polymers

Article
Mechanism of Base-Catalyzed Resorcinol-
Formaldehyde and Phenol-Resorcinol-Formaldehyde
Condensation Reactions: A Theoretical Study
Taohong Li 1,2, * ID
, Ming Cao 1,2 , Jiankun Liang 1,2 , Xiaoguang Xie 3 and Guanben Du 1,2, *
1 The Yunnan Province Key Lab of Wood Adhesives and Glued Products, Southwest Forestry University,
Kunming 650224, China; caominghappy@swfu.edu.cn (M.C.); liangjiankun@swfu.edu.cn (J.L.);
2 Key Lab for Forest Resources Conservation and Utilisation in the Southwest Mountains of China,
Southwest Forestry University, Ministry of Education, Kunming 650224, China
3 School of Chemical Science and Technology, Yunnan University, Kunming 650091, China; xgxie@ynu.edu.cn
* Correspondence: lith.cool@163.com (T.L.); guanben@swfu.edu.cn (G.D.);
Tel.: +86-871-6386-2581 (T.L.); +86-871-6386-3472 (G.D.)

Received: 13 August 2017; Accepted: 5 September 2017; Published: 7 September 2017

Abstract: The base-catalyzed resorcinol-formaldehyde condensation reactions were theoretically

investigated in this study by employing a quantum chemistry method. The condensation reaction includes
two steps: (1) formation of the quinonemethide (QM) intermediate from hydroxymethylresorcinol;
(2) Michael addition between the quinonemethide and resorcinol anion. The first step is the
rate-determining step. Two mechanisms, unimolecular elimination of the conjugate base (E1cb) and
water-aided elimination (WAE), were identified for the formation of QM. The hydroxymethylresorcinol
anion produces neutral QM while the dianion produces a quinonemethide anion (QMA).
The calculated potential energy barriers suggested that the QMA formation is much more favorable.
Although resorcinol-formaldehyde and phenol-formaldehyde condensations share a common mechanism,
the former would be faster if the QMA participates in condensations. The potential energy barriers
for formation of 2-QM, 4-QM, 6-QM, 2-QMA, and 4-QMA were calculated. The results show that the
formations of 6-QM and 4-QMA have relatively lower energy barriers. This rationalized previous
experimental observations that the 2,4-(2,6-) and 6,60 -(4,40 -) methylene linkages were dominant,
whereas the 2,20 -linkage was almost absent. The resorcinol-phenol-formaldehyde co-condensations
were also calculated. The cold-setting characteristic of phenol-resorcinol-formaldehyde co-condensed
resin can be attributed to participation of resorcinol quinonemethides in condensations.

Keywords: resorcinol-formaldehyde; base-catalyzed; condensation; quinonemethide

1. Introduction
It is known that the base or acid catalyzed resorcinol-formaldehyde (RF) reactions can form
polymeric resins which are currently used as wood adhesives and composites [1–7], forms [8],
and organic gels [9–12]. Depending on the F/R molar ratio, RF reactions can result in novlac or resol
type resin [2]. RF resin is generally not used separately as wood adhesive due to its instability and high
cost. Alternatively, in order to accelerate the cure of PF resin, resorcinol is generally mixed with PF resol
resin, obtaining the so-called phenol-resorcinol-formaldehyde (PRF) co-condensed resins [2,3,5–7].
Due to the similarity of chemical structure to phenol, it is natural to believe the RF polymers are
also similar to phenol-formaldehyde (PF) polymers in which methylene bridges can be formed
though reactions between phenol and hydroxymethylphenol or between hydroxymethylphenols.
However, a significant difference between PF and RF resin is that curing of PF resin generally requires
a hot-press temperature above 160 ◦ C, whereas RF and PRF are cold-setting resins. This indicates that

Polymers 2017, 9, 426; doi:10.3390/polym9090426 www.mdpi.com/journal/polymers

Polymers 2017, 9, 426 2 of 14

RF condensations encounter a much lower energy barrier and occur much faster. Superficially, higher
reactivity may be attributed to the stronger nucleophilicity of resorcinol since it has two phenolic
hydroxyl groups which lead to higher electron density on the aromatic ring. In the viewpoint
of a mechanism at the molecular level, formation of reactive intermediates is the key factor that
determines the rate of the condensation reaction. Therefore, comparing the kinetics of the PF and
RF intermediates formation can help us better understand the different behaviors of the two resins.
But, the experimental kinetics study seems to be incapable of doing so because capture and quantitative
description of such short-lived intermediates is currently almost impossible, especially for solution
reactions. Fortunately, theoretical methods have become powerful tools that allow us to explore the
microscopic world of chemical reactions in silico. In the field of wood adhesive resins, pioneering
works that employed molecular mechanics methods are from the group of Professor A. Pizzi [13–17].
These studies have investigated the interactions between resin components and wood cellulose,
aiming to establish theoretical models for resin–wood surface interaction and giving insight into
the relationship between resin structure and performance. The obtained results can help us design
novel resins, but the prerequisite is that we can control the reactions based on our knowledge about
mechanisms. Therefore, understanding the reaction mechanisms and resin chemistry is another task.
Different from molecular mechanics, which is based on force filed theory and mainly deals with
the hydrogen bonding interactions, Van der Waals’ force and their influences on molecule energy
and conformations, quantum chemistry methods that is based on quantum mechanics can deal with
formation and breakage of chemical bonds, kinetics and thermodynamics of chemical reactions through
rigorous calculations on the electronic structures of the involved species. To explain the different
condensation reactivity of PF and RF, quantum chemistry calculations appear to be necessary.
By employing a quantum chemistry method, we recently studied the base-catalyzed phenol-
formaldehyde reactions and the quinonemethide (QM) was confirmed to be the key intermediate
that initiates condensations [18]. RF condensation reactions may share the same mechanism.
But, the chemistry of resorcinol is obviously more complex than phenol. Resorcinol has two hydroxyl
groups and therefore dissociation of the protons can produce a singly charged or doubly charged
anion (dianion) with the presence of a base. As a result, different quinonemethide intermediates
may be formed. As can be seen in Figure 1, five possible intermediates may be formed according to
the proposed mechanism. 2-QM, 4-QM, and 6-QM are neutral species that are similar to phenol
quinonemethides. But, 2-QMA and 4-QMA are anions which may exhibit different reactivity.
According to the results for PF reactions [18], formation of quinonemethide is the rate-determining step
of the overall condensation reaction. Thus, it can be speculated that formations of RF quinonemethides
encounter lower energy barriers. But which of the neutral QMs and QM anions are more favorable?
This is the first issue to be addressed in this study.
Another task of this work is to rationalize the experimentally observed condensed structure
of RF resin. Theoretically, resorcinol has three reactive sites on the ring, namely, the three ortho
positions of the two hydroxyl groups. Thus, three types of methylene linkages through 2,20 -,
2,4-, and 4,40 -condensations should be formed. But, it is interesting that some 13 C NMR studies
indicated that the 2,40 - and 4,40 -linkages were found to be dominant, whereas 2,20 -linkage was
observed to be minor or absent [2–4]. Further, 4,40 -linkage appeared to be more favorable than
2,4-linkage. These observations imply that the 4-position is more reactive than 2-position. A plausible
explanation is the 4-position is more reactive toward formaldehyde, leading to higher concentration of
4-hydroxymethylresorcinol. But the hydroxymethylation reaction should be faster than condensations,
considering that resorcinol is highly reactive toward formaldehyde in the presence of catalysts.
Therefore, different reaction rates of rate-determining steps may be another reason. That is to
say, 2-hydroxymethylresorcinol and 4-hydroxymethylresorcinol may have different rates of forming
reactive intermediates. Which is the more important factor? Or do both of them work? This issue has
never been addressed in detail, according to our survey of the literature.
Polymers 2017, 9, 426 3 of 14

of forming reactive intermediates. Which is the more important factor? Or do both of them work?
Polymers 2017, 9, 426 3 of 14
This issue has never been addressed in detail, according to our survey of the literature.

