions 2, 4, and 6 should indicate the directing effect on
formaldehyde addition reactions. The electron density of the hydro-
xymethyl carbon and hydrogen atoms, along with the substituted
ring carbon in positions 2, 4, and 6 should influence reactivity in
phenolic self-condensation, accelerated self-condensation, and
urethane formation with isocyanate functionalized compounds.
Table 3 presents the relative chemical shifts as a proxy for the
electron density of these particular structural groups.
Addition reactions: electrophilic aromatic substitution
Most of the change in chemical shift relative to phenol is gener-
ated by the addition of an alkali metal hydroxide, but the specific
cation also affects the chemical shift to a lesser extent. Grenier-
Loustalot et al.[3] has previously demonstrated that the alkali
metal cation affects the addition reaction rate. The reactivity of
the alkali metal phenolates toward formaldehyde depends on
the availability of high electron density at positions 2, 4, and 6
ring positions for electrophilic substitution. This electron density
Table 3. 13C NMR chemical shifts for unreacted ring sites of alkali metal phenolates
Counterion
Phenol
2-HMP
4-HMP
Statistics
dC-4
dC-2,6
dC-4
dC-6
dC-2,6
Average
Rank
Li
117.34
115.33
117.08
117.60
118.13
117.10
2
118.59
119.78
118.74
118.06
118.13
118.66
3
119.60
119.79
120.05
120.03
120.21
119.94
5
115.65
115.28
114.82
114.91
114.91
115.11
1
118.64
119.55
119.40
119.67
119.93
119.44
4
117.96
117.95
118.02
118.05
118.26
(Highest eꢁ density = 1
Lowest eꢁ density = 5)
2
1
3
4
5
Na
K
Rb
Cs
Average
Rank
Magn. Reson. Chem. 2013, 51, 95–101
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

R. A. Haupt and S. Renneckar
Table 4. Relative rate of hydroxymethylation
Literature rate constants relative to the phenol + formaldehyde ➞ 2-
HMP reaction = 1.0
Reaction
d Rank
1
2
3
4
5
(Average)
Phenol + Fa ➞ 2-HMPb
Phenol + F ➞ 4-HMP
2-HMP + F ➞ 2,6-DHMPc
2-HMP + F ➞ 2,4-DHMP
4-HMP + F ➞ 2,4-DHMP
2,6-DHMP + F ➞ THMPd
2,4-DHMP + F ➞ THMP
(3)
(2)
(1)
(5)
(4)
1.00
1.18
1.66
1.39
0.71
7.94
1.73
1.00
2.08
1.08
2.58
0.83
3.25
1.25
1.00
1.09
1.98
1.80
0.79
3.33
1.67
1.00
1.46
1.70
3.80
1.02
4.54
1.76
1.00
1.46
1.87
3.35
0.91
4.73
1.75
(1.00)
(1.45)
(1.65)
(2.58)
(0.85)
(4.76)
(1.63)
aFormaldehyde.
bHydroxymethylphenol.
cDihydroxymethylphenol.
dTrihydroxymethylphenol.
1 = Freeman and Lewis 1954 (30 ꢂC).[14]
2 = Minami and Ando 1956 (40 ꢂC).[15]
3 = Zsavitsas and Beaulieu 1968 (30 ꢂC).[16]
4 = Eapen and Yeddanapalli 1968 (30 ꢂC).[17]
5 = Higuchi, Nohno, Morita, and Tohmura 1998 (20 ꢂC, 30 ꢂC, and 40 ꢂC).[5]
H
enhancements generated by intramolecular hydrogen bonding.
These hydrogen bonds chelate with the quinoidal anionic oxygen,
giving it a higher electronegativity. A double bond present
within the chelate ring would stabilize it, enhancing the resonance
contribution. These data show that the composition of PF
hydroxymethylphenols may be directed by catalyst selection
toward specific ring carbon sites during formaldehyde addition
and the chemical shift data can be used as a first approximation
in most cases.
OH
O
OH
OH
H
O
OH
H2O
+
O
O
O
O
O
Implications in condensation reactions: neutral conditions
Under neutral conditions, HMPs are known to form dibenzyl
ether bridges from two hydroxymethyl groups.[18,19] A probable
mechanism for this reaction is shown in Fig. 6a. This reaction
depends on the electron density distribution between the
methylene carbon and the hydroxyl oxygen favoring the oxygen
atom. The relatively electron-rich oxygen will attack the lower
electron density methylene carbon, forming an ether bridge with
the elimination of water. The more deshielded methylene carbon
of 4-HMP would indicate that this reaction may have a lower
activation energy than the same reaction for 2-HMP. This reaction
for the 2-HMP is known to be relatively slow, requiring a high
activation energy input.[18]
O
HO
O
CH2O
+
O
CH2OH
Figure 6. Reaction mechanism related to phenolic monomers. Benzylic
ether formation mechanism (a), carbocation resonance structures for 2
and 4 QMs (b), and mechanism of 4-HMP self-condensation by electrophilic
attack of a QM carbocation resonance structure on an anionic resonance
structure of 4-HMP (c).
Implications in condensation reactions: anionic
(alkaline) conditions
neutral monomer and an anionic monomer.[23,24] One possible
mechanism involves the attack of the electron-rich, hydroxy-
methyl substituted ring carbon by a QM carbocation with
subsequent emission of a formaldehyde molecule as illustrated
in Fig. 6c. For the phenolate salts of 2-HMP and 4-HMP, the 13C
chemical shifts for the substituted ring carbons are found in
the range of 129.62 to 130.12 ppm for 2-HMP and 126.85 to
128.68 ppm for 4-HMP. The higher electron density of the 4-HMP
ring carbon is in accord with the established self-condensation
reactivity order of 4-40 > 4-20 > 2-20.[3,24,25]
The self-condensation reactivity of the alkali metal salts of 2-HMP
and 4-HMP is directed to the electron density at the 2 and 4
hydroxymethyl substituted ring positions. The condensation
kinetic rates will be dependent on the electron density at the
reaction sites and the particular mechanism involved, which is
thought to be mediated by a QM mechanism.[20–22] The reso-
nance structures of QMs that can form from 2-HMP and 4-HMP
are shown in Fig. 6b, in which the carbocation is a powerful
electrophile. The reaction is considered to occur between a
wileyonlinelibrary.com/journal/mrc
Copyright © 2012 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2013, 51, 95–101

Electronic behavior of phenolic monomers in solution and implications for addition and self-condensation
When considering the faster PF reactivity of potassium
observed by Daisy,[2] the 13C chemical shift at C-2 of 2-HMP was
129.77 and 129.96 ppm, whereas at C-4 of 4-HMP it was 126.92
and 127.23 ppm for sodium and potassium, respectively. In both
cases, the electron density at the ring carbon was higher for
sodium than potassium, inconsistent with the expected results
for the QM carbocation mechanism attacking the substituted ring
carbon with the highest electron density. This discrepancy
suggests that the basis for enhanced reactivity may not be found
with the anionic HMP species, but possibly that potassium may
foster a faster rate of QM formation than sodium.
The implication is that an HMP would be more susceptible to
electrophilic substitution by a QM carbocation at the C-4
position than the C-2 position in the anionic form, consistent
with the order of reactivity published in the literature.
Further research could attempt to quantify the relationship
between NMR chemical shifts and ab initio calculations of
theoretical electron density for the model compounds studied
here. Overall, NMR chemical shift data are most useful for interpret-
ing reactivity when combined and corroborated with information
on reaction kinetics, mechanism, and physical properties.
Acknowledgements
Conclusions
This work was made possible with funding from the Dynea Group
and the Graduate School of Virginia Tech.
This research has identified the fundamental importance of the
electrostatic interactions on the chemical shift of the substituted
or unsubstituted phenolic aromatic rings and related this to the
reactivity of compounds previously published in the literature.
Subtle effects relate to how the solvated ionic radius and solvent
dielectric constant affect the strength of the electric field gener-
ated by the ions. Reactivity in turn is dependent on the electronic
structure of the various reactants, monitored by changes to the
chemical shift, and is affected by changes in the alkali metal
hydroxide and solvent.
References
[1] D. Fengel, G. Wegener, Wood: chemistry, ultrastructure, reactions,
W. de Gruyter, Berlin, 1984.
[2] N. Daisy, US Patent 5074946, Borden, Inc., 1991.
[3] M.-F. Grenier-Loustalot, S. Larroque, D. Grande, P. Grenier, D. Bedel.
Polymer 1996, 37, 1363.
[4] M. Higuchi, S. Nohno, M. Morita, S.-I. Tohmura. J. Wood Sci. 1999, 45,
306.