Figure 1.
Figure The proposed
1. The proposed mechanisms
mechanisms for
for formation
formation of
of resorcinol
resorcinol quinonemethides.
quinonemethides.

Finally, in in the synthesis

the synthesis of phenol-resorcinol-formaldehyde
of phenol-resorcinol-formaldehyde resins, PF and resins, PF and RF
RF self-condensations
self-condensations
and PRF co-condensationsand PRF co-condensations
should shouldThe
occur simultaneously. occur simultaneously.
competitive relationshipThe competitive
of the reactions
relationship of the reactions
is also an important issue thatisdeserves
also an attention
importantbecause
issue that
suchdeserves attention
a relationship becausewhether
determines such a
relationship
co-condensed determines
structures whether co-condensed
can be formed structures can be formed efficiently.
efficiently.

2. Theoretical Calculations
By employing
employingthe thedensity
densityfunctional theory
functional [19,20]
theory (DFT)(DFT)
[19,20] method B3LYP B3LYP
method [21–23] with the with
[21–23] standard
the
basis set 6-31++G** [24–27], full structural optimizations were performed
standard basis set 6-31++G** [24–27], full structural optimizations were performed on all the on all the stationary points
on the reaction
stationary points potential energy surfaces
on the reaction potential(PES),
energy including
surfaces reactants, intermediates,
(PES), including reactants, transition states,
intermediates,
and products.
transition The
states, harmonic
and products.vibrational frequencies
The harmonic were calculated
vibrational frequenciesatwere the same theoretical
calculated at thelevel to
same
characterize the nature of the stationary point as a local minimum or first-order
theoretical level to characterize the nature of the stationary point as a local minimum or first-order saddle point (transition
state) onpoint
saddle the PES. The zero-point
(transition state) onvibrational
the PES. The energies (ZPVE)
zero-point were usedenergies
vibrational to correct(ZPVE)
the relative
wereenergies.
used to
Each transition
correct the relativestateenergies.
(TS) wasEachidentified as itstate
transition has a(TS)
unique
was imaginary
identified as frequency. Intrinsic
it has a unique reaction
imaginary
coordinate (IRC)
frequency. analysis
Intrinsic was also
reaction performed
coordinate (IRC)to confirm
analysisthewas connectivity between to
also performed a TSconfirm
and a local
the
minimum. Tobetween
connectivity compareathe TSrelative
and a nucleophilicity
local minimum. of To
different
compare carbons
the of the resorcinol
relative anion and
nucleophilicity of
dianion, the APT (atomic polar tensor) charges were also calculated. To simulate
different carbons of the resorcinol anion and dianion, the APT (atomic polar tensor) charges were the implicit solvent
effects,
also for all theTocalculations
calculated. simulate the theimplicit
self-consistent
solvent reaction
effects, forfield
all(SCRF) method was
the calculations theemployed with
self-consistent
the polarizable
reaction continuum
field (SCRF) methodmodel
was(PCM) [28–30]
employed withby the
defining water as
polarizable the solvent
continuum (Water:
model ε = 78.3553).
(PCM) [28–30]
All the calculations
by defining water as were
the carried
solventout with εthe
(Water: GAUSSIAN
= 78.3553). All 03theprogram package
calculations were(Gaussian,
carried out Pittsburgh,
with the
PA, USA) [31].
GAUSSIAN 03 program package (Gaussian, Pittsburgh, PA, USA) [31].
The potential energy barrier (theoretical activation energy) of each calculated reaction step was
described by bythetherelative
relativeenergy (energy
energy gap) between
(energy a transition
gap) between state and state
a transition a reactant
and or a reactant-like
reactant or
intermediate.intermediate.
reactant-like The relative energies wereenergies
The relative obtained by the
were following
obtained formulas:
by the following formulas:
ΔE(TS)==[(E(TS)
∆E(TS) [(E(TS)++ ZPVE(TS)]
ZPVE(TS)] −
− [E(R)
[E(R) ++ ZPVE(R)]
ZPVE(R)]

∆E(P) = [(E(P) + ZPVE(P)] − [E(R) + ZPVE(R)]

Polymers 2017, 9, 426 4 of 14

Here, E(TS) is the total electron energy of a TS, while E(R) represents the total electron energy of a
local minimum which can be an initial reactant or a reactant-like intermediate. Thus, ∆E(TS) represents
the relative energy between TS and reactant or reactant-like intermediate, namely the energy barrier of
a reaction step. Similarly, ∆E(P) corresponds to the relative energy between a product or product-like
intermediate. In this context, the potential energy profile shown in this paper was demonstrated by
choosing the energy of reactant or a reactant-like intermediate as reference.

3. Results and Discussion

3.1. Formations of Quinine Methide Intermediates

Figure 1 shows two mechanisms for base-catalyzed formation of quinonemethide which was
derived from that identified for phenol-formaldehyde reactions [18]. Pathway (A) demonstrates the
E1cb (unimolecular elimination of conjugate base) mechanism. Pathway (B) is another one in which
intra-molecular water elimination is involved.
According to the proposed mechanisms, three neutral quinonemethide intermediates, 2-QM,
4-QM, and 6-QM, can be produced from singly charged hydroxymethylresorcinol. If a resorcinol
dianion participates in the reaction, a quininemethide anion will be formed. Concentrations of the two
types of intermediate are dependent on the concentration of the alkaline catalyst (or pH).
In the experiment of Christiansen [4], 76% formaldehyde was converted to 4- or 6-hydroxymethyl
groups (4- and 6-position are equal due to symmetry of resorcinol), whereas only 12% formaldehyde
took the form of 2-hydroxymethyl groups. Three factors may be attributed to this result. First, the 4- and
6-positions are more reactive toward formaldehyde due to their stronger nucleophilicity. Another one
is a statistical factor. Namely, two positions versus one. Finally, Durairaj pointed out that the steric
hindrance on the 2-position caused by the two adjacent hydroxyl groups is a key factor [32]. To confirm
the first factor, the APT atomic charges were calculated and are given in Figure 2. In the resorcinol
anion (RA), the carbons numbered as 2, 4, and 6 are negatively charged as −0.574, −0.479, and −0.642,
while carbon 5 is positively charged as 0.378, clearly indicating the three ortho positions are strong
nulceophilic centers. But, according to the charge distribution, the three positions are different
in reactivity as the nucleophilic order appears to be 6 > 2 > 4. Position 4 is less reactive than
position 2, suggesting that, for RA, high reactivity of 6-position is responsible for dominant formation
of corresponding hydroxymethyl groups. In contrast with the statistical factor, nucleophilicity plays
a more important role. For the resorcinol dianion (RDA), the negative charges of the three ortho
positions were almost equal. But, the statistical factor determines 4- and 6-positions have higher
probability in reacting with formaldehyde. Steric hindrance may be partially responsible for the slower
hydroxymethylation in 2-postion, but it cannot rationalize the results that the 2-hydroxymethyl groups
and 2,40 -methylene linkages were produced while 2,20 -methylene linkages were absent. These results
also cannot be explained by nucleophilicity and statistical effects. Slower hydroxymethylation
is not definitely in accordance with slower condensation reactivity. Theoretical calculations on
PF reactions revealed that quinonemethide formation is the rate-determining step in the overall
condensation reaction, and formation of para quinonemethide is faster than the formation of ortho
quinonemethide [18]. This explained the experimental results that ortho-ortho-methylene linkage was
always observed to be minor in comparison with ortho-para- and para-para-methylene linkages. So, it is
also necessary to compare the reactivity of 2-, 4-, and 6-hydroxymethylresorcinol anion (HMRA) in
formation of quinonemethide intermediates.
Polymers 2017, 9, 426 5 of 14
Polymers 2017, 9, 426 5 of 14

Figure 2.
Figure Thecalculated
2. The calculated atomic
atomic charges
charges for
for resorcinol
resorcinol anion
anion (RA)
(RA) and
and dianion
dianion (RDA).
(RDA).

Figure
Figure 33 shows showsthe thecalculated
calculatedstructures
structuresofof thethereactant
reactant intermediates
intermediates andand transition
transition states for
states
the E1cb
for the E1cb formation
formation of of
different
different neutral
neutralquinonemethides
quinonemethidesfrom fromhydroxymethylresorcinol
hydroxymethylresorcinolanions. anions.
The structures
The structures of produced
of produced quinonemethides
quinonemethides are collected in areFigurecollected in Figure 4. In
4. In 2-hydroxymethylresorcinol
2-hydroxymethylresorcinol
anion (2-HMRA), a strong intra-molecular anion (2-HMRA), hydrogen a strong intra-molecular
bond (1.747 Å) was formed hydrogen
betweenbond (1.747 Å)
hydroxymethyl
was
group formed
and the between
carbonyl hydroxymethyl
oxygen. In contrast, groupthe and the carbonyl
hydrogen bond in oxygen. In contrast, the hydrogen
4-hydroxymethylresorcinol anion
bond in 4-hydroxymethylresorcinol anion (4-HMRA) is weaker since
(4-HMRA) is weaker since the bond length is longer (1.814 Å). In 6-HMRA, the hydrogen bond is formed the bond length is longer
(1.814
between Å).two In 6-HMRA, the hydrogen
hydroxyl groups and thebondvery is formed
long bondbetween
length oftwo hydroxyl
2.660 Å indicatesgroupsthe and the very
interaction is
long
almost bond length of
ignorable. The2.660 Å indicates
hydrogen bondingthe effects
interaction
seems is to
almost
be a ignorable.
factor that The hydrogen
stabilizes thesebonding
species.
effects
But, theseemsbond to be aorder
length factoristhat stabilizes these
not consistent with thespecies. But,order,
stability the bond length
as it can order
be seen thatis the
not4-HMRA
consistent is
with
the mostthe stable
stability one,order, as it itcan
therefore wasbechosen
seen that the 4-HMRA
as reference in the is the most
potential stableprofiles.
energy one, therefore it was
In fact, besides
chosen
the hydrogen as reference
bondingininteraction,
the potential energy profiles.
the electronic structures In offact,
thebesides
isomersthe alsohydrogen
play an importantbonding
interaction, the electronic structures of the isomers also play an
role, especially the delocalization extent of electrons in the conjugate system. However, as can important role, especially the
delocalization
be seen from the extent of electrons
potential energy in the conjugate
profiles in Figure system.
5, the E1cb However,
transition as state
can be seen fromhas
(2-E1cb-TS) thea
potential energy profiles in Figure 5, the E1cb transition state (2-E1cb-TS)
higher energy barrier (113.7 kJ/mol) than 4-E1cb-TS (107.3 kJ/mol) by 6.7 kJ/mol. Further, 6-E1cb-TS has a higher energy
barrier
has the(113.7lowest kJ/mol)
energythan 4-E1cb-TS
barrier of 93.5(107.3
kJ/mol.kJ/mol)
Thisby 6.7 kJ/mol.
order Further,
is consistent with 6-E1cb-TS
the order has ofthe lowest
hydrogen
energy
bonding barrier
strength.of 93.5 kJ/mol. This
Obviously, order is consistent
the hydrogen with the
bonding effect order ofthe
suppresses hydrogen
leavingbonding
of an OHstrength.ion and
Obviously, the hydrogen bonding effect suppresses the leaving
the relative bonding strength is the key factor that determines the relative reactivity for HMRAs of an OH ion and the relative
bonding strength is the key
to form quinonemethide factor that In
intermediates. determines the relative
our calculations for PFreactivity
reactionsfor [18],HMRAs
such effect to form
was
quinonemethide
also found. In addition, intermediates. In our calculations2-QM,
the quinonemethides, for PF4-QM,
reactions and[18],
6-QM such areeffect was also
different found.
in stability.
In addition, 6-QM
Specifically, the quinonemethides,
is the most stable 2-QM,
species4-QM,
and is and 6-QM
more are than
stable different
2-QMinand stability.
4-QM Specifically,
by 20.8 and
6-QM is therespectively.
7.4 kJ/mol, most stable Thus,
species and is more
considering bothstable than 2-QM
the energy andand
barriers 4-QM by 20.8
relative and of
stability 7.4different
kJ/mol,
respectively. Thus, considering both the energy barriers and
quinonemethides, formation of 6-QM is energetically the most favorable and condensations involving relative stability of different
quinonemethides,
6-QM are the fastest, formation
whereasofcondensations
6-QM is energetically
related to 2-QM the most favorable
are less and condensations
competitive. The reaction
involving
of 6-QM with 6-QManother
are the monomer,
fastest, whereas condensations
resorcinol related to 2-QM arewill
or hydroxymethylresorcinol, lesslead
competitive.
to 2-6 orThe 4-6
reaction of 6-QM with another monomer, resorcinol or hydroxymethylresorcinol,
methylene linkage. Due to the symmetric structure of resorcinol, they are equal to 2,4 or 4,4 -methylene will lead
0 to 2-6 or
4-6 methylene linkage. Due to the symmetric structure of resorcinol,
linkage. Therefore, in addition to the higher concentration of 4- or 6-hydroxymethyl group, the lower they are equal to 2,4 or
4,4′-methylene
energy barrier linkage. for 6-QM Therefore,
formationinalso addition
playstoanthe higher concentration
important of 4- or 6-hydroxymethyl
role. Correspondingly, because of the
group, the lower energy barrier for 6-QM formation also plays an important
lower concentration of the 2-hydroxymethyl group and higher energy barrier for formation of 2-QM, role. Correspondingly,
because of the linkage
2,20 -methylene lower concentration of the 2-hydroxymethyl
was almost prohibited, or only very minorly groupformed.
and higher energy barrier for
formation of 2-QM, 2,2′-methylene linkage was almost prohibited, or only very minorly formed.
 Polymers 2017,
Polymers 2017, 9,
9, 426
426 66 of
of 14
14
Polymers 2017, 9, 426 6 of 14

Figure 3. The calculated structures of the intermediates and transition states for E1cb formation of
Figure 3.
Figure Thecalculated
3. The calculatedstructures
structuresof
of the
the intermediates
intermediatesand
and transition
transitionstates
statesfor
for E1cb
E1cb formation
formation of
of
quinonemethide from hydroxymethyresorcinol anion.
quinonemethide from
quinonemethide from hydroxymethyresorcinol
hydroxymethyresorcinol anion.

Figure 4. The calculated

4. The calculated structures
structures for the neutral quinonemethides (QM) and quinonemethide
quinonemethide
Figure 4. The calculated structures for the neutral quinonemethides (QM) and quinonemethide
anions (QMA).
anions (QMA).
Polymers 2017, 9, 426 7 of 14
Polymers 2017, 9, 426 7 of 14

Polymers 2017, 9, 426 7 of 14

Figure 5. The
The potential energy profiles for E1cb formations of quinonemethides (QM).

In our
In our recent
recent calculations
calculations on on PFPF condensation
condensation reactions reactions [18],[18], aa new
new mechanism
mechanism which which involves
involves
intra-molecular Figure
water 5.elimination
The potential energy
was profiles for E1cb
identified. For formations
RF, such of mechanism
a quinonemethides was (QM).
shown in Figure 1
intra-molecular water elimination was identified. For RF, such a mechanism was shown in Figure 1 as
as pathway (B).
pathway (B). Here,
Here,
only
only the reaction
the reaction
on carbon 6considered
on condensation
carbon 6 wasreactions
was considered as an example. The calculated
In our recent calculations on PF [18],as an example.
a new mechanism Thewhich
calculated
involves results
results are given in Figures 6 and 7. Note that the 6-HMRA (Figure 3) is an anion, while the
are given in Figures water
intra-molecular 6 andelimination
7. Note that theidentified.
was 6-HMRAFor (Figure 3) isa an
RF, such anion, while
mechanism the intermediate
was shown in Figure 1 (IM)
intermediate
as pathway (IM)(B).6-HMR-IM
Here, only theis areaction
neutralonspecies.
carbon 6Starting from this
was considered as an reactant-like
example. Theintermediate,
calculated a
6-HMR-IM is a neutral species. Starting from this reactant-like intermediate, a water molecule can
waterresults
molecule can beineliminated
are(WE)
given Figures (WE)7.via
6 and Notethethatfour-member
the 6-HMRA ring transition
(Figure 3) is state
an WE-TS, leading the to
be eliminated via the four-member ring transition state WE-TS, leading to anion,
6-QM.while This reaction
6-QM. This reaction encounters aisnotable barrier of Starting
160.4 kJ/mol, indicating that intermediate,
such a mechanism
encounters a notable barrier of 160.4 kJ/mol, indicating that such a mechanism cannot competea with
intermediate (IM) 6-HMR-IM a neutral species. from this reactant-like
cannot compete
water molecule with canE1cb at all. However,
be eliminated (WE) via the for four-member
PF, it was found that solvent
ring transition water molecules
state WE-TS, leading to can
E1cb at all. However, for PF, it was found that solvent water molecules can catalyze the proton
6-QM. This reaction encounters a notable barrier of 160.4 kJ/mol,
catalyze the proton transfer from carbon to oxygen and the energy barrier was significantly lowered indicating that such a mechanism
transfer from carbon to oxygen and the energy barrier was significantly lowered [13]. We call such
[13]. cannot
We call compete
such with E1cb at all. aHowever,
a mechanism water-aided for PF,elimination
it was found(WAE). that solvent It waswaterfound
molecules that can such a
a mechanism
catalyze the a water-aided
proton transferelimination
from carbon (WAE).to oxygen Itandwasthe found
energy that
barriersuch wasa mechanism
significantly loweredcan also be
mechanism can also be applied to RF. 6-W-HMR-IM1 is the complex formed through
applied[13].toWeRF.call
6-W-HMR-IM1
such a mechanism is the acomplex
water-aided formed through(WAE).
elimination inter-molecular
It was found hydrogen
that such bonding
a
inter-molecular hydrogen bonding between 6-HMR and a water molecule. Via the six-member ring
between 6-HMR and
mechanism can aalso waterbe molecule.
applied to ViaRF.the six-member
6-W-HMR-IM1 ringis transition
the complex state formed
WAE-TS,through the proton
transition state WAE-TS, the bonding
proton on carbon 6 migrates to water molecule. Meanwhile, aring
proton
on carbon 6 migrates
inter-molecular to water
hydrogen molecule.betweenMeanwhile,
6-HMR and a aproton on waterVia
water molecule. shifts
the to the oxygen
six-member atom
on water shiftsstate to the oxygen atom protonofonthe hydroxymethyl
6 migrates to group. Consequently, the aproduct-like
of thetransition
hydroxymethyl WAE-TS,
group.the Consequently, carbon
the product-like water molecule.
intermediate Meanwhile,
6-W-HMR-IM2 proton
is formed.
intermediate
on water 6-W-HMR-IM2
shifts tobarrier
the oxygen is atom
formed.of the WAE-TS
hydroxymethyl has energy barrier
group.lower of 98.2thekJ/mol
Consequently, product-likewhich is
WAE-TS has energy of 98.2 kJ/mol which is significantly than WE-TS by 62.2 kJ/mol,
significantly
intermediate 6-W-HMR-IM2 is formed. WAE-TS has energy barrier of 98.2 kJ/mol which strong
lower than WE-TS by 62.2 kJ/mol, indicating that solvent water has a very is
indicating that solvent water has a very strong catalytic effect. Elimination of a two-water complex via
catalytic effect. Elimination
significantly lower than of a two-water
WE-TS complex
by 62.2 kJ/mol, via 6-Disso-TS
indicating that solvent finally leads
water hastoa 6-QM.
very strongThis step
6-Disso-TS finally leads to 6-QM. This step is almost barrierless. The barrier of WAE is slightly higher
catalytic
is almost effect. Elimination
barrierless. The barrier of a two-water
of WAE iscomplex slightly viahigher
6-Disso-TS
thanfinally
that of leads to 6-QM.
E1cb This step
(6-E1cb-TS) by 4.7
than that of E1cb
is almost
(6-E1cb-TS)
barrierless. The
by 4.7 kJ/mol, implying
barrier of WAE
that such a mechanism may also contribute to
kJ/mol, implying that such a mechanism mayis alsoslightly higher than
contribute to the that of E1cb (6-E1cb-TS)
formation by 4.7
of 6-QM. Differently,
the formation of 6-QM.
kJ/mol, implying thatDifferently,
such a mechanism for PF, may
the WAE mechanism was found to of be energetically
Differently,more
for PF, the WAE mechanism was found toalsobecontribute to the
energetically formation
more favorable 6-QM.
than the E1cb
favorable
for PF, the WAE mechanism was found to be energetically more favorable than theonE1cb
than the E1cb mechanism [18]. Specifically, the barrier of the WAE reaction the para
mechanism [18]. Specifically, the barrier of the WAE reaction on the para position was found to be
position was found
mechanism [18].toSpecifically,
be 8.0 kJ/mol lower than
the barrier of thethatWAE of reaction
the E1cbonreaction.
the para This difference
position was found suggests
to be that
8.0 kJ/mol lower
8.0 kJ/mol lower
thanthan thatthat
of ofthetheE1cb reaction. This difference
difference suggests that the WAE and E1cb
the WAE and E1cb mechanisms play E1cb reaction.
different rolesThis for PF and RFsuggestssystems.that the WAE and E1cb
mechanisms
mechanisms playplaydifferent roles
different rolesforforPFPFand
andRF RFsystems.
systems.

Figure 6. The
Figure calculated
6. The structures
calculated structuresofofthe
theintermediates andtransition
intermediates and transition states
states forfor
thethe formation
formation of 6-QM
of 6-QM
(quinonemethide) via the WE (water elimination) and WAE (water-aided elimination)
(quinonemethide) via the WE (water elimination) and WAE (water-aided elimination) mechanisms. mechanisms.
Figure 6. The calculated structures of the intermediates and transition states for the formation of 6-QM
(quinonemethide) via the WE (water elimination) and WAE (water-aided elimination) mechanisms.
Polymers 2017, 9, 426 8 of 14
Polymers
Polymers 2017, 9, 426 88 of 14
Polymers 2017,
2017, 9,
9, 426
426 8 of
of 14
14

Figure
Figure 7.
7. The
The potential energy profiles for formation of 6-QM via the WE and WAE mechanisms.
The potential
potential energy profiles for formation of 6-QM via the WE and WAE
WAE mechanisms.
mechanisms.

Dissociation
Dissociation of of the
the two
two phenolic
phenolic hydroxyl
hydroxyl groups
groups of of resorcinol
resorcinol leads
leads toto aa dianion
dianion which
which should
should
be Dissociation of the two Therefore,
phenolic hydroxyl groups of resorcinol leads to a dianion which
be a stronger nucleophile. Therefore, hydroxymethylation of such a dianion can
a stronger nucleophile. hydroxymethylation of such a dianion can produce
produce
should be a stronger nucleophile.
hydroxymethylresorcinol Therefore, hydroxymethylation of such a dianion can produce
hydroxymethylresorcinol dianion dianion (HMRDA),
(HMRDA), from from which
which aa quinonemethide
quinonemethide intermediate
intermediate can can also
also
hydroxymethylresorcinol
be formed. According to dianion
the (HMRDA),
proposed mechanismfrom which
in a
Figurequinonemethide
1, the formed intermediate can should
quinonemethides also be
be formed. According to the proposed mechanism in Figure 1, the formed quinonemethides should
formed.
be According to the proposed mechanism in Figure 1,for the formed quinonemethides should be
be anions
anions like
like 2-QMA
2-QMA and and 4-QMA.
4-QMA. The The calculated
calculated results
results for thethe formation
formation of of them
them are
are shown
shown in in
anions
Figures like
8 2-QMA
and 9. and
The 4-QMA.
transition The
statescalculated
2-E1cb-TSA results
and for the formation
4-E1cb-TSA haveof them
energy are shown
barriers in
of Figures
76.7 and 8
Figures 8 and 9. The transition states 2-E1cb-TSA and 4-E1cb-TSA have energy barriers of 76.7 and
and
69.1 9. The transition states 2-E1cb-TSA and 4-E1cb-TSA have energy barriers of 76.7 and 69.1 kJ/mol,
69.1 kJ/mol,
kJ/mol, which
which are
are significantly
significantly lower
lower thanthan 2-E1cb-TS
2-E1cb-TS and and 4-E1cb-TS
4-E1cb-TS by by 37.0
37.0 and
and 24.4
24.4 kJ/mol,
kJ/mol,
which are significantly
respectively. lower than 2-E1cb-TS and 4-E1cb-TS by 37.0 and 24.4 kJ/mol, respectively.
respectively. Therefore, condensations related to 2-QMA and 4-QMA should be much faster.
Therefore, condensations related to 2-QMA and 4-QMA should be much faster.
Therefore,formation
Similarly, condensationsof related
4-QMA is to 2-QMA andmore
energetically 4-QMA shouldthan
favorable be much
2-QMA. faster. Similarly,
However, as formation
resorcinol
Similarly, formation of 4-QMA is energetically more favorable than 2-QMA. However, as resorcinol
of a4-QMA
is is energetically more favorable than 2-QMA. However, as resorcinol is a weak acid,
is a weak
weak acid,acid, the
the concentrations
concentrations of of resorcinol
resorcinol dianion
dianion (RDA)
(RDA) and and hydroxymethylresorcinol
hydroxymethylresorcinol
the concentrations
dianion of resorcinol dianion (RDA) and hydroxymethylresorcinol dianion (HMRDA)
dianion (HMRDA) are dependent on the concentration of the base or the base/resorcinol
(HMRDA) are dependent on the concentration of the base or the base/resorcinol molar
molar
are dependent
ratio. on the concentration of the base or the base/resorcinol molar ratio.
ratio.

Figure
Figure 8.
8. The calculated
The calculated structures
calculated structures for
structuresfor the
forthe intermediates
theintermediates and
intermediatesand transition
andtransition states
states
transition forfor
states the
the
for E1cb
E1cb
the formation
formation
E1cb of
formation
of quinonemethide
quinonemethide anions.
anions.
of quinonemethide anions.

The potential energy

Figure 9. The
Figure energy profiles for
for the E1cb
E1cb formation of
of quinonemethide anions (QMAa).
Figure 9.
9. The potential
potential energy profiles
profiles for the
the E1cb formation
formation of quinonemethide
quinonemethide anions
anions (QMAa).
(QMAa).

3.2. RF and
3.2. RF
RF and PRF
PRF Polyaddtion
Polyaddtion Reactions
Reactions
As ititwas
As it wasfound
was found
found forfor
for PF PF
PF reactions
reactions
reactions [18],[18],
[18], once once
once the the quinonemethide
the quinonemethide
quinonemethide is is formed,
is formed,
formed, aa Michael
Michael a addition
Michael
addition
addition
reaction reaction will take place between a neutral quinonemethide
reaction will take place between a neutral quinonemethide and a phenol anion or
will take place between a neutral quinonemethide and and
a a phenol
phenol anion
anion or aa
or
hydroxymethylphenol
hydroxymethylphenol
hydroxymethylphenol anion. anion. Therefore,
anion. Therefore, “polyaddition”
Therefore, “polyaddition”may
“polyaddition” may be
may be aaa more
be more appropriate
more appropriate word
appropriate word
word than
than
“polycondensation”
“polycondensation” to describe the reaction from a mechanistic viewpoint.
“polycondensation” to describe the reaction from a mechanistic viewpoint. Such a reaction was
reaction from a mechanistic viewpoint. Such
Such a reaction was
identified
identified to be much
to be much faster
faster than
than quinonemethide
quinonemethide formation. This feature
formation. This
This feature should
feature should still
should still be
still be kept
be kept
kept for
for
neutral
neutral RF quinonemethides. But does it still hold true for a quinonemethide anion, like 4-QMA? It
RF quinonemethides. But does it still hold true for a quinonemethide anion, like 4-QMA? It
is
is expected
expected that
that the
the reaction
reaction between
between two two anions
anions (or(or electron-rich
electron-rich species)
species) would
would encounter
encounter aa
Polymers 2017, 9, 426 9 of 14

Polymers 2017, 9, 426 9 of 14

neutral RF2017,
Polymers quinonemethides.
9, 426 But does it still hold true for a quinonemethide anion, like 4-QMA? It is
9 of 14
expected that the reaction between two anions (or electron-rich species) would encounter a repulsive
repulsive effect. Then, what if the condensation occurs between a quinonemethide anion and a
repulsive
effect. Then, effect.
what ifThen, what if the occurs
the condensation condensation
betweenoccurs between a quinonemethide
a quinonemethide anion
anion and a neutral and a
monomer?
neutral monomer? To answer these questions, the condensation reactions shown in Figure 10 were
Toneutral
answermonomer? To answer
these questions, the these questions,reactions
condensation the condensation
calculated and the results are shown in Figures 11 and 12.
shown inreactions shown
Figure 10 wereincalculated
Figure 10 and
werethe
calculated
results and the
are shown in results
Figuresare
11shown
and 12.in Figures 11 and 12.

Figure 10. The selected Michael addition reactions of resorcinol anion (RA) and neutral resorcinol
Figure The
10.10.
Figure selected
The Michael
selected addition
Michael additionreactions
reactionsofofresorcinol
resorcinolanion
anion(RA)
(RA)and
andneutral
neutralresorcinol
resorcinol(R)
(R) with quinonemethides.
(R)quinonemethides.
with with quinonemethides.

Figure TheThe
11.11.
Figure calculated structures
calculated of transition
structures statesstates
of transition and product-like intermediates
and product-like for thefor
intermediates Michael
the
Figure reactions
addition 11. The calculated
in Figure structures of transition states and product-like intermediates for the
10.
Michael addition reactions in Figure 10.
Michael addition reactions in Figure 10.
Polymers 2017, 9, 426 10 of 14
Polymers 2017, 9, 426 10 of 14

Figure
Figure 12.
12. The
The potential
potential energy
energy profiles
profiles for
for the
the Michael
Michael addition
addition reactions
reactions in
in Figure 10.
Figure 10.

Reactions (1) and (2) are the reactions of neutral 2-QM and 6-QM with a resorcinol anion.
Reactions (1) and (2) are the reactions of neutral 2-QM and 6-QM with a resorcinol anion.
Theoretical calculations predicted that they have energy barriers of 38.5 and 31.9 kJ/mol,
Theoretical calculations predicted that they have energy barriers of 38.5 and 31.9 kJ/mol, represented
represented by 2,6-Add-TS and 6,6′-Add-TS, respectively, which are much lower than those for
by 2,6-Add-TS and 6,60 -Add-TS, respectively, which are much lower than those for formations of the
formations of the quinonemethides. Therefore, they are indeed faster reactions. As the formation of
quinonemethides. Therefore, they are indeed faster reactions. As the formation of quinonemethide is
quinonemethide is the rate-determining step, the overall condensation reaction can be identified as
the rate-determining step, the overall condensation reaction can be identified as having a unimolecular
having a unimolecular pattern. The formed intermediate products 6,6′-IM and 2,6-IM can evolve to
pattern. The formed intermediate products 6,60 -IM and 2,6-IM can evolve to 6-60 -dimer (4,40 -dimer)
6-6′-dimer (4,4′-dimer) and 2,6-dimer (2,4-dimer), respectively. Several steps of proton transfers or
and 2,6-dimer (2,4-dimer), respectively. Several steps of proton transfers or migrations may occur in
migrations may occur in this process, but they are generally fast reactions. Another reaction leading
this process, but they are generally fast reactions. Another reaction leading to the 2,4-dimer is the one
to the 2,4-dimer is the one between 4-QM or 6-QM and the 2-position of the resorcinol anion. But,
between 4-QM or 6-QM and the 2-position of the resorcinol anion. But, the situation should be similar
the situation should be similar to reaction (2).
to reaction (2).
Differently, reaction (3), the condensation of 4-QMA with neutral resorcinol encounters a
Differently, reaction (3), the condensation of 4-QMA with neutral resorcinol encounters a notable
notable barrier of 103.4 kJ/mol which significantly exceeded the barrier of 4-QMA formation (69.1
barrier of 103.4 kJ/mol which significantly exceeded the barrier of 4-QMA formation (69.1 kJ/mol) and
kJ/mol) and became the rate-determining step. This must be caused by the much lower nucleophilic
became the rate-determining step. This must be caused by the much lower nucleophilic reactivity of
reactivity of neutral resorcinol. It was not our expectation that reaction (4) would encounter a much
neutral resorcinol. It was not our expectation that reaction (4) would encounter a much lower barrier
lower barrier of 63.1 kJ/mol. Although the repulsive effect makes the barrier higher than that of
of 63.1 kJ/mol. Although the repulsive effect makes the barrier higher than that of reactions (1) and
reactions (1) and (2), the barrier is still lower than that of the formation of 4-QMA (69.1 kJ/mol).
(2), the barrier is still lower than that of the formation of 4-QMA (69.1 kJ/mol). Therefore, such a
Therefore, such a condensation pattern is feasible. But, the overall reaction may not be a typical
condensation pattern is feasible. But, the overall reaction may not be a typical unimolecular mechanism
unimolecular mechanism because the 4-QMA formation and following Michael addition have
because the 4-QMA formation and following Michael addition have similar energy barriers. In fact,
similar energy barriers. In fact, 4-QMA may not directly participate in the condensation because it can
4-QMA may not directly participate in the condensation because it can convert to 4-QM or 6-QM
convert to 4-QM or 6-QM through fast protonation of one of the carbonyl groups. Thus, the pathway
through fast protonation of one of the carbonyl groups. Thus, the pathway with the lowest energy
with the lowest energy barrier should be 4-HMRDA→4-QMA→protonation→4-QM (or 6-QM).
barrier should be 4-HMRDA→4-QMA→protonation→4-QM (or 6-QM).
So far, there is no experiment that compares PF and RF condensation reaction rates under the
So far, there is no experiment that compares PF and RF condensation reaction rates under the
same conditions (the same F/P or F/R molar ratio, catalyst concentration, and temperature) that has
same conditions (the same F/P or F/R molar ratio, catalyst concentration, and temperature) that
been reported. But, the calculated potential energy barriers for PF and RF quinonemethides allow us
has been reported. But, the calculated potential energy barriers for PF and RF quinonemethides
to make comparisons since the formation of these intermediates is the rate-determining step.
allow us to make comparisons since the formation of these intermediates is the rate-determining step.
According to our theoretical calculations of PF [18], the lowest energy barrier for E1cb formation of
According to our theoretical calculations of PF [18], the lowest energy barrier for E1cb formation of
neutral para-quinonemethide was calculated to be 99.7 kJ/mol, which is higher than that of 6-QM
neutral para-quinonemethide was calculated to be 99.7 kJ/mol, which is higher than that of 6-QM
(93.5 kJ/mol) by 6.2 kJ/mol, but the WAE mechanism for PF has a lower barrier of 87.8 kJ/mol. In
(93.5 kJ/mol) by 6.2 kJ/mol, but the WAE mechanism for PF has a lower barrier of 87.8 kJ/mol.
this sense, RF condensations are not faster. But, taking into account the participation of the
In this sense, RF condensations are not faster. But, taking into account the participation of the
quinonemethidesanions like 4-QMA, RF condensations should be much faster since the energy
quinonemethidesanions like 4-QMA, RF condensations should be much faster since the energy barrier
barrier for its formation is significantly lower than the neutral PF QMs by 20–30 kJ/mol. Such a big
for its formation is significantly lower than the neutral PF QMs by 20–30 kJ/mol. Such a big difference
difference is sufficient to explain the fact that the RF resin cure is much faster than PF resin, and that
RF resin can be used as PF cure accelerator.
Polymers 2017, 9, 426 11 of 14

is sufficient to explain the fact that the RF resin cure is much faster than PF resin, and that RF resin can
be used as PF cure accelerator.
Polymers
Polymers 2017,
2017, 9,
9, 426
426 11
11 of
of 14
14
Synthesis of phenol-resorcinol-formaldehyde (PRF) co-condensed resin has been a hot
topic [1–3,5,6,8].
Synthesis
Synthesis Inof
ofpreparation of traditional phenol-resorcinol-formaldehyde
phenol-resorcinol-formaldehyde
phenol-resorcinol-formaldehyde (PRF)
(PRF) co-condensed
co-condensed resin resin has resin,
has been
been resorcinol
aa hot
hot topic
topic is
generally added
[1–3,5,6,8].
[1–3,5,6,8]. In into the PF
In preparation resol
preparation of resin. It
of traditional is believed that resorcinol
traditional phenol-resorcinol-formaldehyde can
phenol-resorcinol-formaldehyde resin, react with the hydroxymethyl
resin, resorcinol
resorcinol is is
groupsgenerally added
of PF resin,
generally added into
into the
forming P–CH
the PF
PF 2resol resin.
resin. It
–R methylene
resol It is
is believed
linkages
believed [3].that
Withresorcinol
that can
can react
such a pattern,
resorcinol with
with the
resorcinol
react can be
the
hydroxymethyl
connected to the end
hydroxymethyl groups
of PF
groups of PF
PF resin,
of chains. forming
forming P–CH
resin, Conversely, P–CH 22–R
can methylene
–R phenol
methylene linkages
react with [3].
linkages theWith
[3]. such
such aa pattern,
hydroxymethyl
With pattern, group
resorcinol
resorcinol can
can be
be connected
connected to
to the
the end
end of
of PF
PF chains.
chains. Conversely,
Conversely,
on resorcinol? What will happen if phenol, resorcinol, and formaldehyde are mixed together? can
can phenol
phenol react
react with
with the
the
Howhydroxymethyl
hydroxymethyl
do co-condensations group
group on on resorcinol?
resorcinol?
compete withWhat
What will
will happen
happen if
self-condensations? if phenol,
phenol,
These resorcinol,
resorcinol,
issues have and formaldehyde
and never
formaldehyde are
are
been addressed
mixed
mixed together?
together? How
How do
do co-condensations
co-condensations compete
compete with
with self-condensations?
self-condensations? These
These issues
issues have
have
mechanistically in detail. According to our understanding, the co-condensation efficiency is not only
never
never been been addressed
addressed mechanistically
mechanistically in in detail.
detail. According
According to to ourour understanding,
understanding, the the
dependent on the rates of PF and
co-condensation RF quinonemethides formation, but also on the selectivity of these
co-condensation efficiency
efficiency is is not
not only
only dependent
dependent on on the
the rates
rates ofof PF
PF and
and RF RF quinonemethides
quinonemethides
intermediates
formation, toward PF and RF monomers. Namely, both PF and RF and quinonemethide intermediates
formation, but but also
also on
on the
the selectivity
selectivity of of these
these intermediates
intermediates toward toward PF PF and RF RF monomers.
monomers. Namely,
Namely,
can be formed
both
both PFPF and under
and RF alkaline
RF quinonemethide conditions
quinonemethide intermediates and
intermediates can their
can be relative
be formed formation
formed under
under alkalinerates and
alkaline conditions reactivity
conditions and and theirtoward
their
phenol, resorcinol,
relative
relative formation
formation andrates
theirand
rates hydroxymethyl
and reactivity
reactivity toward compounds
toward phenol, determines
phenol, resorcinol,
resorcinol, and the their
and competitive
their hydroxymethylformations
hydroxymethyl
compounds
of self-
compounds determines
determines the
and co-condensed competitive
polymers.
the competitive To formations
address
formations thisof
of self-
issue,
self- and
andwe co-condensed
carried outpolymers.
co-condensed To
To address
further calculations
polymers. address on
this
this issue,
issue, we
the representative we carried out
out further
further calculations
PRF co-polyaddition
carried reactions
calculations on the
the representative
on which are shownPRF
representative co-polyaddition
in Figure
PRF reactions
13 as reactions
co-polyaddition reactions (5)–(7).
which are shown
which are shown
The calculated results in Figure
inare
Figure 13 as
13 as
given reactions
inreactions (5)–(7).
Figures(5)–(7).
14 andTheThe
15.calculated results are given in Figures 14 and
calculated results are given in Figures 14 and 15.
15.

Figure 13. The

Figure
Figure 13.
Michael
13. The
addition
The Michael
Michael addition
reactions
addition reactions
of
reactions of
phenol
of phenol
anion(PA)
phenol anion
(PA)with
anion (PA)
withresorcinol
resorcinol
with resorcinol
quinonemethides
quinonemethides
quinonemethides
(6-QM and
(6-QM 4-QMA) and the reaction of resorcinol anion (RA) with phenol quinonemethide
(6-QM and 4-QMA) and the reaction of resorcinol anion (RA) with phenol quinonemethide (para-QM).
and 4-QMA) and the reaction of resorcinol anion (RA) with phenol quinonemethide (para-QM).
(para-QM).

Figure
Figure 14.
14. The
The calculated
calculated structures
structures of
of the
the transition
transition states
states and
and product-like
product-like intermediates
intermediates for
for the
the
Figure 14. The calculated structures of the transition states and product-like intermediates for the
Michael
Michael addition
addition reactions
reactions in
in Figure
Figure 13.
13.
Michael addition reactions in Figure 13.
Polymers 2017, 9, 426 12 of 14
Polymers 2017, 9, 426 12 of 14

Figure15.
Figure Thepotential
15.The potentialenergy
energyprofiles
profiles for
for the
the Michael
Michael addition
addition reactions
reactions in
in Figure
Figure 13.
13.

Reaction
Reaction (5) (5) is
is the
the addition
addition between
between the the phenol
phenol anion
anion and
and resorcinol
resorcinol quinonemethide
quinonemethide(6-QM). (6-QM).
The calculated energy barrier for this reaction is 39.8
The calculated energy barrier for this reaction is 39.8 kJ/mol, which is higherkJ/mol, which is higher than reaction (1) by 7.9
than reaction (1)
kJ/mol, and higher
by 7.9 kJ/mol, andthan higherthethan addition of the phenol
the addition of the anion
phenolwithanionphenol
with quinonemethide
phenol quinonemethide by 13.2
kJ/mol
by 13.2 [18].
kJ/mol This[18].result indicates
This result that that
indicates the the co-condensation
co-condensation reaction
reaction between
betweenphenol
phenol and and
hydroxymethylresorcinol
hydroxymethylresorcinol is is less
less competitive
competitivethan thanPF PFandandRF RFself-condensations.
self-condensations. Similarly,
Similarly, reaction
reaction (6)
(6) encounters
encounters a higher
a higher barrierbarrier than self-condensations.
than self-condensations. In contrast,
In contrast, reactionreaction
(7) has the(7) lowest
has the lowest
barrier of
barrier of 27.0 kJ/mol, which is lower than the barrier of reaction (1) by 12.8
27.0 kJ/mol, which is lower than the barrier of reaction (1) by 12.8 kJ/mol and very close to the barrier kJ/mol and very close to
the barrier
of PF of PF self-condensation
self-condensation (26.6 kJ/mol), (26.6suggesting
kJ/mol), suggesting that the co-condensation
that the co-condensation between the between the
resorcinol
resorcinol anion and phenol quinonemethide is energetically more
anion and phenol quinonemethide is energetically more competitive. In this context, if resorcinol, competitive. In this context, if
resorcinol, phenol, and formaldehyde are mixed together, hydroxymethylation
phenol, and formaldehyde are mixed together, hydroxymethylation of resorcinol should be faster than of resorcinol should
be faster than hydroxymethylation
hydroxymethylation of phenol due of phenolreactivity
to higher due to higher reactivity
of resorcinol. Inof resorcinol.
addition In addition
to faster formation to
faster formation of resorcinol quinonemethides, RF self-condensation
of resorcinol quinonemethides, RF self-condensation occurs prior to PRF co-condensation, but it is occurs prior to PRF
co-condensation,
unlikely that the PRF but reactions
it is unlikely can bethat the PRFsince
excluded reactions canbetween
the gap be excluded since barriers
the energy the gap are
between the
moderate.
energy Asbarriers
mentioned are in
moderate.
the literature [3], a general strategy in the synthesis of PRF is to let resorcinol react
withAs mentioned in the literature
hydroxymethylphenol or with [3], a general
PF resin in excessstrategy in the synthesis
of un-reacted of PRF isgroups.
hydroxymethyl to let resorcinol
In such a
react with
type of hydroxymethylphenol
resin, the polymers are resorcinol or with PF resin in excess
terminated. In theof un-reacted
curing process, hydroxymethyl
condensationsgroups.can occurIn
such a type of resin, the polymers are resorcinol terminated. In the
between terminal resorcinol moieties, or between phenol and hydroxymethylatedresorcinol. In both curing process, condensations
can occur
cases, fast formationbetween terminal quinonemethides
of the resorcinol resorcinol moieties, or tobetween
can contribute the curing phenol
condensations.and
hydroxymethylatedresorcinol.
This is why PRF resin bears theIn both cases,characteristic.
cold-setting fast formation of the resorcinol quinonemethides can
contribute to the curing condensations. This is why PRF resin bears the cold-setting characteristic.
4. Conclusions
4. Conclusions
Base-catalyzed resorcinol-formaldehyde condensation reactions were theoretically investigated
Base-catalyzed
in this study by employing resorcinol-formaldehyde
a quantum chemistry condensation
method.reactions were theoretically
The experimentally observedinvestigated
polymer
in this study
structures wereby rationalized
employing aaccording quantum to chemistry
the identifiedmethod. The experimentally
mechanisms observedenergetics.
and the calculated polymer
structures
The resultswere
wererationalized
discussed and according
compared to the
with identified mechanisms
those obtained and the
previously calculated energetics.
for phenol-formaldehyde
The resultsThe
reactions. were maindiscussed and compared
conclusions have beenwith drawn those obtained previously for phenol-formaldehyde
as follows:
reactions. The main conclusions have been drawn as follows:
(1) The condensation reaction includes two steps: Formation of a quinonemethide (QM) intermediate
(1) Thefromcondensation reaction includes
hydroxymethylresorcinol, followedtwo steps: Formation
by Michael addition betweenof a the quinonemethide
quinonemethide (QM)
and
intermediate
the resorcinolfrom anion.hydroxymethylresorcinol, followed by Michael addition between the
(2) quinonemethide
Two mechanisms, and the resorcinol
unimolecular anion. of conjugate base (E1cb) and water-aided elimination
elimination
(2) Two(WAE), mechanisms,
were identified unimolecular elimination
for the formation of conjugate Hydroxymethylresorcinol
of quinonemethide. base (E1cb) and water-aided anion
elimination (WAE), were identified for the
produces a neutral quinonemethide while the dianion produces a quinonemethide anion. formation of quinonemethide.
Hydroxymethylresorcinol
The calculated potential energy anion produces
barriers suggested a neutral quinonemethide
that quinonemethide anion while the dianion
formation is much
produces a quinonemethide
more favorable. anion. The calculated potential
Although resorcinol-formaldehyde energy barriers suggested
and phenol-formaldehyde share common that
quinonemethide anion formation is much more favorable. Although
condensation mechanisms, the former would be faster if the quinonemethide anion is involved. resorcinol-formaldehyde
(3) and
The phenol-formaldehyde
potential energy barriers share
for common
formationcondensation
of 2-QM, 4-QM, mechanisms,
6-QM, 2-QMA, the former
and 4-QMA would be
were
faster if the quinonemethide anion is involved.
calculated. The results show that the formation of 6-QM and 4-QMA have relatively lower
Polymers 2017, 9, 426 13 of 14

energy barriers. In addition to higher reactivity of the position-6 hydroxymethylation reaction,

the previous experimental observation that the 2,4-(2,6-) and 6,60 -(4,40 -) methylene linkages were
dominant polycondensed structures has been rationalized.
(4) The cold-setting characteristic of phenol-resorcinol-formaldehyde co-condensed resin can be
attributed to participation of resorcinol quinonemethides in condensations.

Acknowledgments: This work was supported by the Programs of the National Natural Science Foundation of
China (NSFC) (31360159 and 51273163).
Author Contributions: Taohong Li and Guanben Du contributed to the theoretical calculations and editing of
the paper. Ming Cao and Jiankun Liang contributed to the collection of references and editing of the paper.
Xiaoguang Xie contributed to the analysis of the calculated results.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Chow, S.; Steiner, P.R. Determination of resorcinol content in phenol-resorcinol-formaldehyde resins by
infrared spectrometry. Holzforschung 1978, 32, 120–122. [CrossRef]
2. Kim, M.G.; Amos, W.L.; Barnes, E.E. Investigation of a resorcinol-formaldehyde resin by l3 C-NMR
spectroscopy and intrinsic viscosity measurement. J. Polym. Sci. A 1993, 31, 1871–1877. [CrossRef]
3. Scopelitis, E.; Pizzi, A. The chemistry and development of branched PRF wood adhesives of low
resorcinol content. J. Appl. Polym. Sci. 1993, 47, 351–360. [CrossRef]
4. Christiansen, A.W. Resorcinol-formaldehyde Reactions in dilute solution observed by C13 NMR spectroscopy.
J. Appl. Polym. Sci. 2000, 75, 1760–1768. [CrossRef]
5. Pizzi, A.; Pasch, H.; Simon, C.; Rode, K. Structure of resorcinol, phenol, and furan resins by MALDI-TOF
mass spectrometry and 13 C NMR. J. Appl. Polym. Sci. 2004, 92, 2665–2674. [CrossRef]
6. Yang, I.; Ahn, S.; Choi, I.; Kim, H.Y.; Oh, S. Adhesives formulated with chemically modified okara and
phenol-resorcinol-formaldehyde for bonding fancy veneer onto high-density fiberboard. J. Ind. Eng. Chem.
2009, 15, 398–402. [CrossRef]
7. Sauget, A.; Zhou, X.; Pizzi, A. Tannin-resorcinol-formaldehyde resin and flax fiber biocomposites.
J. Renew. Mater. 2014, 2, 173–181. [CrossRef]
8. Liu, J.; Chen, R.; Xu, Y.; Wang, C.P.; Chu, F. Resorcinol in high solid phenol-formaldehyde resinsfor
foams production. J. Appl. Polym. Sci. 2017, 134, 44881.
9. Lin, C.; Ritter, J.A. Effect of synthesis pH on the structure of carbon xerogels. Carbon 1997, 35, 1271–1278.
[CrossRef]
10. Liang, C.; Sha, G.; Guo, S. Resorcinol-formaldehyde aerogels prepared by supercritical acetone drying.
J. Non-Cryst. Solids 2000, 271, 167–170. [CrossRef]
11. Al-Muhtaseb, S.A.; Ritter, J.A. Preperation and properties of resorcinol-formaldehyde organic and carbon gels.
Adv. Mater. 2003, 15, 101–114. [CrossRef]
12. Yu, J.; Guo, M.; Muhammad, F.; Wang, A.; Zhang, F.; Li, Q.; Zhu, G. One-pot synthesis of highly ordered
nitrogen-containing mesoporous carbon with resorcinol-urea-formaldehyde resin for CO2 capture. Carbon
2014, 69, 502–514. [CrossRef]
13. Pizzi, A.; Eaton, N.J. A conformational analysis approach to phenol-formaldehyde resins adhesion to
wood cellulose. J. Adhes. Sci. Technol. 1987, 1, 191–200. [CrossRef]
14. Pizzi, A.; De Sousa, G. On the resolution of dihydroxydiphenylmethanes on achiral crystalline Cellulose
II—Correlation of experimental and calculated results. Chem. Phys. 1992, 164, 203–216. [CrossRef]
15. Pizzi, A.; Maboka, S. Calculated values of the adhesion of phenol-formaldehyde oligomers to
crystalline CelluloseI. J. Adhes. Sci. Technol. 1993, 7, 81–94. [CrossRef]
16. Sedano-Mendoza, M.; Alnarran-Lopez, P.; Pizzi, A. Natural lignans adhesion to cellulose: Computational vs
experimental results. J. Adhes. Sci.Technol. 2010, 24, 1769–1786. [CrossRef]
17. Alnarran-Lopez, P.; Pizzi, A.; Navarro-Santos, P.; Hernandes-Esparza, R.; Garza, J. Oligolignols within
lignin-adhesive formulations drive their Young’s modulus: A ReaxFF-MD study. Int. J. Adhes. Adhes. 2017.
[CrossRef]
Polymers 2017, 9, 426 14 of 14

18. Li, T.; Cao, M.; Liang, J.; Xie, X.; Du, G. Theoretical confirmation of the quinonemethide hypothesis for the
condensation reactions in phenol-formaldehyde resin synthesis. Polymers 2017, 9, 45. [CrossRef]
19. Hohenberg, P.; Kohn, W. Inhomogeneous electron gas. Phys. Rev. 1964, 136, B864–B871. [CrossRef]
20. Kohn, W.; Sham, L.J. Self-consistent equations including exchange and correlation effects. Phys. Rev. 1965,
140, A1133–A1138. [CrossRef]
21. Lee, C.; Yang, W.; Parr, R.G. Development of the Colle-Salvetti correlation-energy formula into a functional
of the electron density. Phys. Rev. B 1988, 37, 785–789. [CrossRef]
22. Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Results obtained with the correlation-energy density 17.
functionals of Becke and Lee, Yang and Parr. Chem. Phys. Lett. 1989, 157, 200–206. [CrossRef]
23. Becke, A.D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98,
5648–5652. [CrossRef]
24. Ditchfield, R.; Hehre, W.J.; Pople, J.A. Self-consistent molecular orbital methods. 9. Extended Gaussian-type
basis for molecular-orbital studies of organic molecules. J. Chem. Phys. 1971, 54, 724. [CrossRef]
25. Hehre, W.J.; Ditchfield, R.; Pople, J.A. Self-consistent molecular orbital methods. 12. Further extensions of
Gaussian-type basis sets for use in molecular-orbital studies of organic-molecules. J. Chem. Phys. 1972, 56,
2257–2261. [CrossRef]
26. Clark, T.; Chandrasekhar, J.; Spitznagel, G.W.; Schleyer, P.V. Efficient diffuse function-augmented basis-sets
for anion calculations. 3. The 3-21+G basis set for 1st-row elements, Li-F. J. Comput. Chem. 1983, 4, 294–301.
[CrossRef]
27. Frisch, M.J.; Pople, J.A.; Binkley, J.S. Self-consistent molecular orbital methods. 25. Supplementary functions
for Gaussian Basis Sets. J. Chem. Phys. 1984, 80, 3265–3269. [CrossRef]
28. Miertus, S.; Tomasi, J. Approximate evaluations of the electrostatic free energy and internal energy changes
in solution processes. J. Chem. Phys. 1982, 65, 239–245. [CrossRef]
29. Miertus, S.; Scrocco, E.; Tomasi, J. Electrostatic interaction of a solute with a continuum. A direct utilization
of AB initio molecular potentials for the prevision of solvent effects. J. Chem. Phys. 1981, 55, 117–129.
30. Cossi, M.; Barone, V.; Mennucci, B. Ab initio study of ionic solutions by a polarizable continuum
dielectric model. Chem. Phys. Lett. 1998, 286, 253–260. [CrossRef]
31. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Montgomery, J.A., Jr.;
Vreven, T.; Kudin, K.N. Gaussian 03, Revision B.01, Software for Quantum Chemistry Calculation; Gaussian, Inc.:
Pittsburgh, PA, USA, 2003.
32. Durairaj, R.B. Resorcinol, Chemistry, Technology and Applications; Springer: Berlin/Heidelberg, Germany;
New York, NY, USA, 2005.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).