[5] M. Higuchi, S.-I. Nohno, S.-I. Tohmura. J. Wood Sci. 1998, 44, 198.
[6] A. H. Conner, L. F. Lorenz, K. C. Hirth. J. Appl. Polym. Sci. 2002, 86,
3256.
•
Alkali metal cations act as an electron withdrawing force
on the phenolic monomer’s electronic structure, with the
strength of their effect mostly dependent on the proximity
of the phenoxide anion and resonance effects. The effect of
the cation is related to the dielectric screening of their inner
shell electrons and the local dielectric effects of the solvation
shell. The greatest effect is deshielding at the C-1 carbon,
whereas the C-3 and C-5 carbons are mostly unaffected by
the perturbation.
[7] N. Kamo, J. Tanaka, M. Higuchi, T. Kondo, M. Morita. J. Wood Sci.
2006, 52, 325.
[8] R. A. Haupt, US Patent US20110151180 Rapid curing aldehyde
resinpolyisocyanate composition and method for producing hybrid
polymer Sylvanova, Inc. 2010, 56 pp.
[9] P. Yang, T. Li, J. Li, X. Zhu, Y. Xi. Prog. React. Kinet. Mech. 2010, 35, 93.
[10] T. Kremer, P. v. R. Schleyer. Organometallics 1997, 16, 737.
[11] P. C. Hiemenz, R. Rajagopalan (Eds), Principles of Colloid and Surface
Chemistry (3rd edn, Revised and Expanded), Dekker, New York, 1997.
[12] J. N. Israelachvili, Intermolecular and Surface Forces, McGraw-Hill
Publishing Co. Japan, Ltd, London, 1991.
[13] K. C. Eapen, L. M. Yeddanapalli. Makromol. Chem. 1968, 119, 4.
[14] J. H. Freeman, C. W. Lewis, J. Am. Chem. Soc. 1954, 76, 2080.
[15] T. Minami, T. Ando. Kogyo Kagaku Zasshi 1956, 59, 668.
[16] A. A. Zavitsas, R. D. Beaulieu, J. R. Leblanc. J. Polym. Sci., Part A: Polym.
Chem. 1968, 6, 2541.
•
Solvents shift the electron density, as revealed by NMR
chemical shift, probably due to hydrogen bonding interac-
tions with the hydroxyl groups. Methanol was found to be
more deshielding than water, lowering electron density at
reactive sites. For neutral compounds, DMSO was more
shielding of the carbon backbone than THF, but was more
deshielding of the hydroxyl protons.
[17] K. C. Eapen, L. M. Yeddanapalli. Die Makromolekulare Chemie 1968,
119, 4.
•
•
The hydroxymethyl substituted carbon at C-2 or C-4 is
deshielded, whereas the unsubstituted C-2, C4, and C-6
positions when shielded are thus made more reactive toward
formaldehyde addition. The correct order of reactivity was
predicted for formaldehyde addition for four of five reactions.
Hydroxymethyl substitution on the C-2 or C-4 position has
the overall effect of deshielding the C-2 ring carbon, whereas
the C-4 position is more shielded than their respective ring
carbons in the neutral state. The chemical shift of the substituted
C-2 position is more deshielded in the anionic form than C-4.
[18] P. Christjanson, T. Pehk, K. Siimer, J. Paju. J. Appl. Polym. Sci. 2008,
107, 1226.
[19] M. M. Sprung, M. T. Gladstone. J. Am. Chem. Soc. 1949, 71, 2907.
[20] L. Diao, C. Yang, P. Wan. J. Am. Chem. Soc. 1995, 117, 5369.
[21] M. Higuchi, T. Urakawa, M. Morita. Polymer 2001, 42, 4563.
[22] A. Merijan, P. D. Gardner. J. Org. Chem. 1965, 30, 3965.
[23] R. A. Haupt, S. Waago, in Wood Adhesives (Eds: A. W. Christiansen,
A. H. Conner), Forest Products Laboratory, Madison, 1995, p. 220.
[24] L. M. Yeddanapalli, D. Joseph Francis. Die Makromolekulare Chemie
1962, 55, 74.
[25] M. G. Kim, L. W. Amos, E. E. Barnes. Ind. Eng. Chem. Res. 1990, 29,
2032.
Magn. Reson. Chem. 2013, 51, 95–101
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc