novel liquid extraction method for detecting native-wood ......forest products charles e. frazier...
TRANSCRIPT
Novel Liquid Extraction Method for Detecting Native-wood Formaldehyde
Mohammad Tasooji
Thesis submitted to the faculty of the Virginia Polytechnic Institute and State
University in partial fulfillment of the requirements for the degree of
Master of Science
in
Forest Products
Charles E. Frazier
Maren Roman
Scott H. Renneckar
May 15th , 2014
Blacksburg, VA
Keywords: Native-wood formaldehyde emission, liquid extraction, poly (allylamine) beads,
water extraction
Novel Liquid Extraction Method for Detecting Native-wood Formaldehyde
Mohammad Tasooji
Abstract
New vigorous regulations have been established for decreasing the allowable formaldehyde
emissions from nonstructural wood based composites. Two main sources of formaldehyde
emission in non-structural wood based composites are adhesive and wood. Adhesives are quite
well known and great efforts have been conducted to decrease their formaldehyde content;
however formaldehyde emission from wood has received little attention and it is not completely
understood. Wood-borne formaldehyde emission exists in a complex equilibrium in wood
matrix. The reaction between formaldehyde and wood hydroxyl groups/water can hinder the
complete formaldehyde extraction. In order to have a complete formaldehyde extraction, a
stronger nucleophile than hydroxyl and water groups is needed.
In this study cross-linked poly (allylamine) (PAA) beads were synthesized and used as a
strong nucleophile to extract all the biogenic and synthetic free-formaldehyde within the woody
matrix of never-heated and heat-treated Virginia pines; the results were compared to simple
water extraction. A new formaldehyde capturing device was also developed using a serum bottle.
Results showed that there was no advantage of using PAA beads over simple water
extraction for extracting woody matrix free-formaldehyde. This means that simple water
extraction can extract all the free-formaldehyde from the woody matrix. It was also found that
thermal treatment resulted in generating more wood-borne formaldehyde. The other important
finding was the new developed formaldehyde capturing device. The device was very promising
for detecting wood-borne formaldehyde from very small pieces of wood (5-70 mg) and can be
very useful in future studies.
iii
Acknowledgment
Foremost I would like to thank my advisor Prof. Charles E. Frazier for his continuous support,
his enthusiasm, patience, and motivation throughout this study. He is for sure the best teacher I
have ever had; I have learned a lot from him and I’m still learning.
I would also like to thank my thesis committee members Prof. Maren Roman and Prof. Scott
Renneckar for their support and insightful comments.
The friendship, kindness and support of my friends: Xing Yang, Guigui Wan, Wei Zhang, Kyle
Mirable, and Sarah Blosch definitely helped me a lot throughout this study. This thesis would not
have been done without the help of David Jones. I would also like to acknowledge Doctor Ann
Norris and Debbie Garnand for their support and kindness.
Without the endless love, care and support of my dear wife and best friend Sarah this work
would not be possible. I also like to thank my dear family for their all-time support, kindness,
and love.
iv
Table of Contents
Abstract ......................................................................................................................................................... ii
Acknowledgment ......................................................................................................................................... iii
1. Introduction .......................................................................................................................................... 1
2. Literature review ................................................................................................................................... 2
2.1. Wood-based composites formaldehyde emission ....................................................................... 2
2.1.1. Structural wood-based composites ...................................................................................... 2
2.1.2. Non-structural wood-based composites ............................................................................ 11
2.2. Formaldehyde regulations .......................................................................................................... 17
2.2.1. Test methods for formaldehyde emission .......................................................................... 19
2.2.2. Formaldehyde analysis ........................................................................................................ 21
2.3. Formaldehyde emission from solid wood, a review ................................................................... 24
2.3.1. Effect of wood species .............................................................................................................. 24
2.3.2. Effect of wood maturity ............................................................................................................ 24
2.3.3. Effect of wood components ...................................................................................................... 25
2.3.4. Effect of thermal treatment ...................................................................................................... 25
2.3.5. Effect of moisture content and particle size ............................................................................. 26
2.4. Wood-Formaldehyde interaction ............................................................................................... 26
2.5. Cross-linked Poly allylamine as formaldehyde adsorbent .......................................................... 28
2.6. References .................................................................................................................................. 30
3. Novel Liquid extraction method for detecting Native-wood Formaldehyde ..................................... 38
3.1. Introduction ................................................................................................................................ 38
3.2. Experimental ............................................................................................................................... 41
3.2.1. Materials ............................................................................................................................. 41
3.2.2. Methods .............................................................................................................................. 42
3.3. Results and discussion ................................................................................................................ 46
3.4. Conclusions ................................................................................................................................. 58
3.5. References .................................................................................................................................. 58
4. Appendices .......................................................................................................................................... 62
Appendix A : The effect of HMR on the rheological and physical properties of wood .......................... 62
Appendix B : The effect of Ionic liquid on lignocellulose rheology ......................................................... 81
v
LIST OF FIGURES
Figure 1-1. Complex equilibria of the biogenic and synthetic formaldehyde exists in woody matrix ......... 1
Figure 2-1. The reaction between phenol and formaldehyde under alkaline condition results in methylol
which converts to quinone methide and/or Benzylcarboinium. (Benzylcarbonium formation is more
favorable than quinone methide) ................................................................................................................. 4
Figure 2-2. The reactions of quinone methide and benzylcarbonium with phenoxide and methylol;
results in linear polymer with methylene and methylene ether linkages. ................................................... 5
Figure 2-3. Highly branched Resol after curing at elevated temperature. ................................................... 6
Figure 2-4. Curing Resol at elevated temperature results in converting methylene ether linkages into
methylene linkages and formaldehyde elimination. .................................................................................... 6
Figure 2-5. The possible structures result from phenol-formaldehyde reactions (Myers, 1986b). ............. 7
Figure 2-6. The confirmed structures existed in cured PF resin (Maciel et al., 1984) .................................. 7
Figure 2-7. Synthesis reaction of pMDI (Pizzi and Mittal, 2003) ................................................................... 9
Figure 2-8. pMDI reactions with woody matrix hydroxyl groups and water which results in urethane and
urea respectively (Pizzi and Mittal, 2003) ................................................................................................... 10
Figure 2-9. Some of the possible urea-formaldehyde reactions ................................................................ 12
Figure 2-10. The structures available in cured UF resin which are subjected to hydrolysis ....................... 13
Figure 2-11. The hydrolysis of methylene ether bond which results in formaldehyde generation ........... 13
Figure 2-12. Reaction between formaldehyde and acetylacetone reagent results in DDL ........................ 22
Figure 2-13. Reaction between formaldehyde and sodium sulfite in water results in sodium hydroxide. 23
Figure 2-14. Xylose forms oxymethylfufural and further furfural and formaldehyde at elevated
temperature ................................................................................................................................................ 26
Figure 2-15. Thermal degradation of xylose results in acetic acid and formaldehyde ............................... 26
Figure 2-16. The reactions of formaldehyde monomer (methylene glycol) with cellulose which results in
hemiacetals and acetals. ............................................................................................................................. 27
Figure 2-17. Cross-linked poly allylamine beads ......................................................................................... 29
Figure 2-18. PAA-formaldehyde adsorption (up) and desorption (down) reactions .................................. 29
Figure 3-1. Complex equilibria of the Biogenic formaldehyde exists in woody matrix .............................. 40
Figure 3-2. Harvesting Virginia pine (P.Viginiana), Fishburn forest Blacksburg Virginia ............................ 41
Figure 3-3. Cross-section of the Virginia Pine (P.Viginiana) log containing compression wood ................ 41
Figure 3-4. Formaldehyde reaction with PAA beads to form the hemiaminal, where after dehydration
leads to the imine. ...................................................................................................................................... 47
Figure 3-5. Calibration curve used for this study (slope=376.26). .............................................................. 48
Figure 3-6. Acidolytic cleavage of the β-O-4 lignin linkage leading to formaldehyde release. .................. 52
Figure 3-7. Depiction of how other lignin linkages can also lead to formaldehyde release through acid
catalysis. ...................................................................................................................................................... 53
Figure 3-8. Demonstration of PAA bead color change observed during the sorption phase for never-
heated specimens. ...................................................................................................................................... 55
vi
Figure 3-9. Comparison of the recovered formaldehyde from Virginia pine specimens with and without
using the PAA beads, never-heated and 200°C heat treated ..................................................................... 56
Figure 3-10. Depiction of lignocellulose hemiacetal formation and subsequent hydration/hydrolysis that
favors formaldehyde release from the woody matrix. ............................................................................... 56
Figure 4-1. The effect of wood species on dynamic mechanical responses of DMG-plasticized HMR-
treated specimens. Average first heating scans of pine (top) and poplar (bottom) specimens. Error bars
represent ±1 standard deviation (n=3). ...................................................................................................... 69
Figure 4-2. The effect of wood species on dynamic mechanical response of DMF-plasticized HMR-treated
specimens. Average first cooling scans of pine (top) and poplar (bottom) specimens. Error bars represent
±1 standard deviation (n=3). ....................................................................................................................... 69
Figure 4-3. The effect of wood species on dynamic mechanical response of NMP-plasticized HMR-
treated specimens. Average first heating scans of pine (top) and poplar (bottom) specimens. Error bars
represent ±1 standard deviation (n=3). ...................................................................................................... 70
Figure 4-4. The effect of wood species on dynamic mechanical response of NMP-plasticized HMR-
treated specimens. Average first cooling scans of pine (top) and poplar (bottom) specimens. Error bars
represent ±1 standard deviation (n=3). ...................................................................................................... 70
Figure 4-5. Softwood and hardwood lignin monomers .............................................................................. 73
Figure 4-6. The effect of DMF soaking time on the storage modulus at Tg (up) and the Tg (bottom) of
HMR-treated poplar. Error bars represent ±1 standard deviation (n=3). .................................................. 74
Figure 4-7. The effect of NMP soaking time on the storage modulus at Tg (up) and the Tg (bottom) of
HMR-treated poplar. Error bars represent ±1 standard deviation (n=3). .................................................. 75
Figure 4-8. The volumetric swelling (%) of neat-HMR samples immersed in DMF and NMP plasticizers as
a function of soaking time. Error bars represent ±1 standard deviation (n=3). ......................................... 76
Figure 4-9. Neat-HMR samples after 9 days soaked in DMF and NMP solvents ........................................ 76
Figure 4-10. Comparison between the volumetric swelling (%) of HMR-treated (25⁰C, and 70⁰C) poplar
and pine after 48 hr soaking in water. Error bars represent ±1 standard deviation (n=4). ....................... 78
Figure 4-11. DMA scans of IL-DMA specimens as a function of grain orientation. Average first heating
scans are shown with error bars representing ±1 standard deviation (n=3). ............................................ 84
Figure 4-12. The gap changing of ionic liquid saturated specimens (IL-DMA) as a function of grain
orientation. Average first heating scans are shown with error bars representing ±1 standard deviation
(n=3). ........................................................................................................................................................... 85
Figure 4-13. Disintegrated RT and TR specimens after DMA temperature ramp ....................................... 86
Figure 4-14. The dynamic mechanical response of water-DMA specimens; First heat and first cool scans
are shown. Average Tan δ, storage and loss modulus of first heat and cool scans, error bars representing
±1 standard deviation (n=3). ....................................................................................................................... 87
vii
LIST OF TABLES
Table 2-1. Recent formaldehyde emission standards for wood-based composites in Europe, Japan, and
the U.S.A. .................................................................................................................................................... 18
Table 3-1. Formaldehyde sorption efficiency using PAA beads in formaldehyde solution (Primary
NH2/CH2O=200; n=3). ................................................................................................................................ 49
Table 3-2. Formaldehyde sorption efficiency using PAA beads in formaldehyde solution with NH2/CH2O
= 10, 100, and 1000; n=3............................................................................................................................. 49
Table 3-3. Sequential and cumulative formaldehyde desorption efficiency over 4 sequential acid-washes;
n=3. ............................................................................................................................................................. 51
Table 3-4. Measured quantities of formaldehyde water-extracted from Virginia pine specimens, never-
heated and heated at 200°C, 60 min. n = 3. ............................................................................................... 54
Table 3-5. Sorption efficiency of the PAA beads in the presence of wood specimens, never-heated and
200°C heat treated ...................................................................................................................................... 54
Table 3-6. Measured quantities of formaldehyde extracted using PAA beads from Virginia pine
specimens, never-heated and heated at 200°C, 60 min. n = 3. .................................................................. 57
Table 4-1. HMR 50% conc. formulation ...................................................................................................... 66
Table 4-2. Glass transition temperatures (Tg: temp at tan δ max), and ΔTgs of DMF-plasticized control
and HMR-treated specimens, for poplar and pine species. Average values are shown (n=3), and numbers
in parenthesis are standard deviations....................................................................................................... 71
Table 4-3. Glass transition temperatures (Tg: temp at tan δ max), and ΔTgs of NMP-plasticized control
and HMR-treated specimens, for poplar and pine species. Average values are shown (n=3), and numbers
in parenthesis are standard deviations....................................................................................................... 72
Table 4-4. The volume change (%) and weight gain (%) of poplar and pine after HMR treatment (25⁰C,
70⁰C). Average values are shown (n=4), and numbers in parenthesis are standard deviations. ............... 77
Table 4-5. The glass transition temperatures of water-DMA specimens (Tg: temp at tan δ max). Average
(n=3) values with standard deviation in parenthesis. ................................................................................. 87
1
1. Introduction
In 2004, formaldehyde was classified as “carcinogenic to humans” by the international
Agency for Research on Cancer. In July 2010, new regulations were established for decreasing
the allowable formaldehyde emissions from nonstructural wood composites. Several studies have
been conducted on measuring and lowering the formaldehyde emissions of wood composites
adhesive binders, but native-wood formaldehyde emission has received little attention and it is
not completely understood. Wood formaldehyde emissions vary broadly and there is confliction
among different reports regarding this concept; but it is revealed that native emissions are
dependent on wood species, tissue maturity, wood components, wood moisture content and
thermal treatments (during drying and hot pressing) employed in the wood industry (Birkeland et
al., 2010; Bohm et al., 2012; Boruszewski et al., 2011; Meyer and Boehme, 1997; Roffael et al.,
2012; Schäfer and Roffael, 2000; Weigl et al., 2009; Young, 2004).
Biogenic and synthetic formaldehyde exists in a complex equilibrium within the woody matrix,
as crudely depicted in Figure 1-1.
Figure 1-1. Complex equilibria of the biogenic and synthetic formaldehyde exists in woody
matrix
Formaldehyde can be sorbed by wood by diffusing into the woody matrix and reacting
(physically/chemically) with the existed hydroxyl groups and water. This formaldehyde sorbing
ability of wood could make it difficult to extract all the available formaldehyde within the woody
matrix using one simple water-wash. A complete formaldehyde extraction can be achieved by
conducting several water washes, which is a time consuming process, or by using a more
efficient formaldehyde sorption media to remove all the available wood formaldehyde in one
step. Poly(allylamine) (PAA) beads, due to having primary amines, are strong formaldehyde
scavengers. Primary amines are stronger nucleophiles than the existed hydroxyl groups and
water in woody matrix for capturing formaldehyde. It is anticipated that by conducting one step
2
water-washing of wood specimens in the presence of PAA beads one can extract all the available
wood-sorbed formaldehyde.
In this study, a novel method was developed to capture native wood formaldehyde. In this
method small wood flake was placed in a sealed container (serum bottle); the container including
wood flake underwent heat treatment (25⁰C or 200⁰C) followed by water extraction and the
formaldehyde of the solution was quantified using the fluorimteric acetylacetone determination.
The PAA beads were also used for capturing wood-borne formaldehyde; in such a way that
following the above mentioned method, after the heat treatment and water extraction, the PAA
beads were introduced to the water/flake mixture to let the beads sorb all the available woody
matrix free-formaldehyde. The wood specimen was then removed and the PAA beads were
hydrolyzed using sequential acid-washes. The mass of formaldehyde taken from each acid-wash
sequence was then measured using the fluorimteric acetylacetone determination. Finally the
amount of wood borne formaldehyde obtained from both water extraction and water extraction in
the presence of the PAA beads were compared.
Before using the PAA beads with wood specimen, the formaldehyde sorption and desorption
efficiency of the beads were tested using standardized formaldehyde solution.
2. Literature review
2.1. Wood-based composites formaldehyde emission
2.1.1. Structural wood-based composites
Structural wood-based composites are referring to the wood-based composites, such as
structural particleboard, oriented strand board (OSB), plywood, and other materials commonly
used for exterior structural applications. Since these products are mainly used for outside
structural purposes, they should have superior water resistance property; which is provided by
the type of adhesive used in manufacturing these composites. Phenol-formaldehyde (PF), and
polyphenol polyisocyanate or polymeric MDI (pMDI) are the main type of adhesives employed
in structural wood-based composites industry.
3
Thermoset PF adhesives are synthesized by reacting phenol with formaldehyde under either
acidic or alkaline conditions. Novolak is the name of PF adhesive which is synthesized using
acid catalysis with a deficiency of formaldehyde, while Resol is the PF adhesive synthesized
under alkaline condition with excess of formaldehyde. Resol type PF is commercially used as
wood binder in structural wood based composites. In the manufacture of structural plywood,
Resol is applied to the wood veneers by roller or extrusion coating. The coated veneers are then
cross-grained, stacked, and cured in a press for 5-10 min at 120-130°C. For manufacturing
structural particleboard, Resol is sprayed onto the wood chips, or sprayed/spread by continuous
blenders. The wood chips covered with resin are formed into a mat and then pressed for 5-12
s/mm, based on the mat thickness, press temperature and moisture content at 190-230°C (Pizzi
and Mittal, 2003).
The reaction between phenol and formaldehyde under alkaline condition is depicted in Figure
2-1. The reaction begins by phenoxide formation from the phenol’s hydroxyl group after being
deprotonated by the base catalyst. The negative charge on phenoxide is accommodated by
resonance forms with the charge on ortho and para positions, which causes phenoxide to be tri-
functional. The initial reaction between phenoxide and formaldehyde can take place at the 2-,4-
(ortho), or 6- (para) positions of the benzene ring (only the reaction for 2- (ortho) position is
shown in Figure 2-1) which results in methylol formation. Methylol then turns into quinone
methide or carbonium ion which reacts with phenols or other methylols to form a linear polymer
with methylene and/or methylene ether linkages (Figure 2-2); linear polymers forms highly
branched structures under curing condition (Figure 2-3). Curing resin at elevated temperature
(over 100°C) results in condensation reaction between methylols and formation of methylene
ether linkages. High temperature curing also results in converting methylene ether linkages to
methylene linkages with the elimination of formaldehyde (Figure 2-4) (Pizzi and Mittal, 2003).
4
OH
OH-
O- O
H
O
H
O
H
O
HO
HH
O
H O-
H
H
O-
OH
H
H
O
CH2
Quinone methide
OH-
+
Phenoxide
Methylol
OH
CH2
+
and/or
Benzylcarbonium
Formaldehyde
Figure 2-1. The reaction between phenol and formaldehyde under alkaline condition results in
methylol which converts to quinone methide and/or Benzylcarboinium. (Benzylcarbonium
formation is more favorable than quinone methide)
5
O-
O
CH2
OOH
H
Tautomerizes
OHOH
O
CH2
OH
OH
H
HOH
O
OH
OH
CH2
+
O
H
OH
OH
H
H
OHOH
OH
O
OH
Quinone methide
Quinone methide
Benzylcarbonium
Figure 2-2. The reactions of quinone methide and benzylcarbonium with phenoxide and
methylol; results in linear polymer with methylene and methylene ether linkages.
6
O
OH
OH
OH
OH OH
OH
OH
OH
OHOH
OH
Figure 2-3. Highly branched Resol after curing at elevated temperature.
OH
O
OH OHOH
Curing +
O
Formaldehyde
Figure 2-4. Curing Resol at elevated temperature results in converting methylene ether linkages
into methylene linkages and formaldehyde elimination.
In curing process (130-200c) of alkaline resole, the existence of residual methylols and
methylene linkages was proved (Figure 2-5, Top left) (Knop and Scheib., 1979). At temperatures
above 130c significant amounts of methylene ether bridges between phenolic rings are formed.
The existence of hemiacetals in formaldehyde-phenol solutions can be also possible (Figure 2-5,
bottom) (Kopf and Wagner, 1973); however hamiacetals cannot withstand typical alkaline cure
conditions and may re-form after curing (Myers, 1986b).
7
OH OH OH
OH
OH
O
OH
OH
O OH O OH
Methylene l inkage Methylol
Phenol hemiacetal Benzyl alcohol hemiacetal
Methylene ether linkage
Figure 2-5. The possible structures result from phenol-formaldehyde reactions (Myers, 1986b).
In 1984, Maciel and coworkers studied the PF cured resin using C-13 NMR and they
confirmed the availibility of below structures (Figure 2-6) as well (Maciel et al., 1984).
O
Ar
OH OH OH OH O
H
Figure 2-6. The confirmed structures existed in cured PF resin (Maciel et al., 1984)
None of the above structures (Figure 2-6), except the hemiacetal one, is expected to be even
moderately unstable. This shows the high stability of cured PF (Emery, 1985; Myers, 1983;
Roffael, 1978) which results in very low formaldehyde emission from boards made of PF
However, it is stated that 3.6 percent of the cured PF weight, was released as formaldehyde after
soaking the sample for 5 hours in boiling water (Baurichtlinie, 1980). Rosenberg also found that
by placing cured PF in 1N sulfuric acid at 78⁰C, a fast initial formaldehyde release took place
(Rosenberg, 1978). The other study (Myers, 1986b) also indicated the formaldehyde release of
PF at high relative humidity conditions.
The amount of free formaldehyde released from plywood panels bonded with PF resin was
determined using both gas analysis (EN 717-2) and small-chamber (EN 717-1) methods. Results
were compared with the free formaldehyde emissions of the particleboards and MDFs bonded
8
with UF (Urea-formaldehyde) and MUF(Melamine-Urea formaldehyde) resins (Salem et al.,
2011b). Results showed that the free formaldehyde emission of plywood panels was very lower
than that of particleboard or MDF and it was 0.005-0.007 mg/m3 using small chamber method.
These values are very close to the wood-borne formaldehyde emission (0.008-0.01 ppm for
spruce wood flakes). Same low formaldehyde emission from PF bonded wood-based composites
was also reported in other studies (Bohm et al., 2012; Salem et al., 2012). This is due to the
stable C-C bonding in the cured PF resin against hydrolytic attack. This high stability is also the
reason that PF resins are considered as waterproof while UF is not. At such low levels of emitted
formaldehyde the PF bonded boards are considered to be formaldehyde free (Salem et al., 2011b;
Salem et al., 2012).
In another study (Myers and Nagaoka, 1981) static method (Paper sorption test, and the
Japanese desiccator test) and dynamic method (air-flowing chamber test) were used to measure
the formaldehyde emission from PF and UF bonded boards. Results showed that the
formaldehyde emission from PF bonded boards were around 20 folds less than that of UF
bonded boards.
The eluted formaldehyde from PF-bonded boards at different relative humidity (0-20%) was
measured (for 40 days) using gas elution method and the results were compared to that of UF-
bonded boards (Myers, 1986a). It is found that the formaldehyde removal of the PF-bonded
board was approximately 10-fold less than that for the UF-bonded board. However the effect of
increasing relative humidity was the same on both UF and PF boards which the eluted
formaldehyde increased by increasing relative humidity.
Jiang and coworkers studied the volatile organic compounds (VOC) emission from different
types of particleboards. Results showed that although the formaldehyde emission of PF bonded
particleboard was not significant but methanol emission was high (75% of the total VOC
released from PF particleboard) (Jiang et al., 2002).
The other type of adhesive which is used for manufacturing structural wood-based
composites (specifically OSB) is pMDI. OSB is produced with distinct layers of wood strands
and often pMDI is used for binding core layers while PF is used for face layers. The synthesis of
pMDI starts with the reaction between aniline and formaldehyde under acidic condition; reaction
9
takes place at ortho and para positions to the NH2 group which leads to the formation of 4,4’-,
2,4’-, and 2,2’- methylenedianiline compounds. These diamines react to form more methylene
bridged polyphenylene polyamines. The acidic polyamine mixture is then neutralized and reacted
with phosgene in high boiling aromatic solvents such as chlorobenzene. Thereafter the solvent is
removed and the product is isolated. The final product is a mixture of monomeric and oligomeric
poly isocyantes which is known as pMDI in the wood industry (Figure 2-7) (Pizzi and Mittal,
2003).
O
H H
+
NH2
H+
NH2NH2
NH2 NH2
NH2
NH2
4, 4'-methylenedianil ine2, 2'-methylenedianil ine
2- 4'-methylenedianil ine
NH2 CH2 NH2
CH2 NH2
n
Mixture of methylene bridged polyamines
n= 1-12
Aniline
Phosgenation
NCOOCN
4, 4'-methyleneisocyanate
NCO
NCO
2- 4'-methyleneisocyanate
OCN CH2 NCO
CH2 NCO
n
Mixture of methylene bridged polyisocyanate
n= 1-12
Figure 2-7. Synthesis reaction of pMDI (Pizzi and Mittal, 2003)
10
The curing of pMDI is followed by the pMDI reaction with wood components. The general
reaction of pMDI with hydroxyl groups (within cellulose, lignin and hemicellulose) and water in
woody matrix is depicted in Figure 2-8; the reaction between Isocyanate and hydroxyl groups
results in urethane formation which can react with another isocyanate to form allophanate. The
reaction between isocyanate and water results in urea linkage by elimination of CO2; urea
compound can react with another isocyanate to form Biuret. The formation of urethane is
mentioned in some studies (Weaver, 1995; Zhou and Frazier, 2001), but since both the urathane
and allophanate are not thermally stable they can be degraded at high temperatures (above
165⁰C) (Zhou and Frazier, 2001). On the other hand urea compounds are very stable and their
formation is essentially irreversible. pMDI resin can flow into the micrometer and even angstrom
voids of wood and penetrate into the amorphous components of the wood cell wall and the
subsequent curing can result in a type of interpenetrating polymer network (IPN). This restricts
wood polymer motions (Marcinko et al., 1998; Marcinko et al., 1999) and lead to more
dimensional stability and water resistance for OSB panels.
R-NCO + HO-R' R NH C
O
O R'
R-NCO
R NH C
NR C
O
O R'
O
R-NCO+ H-OH R NH C
O
OHR-NCO
-CO2
R NH C
O
NH R
R-NCO
R NH C
NR C
O
NH R'
O
Urathane Allophanate
Carbamic Acid Urea Biuret
Figure 2-8. pMDI reactions with woody matrix hydroxyl groups and water which results in
urethane and urea respectively (Pizzi and Mittal, 2003)
Although the pMDI synthesis starts with the reaction between aniline and formaldehyde
under acidic condition, based on the reactions shown above, there is no formaldehyde release
from pMDI adhesive (Pizzi and Mittal, 2003; Roffael, 2006; Wang et al., 2004).
pMDI was also combined with UF, MUF (Tinkelenberg et al., 1982), and PF (Wang et al.,
2007) to decrease the overall formaldehyde emission and increase the strength of the final
product. By adding approximately 10% pMDI to UF adhesive, the formaldehyde emission
became lower than standard formaldehyde emission limit and also the mechanical properties of
11
resulted boards had a significant improvement (Tinkelenberg et al., 1982). Mixing pMDI with PF
also increased both mechanical and physical properties and decreased the formaldehyde emission
(Wang et al., 2007).
Wang and coworkers also mixed pMDI with UF in order to decrease the formaldehyde
emission of the final particleboard (Wang et al., 2004). It was stated that most of UF bonded
particleboards release formaldehyde between 35 and 60 mg/100gboard which by adding pMDI to
the UF, the emissions became lower than 9 mg/100gboard. They also found that adding the
acidic agent NH4Cl to UF-pMDI mixture did not improve the curing. By omitting the acidic
agent and finding the optimum pressing time and temperature (175C and 4.5 min) they produced
boards with formaldehyde emission below the E1 standard ( 9mgCH2O /100gboard).
Jiang et al. studied the volatile organic compounds (VOC) emission from different types of
particleboards. Results showed that although the formaldehyde emission of pMDI bonded
particleboard was very low (lower than UF and PF bonded boards) significant emission of acetic
acid was found (Jiang et al., 2002).
Using pMDI is also very common in manufacturing boards from agricultural residues such
as, wheat straw and rice straw. Grigoriou tested various UF-pMDI ratios in bonding wheat straw
composites and he found that adding pMDI to UF can significantly improve dry and wet strength
and swelling properties of the final particleboard (Grigoriou, 2000).
2.1.2. Non-structural wood-based composites
Non-structural wood-based composites are referring to the wood-based composites, such as
particleboard, medium density fiberboard (MDF) and other materials commonly used for interior
applications.
Urea-formaldehyde resin (UF) is the most abundant thermoset amino resin used in producing
non-structural wood based composites. The advantages of using UF are as follows: initial
solubility in water, hardness, nonflammablity, low price, good thermal properties, and
transparent color after curing (Pizzi and Mittal, 2003). UF is synthesized by the reaction between
urea and formaldehyde; this reaction first results in monomethylol urea which can also react with
12
formaldehyde to form dimethylol urea. The resulted methylols can further react with each other
or also urea to produce methylene ether and methylene linkages, respectively (Figure 2-9). The
reactions continue and more branched structures are produced, for example a cross-linked
structure can be formed by the reaction between methylene diurea and a methylol urea. The
polymeric and branched structures can be formed at elevated temperature (above 120⁰C) and
under acidic condition (pH below 5). Under curing condition, the structures shown in Figure 2-
10 are presumed to be available which are subjected to hydrolysis. Most, if not all, the reactions
forming these structures are reversible, which release water in the forward direction and
formaldehyde in the reverse direction (Figure 2-11) (Dunky and Lederer, 1982; Ebdon et al.,
1984; Kopf and Wagner, 1973; Kumlin and Simonson, 1978; Ludlam and King, 1984; Myers,
1986b; Nair and Francis, 1983; Slonim et al., 1978).
O
NH2NH2
+ OHOH
H
H
O
NH2 NH OH + H2O
Monomethylol Urea
O
NH2 NH OH+ OHOH
H
H
O
NH NH OHOH
Dimethylol Urea
+ H2O
O
NH2 NH OH
+O
NH2NH2
O
NH2 NH
O
NH NH2
Methylenediurea
+
O
NH2 NH OH+
O
NH2 NH OH
O
NH NH2
O
NH2 NH O+ H2O
Dimethyleneurea ether
O
NH NH
O
NH NH+
O
NH NH OH
O
NH N
O
NH NH
NH
O NH
H2O
+ H2O
Figure 2-9. Some of the possible urea-formaldehyde reactions
13
NC
O
NR
CH2
NR
C
O
N
NR
C
O
NNC
O
NR O
NC
O
NR OH
NC
O
NR O OH
n
Methylene
Methylene ether
Hemiacetal
Acetal
Figure 2-10. The structures available in cured UF resin which are subjected to hydrolysis
O
NH NH2
O
NH2 NH O
O
NH2 NH
O
NH NH2
Methylenediurea
CH2O+
Figure 2-11. The hydrolysis of methylene ether bond which results in formaldehyde generation
In an acid-catalyzed UF-bonded board, formaldehyde can be in form of dissolved methylene
glycol monomer and oligomers, paraformaldehyde (the reaction of formaldehyde with itself),
Hexamethylenetetramine (the formaldehyde reaction with UF curing catalyst (ammonia)),
chemically bonded to UF resin, chemically bonded to wood (amidomethylene ethers with
cellulose), cellulose hemiacetals and acetals. Each of the mentioned form of formaldehyde is a
potential source of formaldehyde emission. The weakly held formaldehyde forms were stated as
methylene glycol, cellulose hemiacetals, amidomethylols, and cellulose amidomethylene ethers
(Myers, 1986a, b).
To prove the reaction between formaldehyde and wood, the initial formaldehyde released
from neat UF resins, which were cured at similar conditions to particleboard press, was measured
(Myers, 1986b). It was found that the neat UF formaldehyde emission was significantly higher
than that for UF bonded particleboard. This indicates that the woody matrix strongly mitigates
the influence of resin hydrolysis on board emission (Myers, 1986b). Since formaldehyde can
react with cellulose hydroxyls (preferably secondary hydroxyls) to form hemiacetals and acetals
14
(under elevated temperature and acidic conditions) (Meyer et al., 1976; Walker, 1975) it is
possible that formaldehyde could be bound to the woody matrix during the hot pressing of UF
bonded particleboard (Myers, 1986b). Lignin methylolation can also be expected during the
pressing (Rosenberg, 1978).
The acid-catalyzed hydrolysis of methylene-diurea was also studied (Braun et al., 1983). It is
stated that the initial hydrolysis products are urea and monomethylol urea, while the latter one
breaks into urea and formaldehyde. They found that 20 percent degradation to formaldehyde
happened at 80⁰C in pH 5 after 3 hours; while around 40 percent degradation to formaldehyde
was resulted at pH 3 in only 30 minutes. They found the degradation rate of 0.5 percent per hour
under pH 3 and 25⁰C. They concluded that even the methylene linkages, which presumed to be
stable in cured resin, degrades significantly when the cured resin dissolved in acidic solution at
25⁰C (Braun et al., 1983).
In another study, formaldehyde removal using gas elution method was conducted to detect
the formaldehyde emission of wood-based composite boards. Variables in this study were time,
gas flow, sample comminution, gas type and humidity. Results showed that formaldehyde
emission from shredded UF bonded board was only slightly faster than from the larger pieces.
This similarity can show that formaldehyde release is not macro-diffusion takes place inside the
voids but it’s either due to micro-diffusion inside the wood or it’s because of bond rupture. The
elution process was slow and did not have any end point (plateau line) after 10 days
measurement. Regarding the effect of different eluent gases (N2, CO, and CO2) on formaldehyde
removal, three gases provided no differentiation between formaldehyde states in UF board. To
see the effect of N2 relative humidity (RH) on formaldehyde removal, formaldehyde emission
was measured using dry, 20% RH, and 80% RH N2; results showed that by increasing RH% the
eluted formaldehyde increased (Myers, 1986a).
Using weighing bottle method, the grounded sieved UF bonded board eluted formaldehyde
was measured. The results were compared to formaldehyde-sorbed pined wood and PF bonded
board. It was found that at high relative humidity (80%) there was no end point for eluted
formaldehyde from UF board after 30 days while for formaldehyde-sorbed wood, it became
constant after only about 5 days. At day 20 the eluted formaldehyde from UF board exceeded the
formaldehyde-sorbed wood formaldehyde. The fast formaldehyde removal from formaldehyde-
15
sorbed wood samples can be due to the very weakly bonded formaldehyde to wood (perhaps
hemiacetal or methylene glycol). For UF board in contrast, it contains little amount of loosely
bound formaldehyde and great amounts of more strongly bound formaldehyde which needs more
time to be eluted. Results also showed that the formaldehyde emission of UF bonded board was
an order of magnitude higher than PF bonded board (Myers, 1986a).
The formaldehyde emission during time under free access of air condition was also
investigated in another study (Wiglusz et al., 1990). They stated that formaldehyde emission
decreased by more than 50% after 3 month and after that the emission decrease was slight,
however it continued even after 1-3 years.
The above formaldehyde emission data (Myers, 1986a) is related to the traditional UF
formulation. Frihart and coworkers measured the formaldehyde emission of UF bonded
particleboards composed of new lower formaldehyde emitting UF and also tested the effect of
temperature and relative humidity on the formaldehyde emission (Frihart et al., 2012). Results
showed that by increasing both relative humidity and temperature the formaldehyde emission is
also increased. Same results were also reported in older studies (Myers, 1985; Myers and
Nagaoka, 1981).
Formaldehyde measurement of grounded UF bonded board using water extraction method
was also investigated (Myers, 1986a). Sample was placed in a stoppered flask and water (pH 3)
was added. Results showed that the formaldehyde extraction continued rapidly after 6 days. By
comparing the results of this method to that of the perforator method (will be discussed later), the
measured formaldehyde using water extraction after 30 days was significantly (~30 times)
higher.
The formaldehyde emission from different wood-based composites was measured using
small chamber (EN 717-1) and gas analysis (EN 717-2) (Salem et al., 2011b) .Three types of
particleboard (uncoated, laminated and veneered) bonded with UF resin and two types of MDF
(uncoated and laminated) bonded with melamine-urea formaldehyde (MUF) resin were tested for
formaldehyde emission. The effect of board thickness on formaldehyde emission was also
investigated. The results were compared to the formaldehyde emission from PF bonded
plywoods. It was found that veneered particleboards had the highest formaldehyde emission
16
among all the boards; which was also found in another study (Salem et al., 2012). However other
studies (Nemli and Colakoglu, 2005; Park et al., 2011; Salem et al., 2011a) reported that using
surface coatings helps to significantly decrease the formaldehyde emission of particleboard
panels.
Regarding the laminated samples, the formaldehyde emission was lower than veneered
samples for both MDF and particleboard. The formaldehyde emission from PF bonded boards
was significantly lower than UF and MUF bonded boards. Park and coworkers reported the same
results (Park et al., 2011). The reason for high formaldehyde emission of UF bonded is its
susceptibility to hydrolysis under normal conditions. It is stated that MUF due to having C-N
bonding in its structure is more stable than UF which is why the formaldehyde emission of MUF
bonded boards are lower. It was also found that by increasing board thickness the formaldehyde
emission also increased for almost all the boards (Salem et al., 2011b), (Salem et al., 2012).
Wang and Gardner developed a method to measure the formaldehyde emission during
particleboard pressing. They stated that the formaldehyde emission increased by increasing press
temperature and time, mat resin content and moisture, and board density (Wang and Gardner,
1999).
The other adhesive that is used in producing semi-exterior and exterior wood-based
composites is melamine formaldehyde (MF). Since MF is expensive it is mixed with urea to
produce melamine-urea formaldehyde (MUF) (Pizzi and Mittal, 2003). MF has much higher
resistance to water than UF and less formaldehyde emission (Kim and Kim, 2005). MF is also
used for impregnating paper sheets for making self-adhesive overlays for the surface of wood-
based composites. The reaction between melamine and formaldehyde is similar to that of UF
resin; however it happens easily and completely. Up to six formaldehyde molecules can be
bonded to one molecule of melamine and unlike UF a complete methylolation can occur.
Another difference between MF and UF is that MF can undergo condensation reaction under
acidic, neutral and even slightly basic conditions. In curing MF at temperature higher than
100⁰C, no formaldehyde is released; and when the temperature is close to 150⁰C, only small
amount of formaldehyde is released. However it is well known that under the same conditions
for curing UF, there is a lot of formaldehyde emission (Pizzi and Mittal, 2003).
17
There are few publications on measuring the formaldehyde emission from MF bonded
boards. In one study the formaldehyde emission of MF and UF bonded MDF boards was
measured (Kim and Kim, 2005). They found that the formaldehyde emission of MF bonded
MDF was around 12 times lower than that of UF bonded MDF using desiccator method (will be
discussed later) and around 4 when perforator method was used.
2.2. Formaldehyde regulations
After the discovery of formaldehyde by a Russian scientist (Alexander Michailowitsch
Butlerow) in 1855, the technical synthesis of formaldehyde from dehydration of methanol was
achieved by a German chemist (Wilhelm von Hofmann). Between 1900 and 1930 formaldehyde
based adhesives were used to produce wood-based composites, and after 1950 these products
became an alternative to solid wood for making furniture and constructional purposes. The first
health issues (mainly eyes irritation) were reported in the middle of 1960. In 1977 a guideline
value of 0.1 ppm was proposed by former German Federal Agency of Health. Criteria for the
regulation of formaldehyde emissions from wood-based composites were determined in 1981 in
Germany and Denmark and in 1985 in the United States. Due to the continuously increase of
formaldehyde concentrations in air, studies focused more on measuring indoor formaldehyde
emissions in 1990s; and an increase in formaldehyde emission from furniture and mobile homes
was reported. In 2004 formaldehyde was classified as carcinogenic for humans; and as a
consequence, new guideline values and regulations were established (Salthammer et al., 2010).
International Agency for Research on Cancer (IARC) established that formaldehyde is
undetectable by smell at concentrations below 0.1 ppm, between 0.1 and 0.5 ppm it is detectable
by smell and causes some slight eyes irritation, between levels 0.5 and 0.1 ppm it results in eyes,
nose and throat irritation and over 0.1 ppm, it produces extreme discomfort (IARC, 2004).
The formaldehyde emission guidelines are different in different countries. In Europe the
standard regulation for formaldehyde emission is known as EN 13986 and based on the amount
of formaldehyde emission from the wood-based composites, they are classified as E1 and E2
classes (E1 is more common). The most common test methods used in European standards are
Perforator method (EN 120), European small chamber method (EN 717-1), and gas analysis
method (EN 717-2) (Table 1). In Japanese standard, the board are classified with stars (2-4
starts), which 4 stars representing the lowest amount of formaldehyde emission. The desiccator
18
test method is the most common method in Japanese standard. It is also important to be
mentioned that Japanese standard is more restricted and has less formaldehyde emission values
than European standard (Table 1) (Salem and Bohm, 2013). In the United States the new
regulation is established and started from January 1st 2013. The new standard, which is extended
to all 50 states, is the most rigorous standard in the world (U.S.A congress, 2010). The test
method used for the United States standard is commonly large chamber (Table 2-1).
Table 2-1. Recent formaldehyde emission standards for wood-based composites in Europe,
Japan, and the U.S.A.
Country Standard Test method Board class Limit value
Europe EN 13986
EN 717-1 E1
PB, MDF, OSB
≤ 0.1 ppm
EN 120 ≤ 8mg/100g o.d. board
EN 717-1 E1
PLW
≤ 0.1 ppm
EN 717-2 ≤ 3.5 mg/(h.m2)
EN 717-1 E2
PB, MDF, OSB
> 0.1 ppm
EN 120 > 8 ≤ 30 mg/100g o.d.
board
EN 717-2 E2
PLW
> 0.1 ppm
EN 717-1 >3.5 ≤ 8 mg/(h.m2)
Japan JIS A 5908 &
5905
JIS A 1460
(Desiccator)
F**
≤ 1.5 mg/L
F***
≤ 0.5 mg/L
F****
≤ 0.3 mg/L
USA ANSI A 208 ASTM E1333
(Large chamber)
PB ≤0.09 ppm
MDF ≤ 0.08 ppm
PB: Particleboard, MDF: medium density fiberboard, OSB: oriented strand board
19
2.2.1. Test methods for formaldehyde emission
The most common methods used for detecting formaldehyde emission from wood/wood-based
composites samples are as follows:
2.2.1.1. Chamber methods
There are different standards such as, EN 717-1 (European Committee for Standardization,
2004), JIS A 1901 (Japanese Standards Organization, 2003), JIS A 1911 (Japanese Standards
Organization, 2006), ASTM E 1333 (American Society for Testing and Materials, 1996), and
ASTM D 6007 (American Society for Testing and Materials, 2002) which are using chamber
method for detecting formaldehyde emission. In all these standards, samples of known surface
area are placed inside a chamber (each standard has different chamber size) and the
formaldehyde emission of the samples is tested with regard to temperature, relative humidity, air
exchange, air velocity, and sample loading factor (ratio of the surface of the sample to the
volume of the chamber) under standardized conditions. The formaldehyde emitted from the
samples mixes with the air in the chamber and the air is sampled periodically by drawing the
chamber’s air through gas washing bottles containing water. Formaldehyde dissolves in water
and the formaldehyde concentration of water is then determined using acetylacetone method
(will be discussed later).
Chamber method is capable of measuring formaldehyde concentration in air and also
provides sample specific emission rate (SER). SER can be based on sample length, volume, area,
and unit and the units are μg.m-1
.h-1
, μg.m-3
.h-1
, μg.m-2
.h-1
, and μg.unit.h-1
, respectively. In most
cases, the chamber method results are reported as area emission rate or chamber concentration in
the steady state (when formaldehyde concentration in the chamber remains constant).
2.2.1.2. Gas analysis method
This method is specifically related to the EN 717-2 standard (European Committee for
Standardization, 1994). In this method the samples are placed in a closed chamber (pre-heated at
60⁰C) and the formaldehyde emitted from the samples mixes with the air chamber and the air is
sampled by continually drawing through gas washing bottles containing water. Formaldehyde
dissolves in water and the formaldehyde concentration of water is then determined using
20
acetylacetone method. The formaldehyde emitted from the samples is calculated based on the
formaldehyde mass (mg), the sample’s exposed area (m2) and the sampling time (h). The
samples emitted formaldehyde is then reported in mg.m-2
.h-1
.
2.2.1.3. Perforator method
In this method the formaldehyde content of the sample is determined using perforator device.
This method is specifically related to EN 120 standard (European Committee for
Standardization, 1992). Small Samples with known mass and moisture content are placed in a
round bottom flask contains toluene. The flask is connected to the perforator device which
contains distilled water. The condenser and the gas absorption equipment are then connected.
Gas absorption equipment contains known volume of distilled water. Samples are extracted by
toluene at high temperature (110⁰C, 3hr). After everything cools down, the perforator device is
washed with known volume of water and the toluene in rinsing water is discarded. The water in
gas absorption equipment is mixed with the rinsing water. The formaldehyde of water is then
determined using acetylacetone method. The samples emitted formaldehyde is reported in
mgCH2O/100 gsample.
2.2.1.4. Flask method
This method has been developed by Roffael in 1975; thereafter it was modified in different
ways and finally published as EN 717-3 standard (Standardization, 1996). One to three samples
are suspended over water in a closed polyethylene bottle (40⁰C, 3hr). The emitted formaldehyde
from samples is absorbed by water. The formaldehyde concentration of water is then determined
using acetylacetone method. The samples emitted formaldehyde is reported in mgCH2O/1000 g
sample.
2.2.1.5. Desiccator method
There are different standards such as: ASTM D 5582 (American Society for Testing and
Materials, 2006), JIS A 1460 (Japanese Standards Organization, 2001), and JAS (Ministry of
Agriculture and Forestry, 2003) based on this method. The desiccator method is very similar to
flask method; samples with known surface area placed over water for 24 hours at a constant
temperature in a desiccator. Formaldehyde dissolves in water and the water is analyzed for
21
formaldehyde concentration using acetylacetone (JIS A 1460) or chromotropic acid (ASTM D
5582) method. The samples emitted formaldehyde is reported in mg/L.
Different studies were reported the relationship between the mentioned methods by
conducting linear correlation analyses. It was found that the variations between the results from
different methods can be due to the differences in test conditions. Factors such as edge sealing,
sample conditioning, temperature, sample type, sample thickness, and the resin that is used for
making the sample all have significant effect on the final results (Risholm-Sundman et al., 2007;
Salem et al., 2011b).
There are also other methods for measuring formaldehyde emission. For In-situ
formaldehyde detection which provides real-time measurements for outdoor environment,
Spectroscopic techniques are used such as: differential optical absorption spectroscopy (DOAS),
Fourier transform infrared absorption (FTIR), laser induced fluorescence spectroscopy (LIFS),
and tunable diode laser spectroscopy (TDLS). However these methods have low sensitivity for
formaldehyde detection and also they required expensive and elaborate instrumentation
(Vairavamurthy et al., 1992). For fast and simple indoor formaldehyde measurements, sensor
technology is developed. Microgas sensor (Lv et al., 2008), cataluminescence-based gas sensor
(Zhou et al., 2006), MWCNTs semiconductor gas sensor (Wang et al., 2008), sensors based on
enzyme reactions (Achmann et al., 2008), conductometric biosensor (Vianello et al., 2007), and
amine-functionalized colorimetric sensor (Feng et al., 2010) are some of examples of sensors
developed for formaldehyde detection. Although most of the mentioned methods have good
sensitivity, they suffer from high detection limits (Salthammer et al., 2010).
2.2.2. Formaldehyde analysis
2.2.2.1. Quantitative analysis of formaldehyde
In order to determine the formaldehyde concentration of water-formaldehyde solution,
different determination methods can be used such as: Pararosaniline (Miksch et al., 1981),
MBTH (Altshuller and Leng, 1963), DNPH (De Andrade and Tanner, 1992), chromotropic acid
(Altshuller et al., 1961), and acetylacetone (acac) (Nash, 1953) methods. Among these methods
acetylacetone is a widely applied standard procedure and is highly specific for formaldehyde.
This method is used in this study and described below.
22
2.2.2.2. Acetylacetone formaldehyde determination method
Acetylacetone reagent consists of ammonium acetate (2M), acetic acid (0.05M) and
acetylacetone (0.02M) in water (1L); it is sensitive to light and should be stored in cool and dark
place (Bunkoed et al., 2010) and once the color is yellow it should be discarded. In order to
determine the formaldehyde concentration of the sample, same volumes of sample solution and
acetylacetone reagent are mixed together (10min, 60⁰C); based on Hantzsch reaction, the
cyclization of acetylacetone, ammonium acetate, and formaldehyde leads to the formation of
dihydropyridine 3,5- diacetyl-1,4-dihydrolutidine (DDL) (Figure 2-12). Quantification of DDL
can be performed by both UV/vis (412 nm) and fluorescence (410nm excitation, 510 nm
emission) spectroscopies (Belman, 1963; Nash, 1953). UV/vis spectroscopy is used when the
concentration of formaldehyde is over 0.4 μg/ml and when it’s below this number fluorescence
spectroscopy should be used (Belman, 1963; Myers and Nagaoka, 1981). In this study due to
having samples with low formaldehyde concentrations (below 0.4 μg/ml) fluorescence
spectroscopy is used.
Figure 2-12. Reaction between formaldehyde and acetylacetone reagent results in DDL
2.2.2.3. Calibration curve
When using the acetylacetone method, it is important to have a proper calibration curve.
Calibration curve is made by using standardized formaldehyde solution (will be described later).
Different solutions with different known formaldehyde concentrations are prepared and reacted
with acetylacetone as discussed above. The obtained fluorimetric intensity for each sample is
then plotted as a function of formaldehyde mass (Figure 3-5) and the calibration curve is
produced. The resulted calibration curve should be highly linear; it is reported that in case of
H H
O O O
NH3+
O
NH
O
+
DDL
2
23
using fluorescence spectroscopy, the calibration curve is linear up to formaldehyde concentration
of 0.4 μg/ml. One can use the calibration curve to determine the formaldehyde mass by
subtracting the blank fluorimetric intensity from the unknown formaldehyde solution
fluorimetric intensity and divided the number by the slope value of calibration curve.
2.2.2.4. Standard formaldehyde solution
In order to prepare standard formaldehyde solution, known volume of formaldehyde solution
(37% weight CH2O in water) is diluted in known volume of water. There are different
standardization methods such as: sodium sulfite, alkaline peroxide, iodimetric, ammonium
chloride, mercurimetric, potassium cyanide, and hydroxylamine hydrochloride, Methone or
Dimedon (Walker, 1975). Among these methods sodium sulfite due to its accuracy, simplicity
and rapidity is probably the best method for formaldehyde standardization (Walker, 1975). In
this method, sodium sulfite solution (1M, 20ml) is added to the formaldehyde solution (50 ml)
and after adding few drops of indicator (thymolphthalein) the color of solution changes to blue.
Each formaldehyde molecule reacts with one molecule of sodium sulfite to produce a sodium
hydroxide molecule (Figure 2-13). The solution is then titrated using HCl (0.1 M) till a clear-
color solution is reached. The consumed volume of HCl is recorded and by using the below
equation the accurate formaldehyde concentration can be calculated:
Where: V=consumed HCl (ml), N=precise normality of HCl.
CH2O + Na2SO3 + H2O NaOH + CH2(NaSO3)OH
Figure 2-13. Reaction between formaldehyde and sodium sulfite in water results in sodium
hydroxide
24
2.3. Formaldehyde emission from solid wood, a review
Formaldehyde is naturally forming in wood matrix (Meyer and Boehme, 1997; Que and Furuno,
2007). All the wood components including cellulose, hemicellulose, lignin and some extractives
have potential to release formaldehyde (Schäfer and Roffael, 2000). It is found that the wood-
borne formaldehyde emission varies and it depends on so many factors such as: wood species
(Bohm et al., 2012; Meyer and Boehme, 1997; Weigl et al., 2009; Young, 2004), tissue maturity
(Weigl et al., 2009), wood components (Schäfer and Roffael, 2000), temperature and process
treatments employed in the industry (Birkeland et al., 2010; Schäfer and Roffael, 2000; Young,
2004),moisture content (Meyer and Boehme, 1997) and particle size (Boruszewski et al., 2011;
Roffael et al., 2012).
2.3.1. Effect of wood species
The formaldehyde emission of different wood species such as: Beech, Douglas-fir, Oak,
Spruce, and Pine was measured (Meyer and Boehme, 1997). The highest and lowest
formaldehyde emitted values were for wet oak (0.6 mgCH2O/100gwood) and dry beech (0.16
mgCH2O/100gwood), respectively (Meyer and Boehme, 1997). Weigl and coworkers found
slightly higher formaldehyde emitted values than what Mayer and Boehme, 1997 reported, which
can be due to different sample size and shape (Weigl et al., 2009). Weigl et al., 2009 found that
softwood species had highest formaldehyde emission than hardwood species. Same result was
also reported in another study (Jiang et al., 2002). However Bohm and coworkers, 2012 did not
find a clear difference between softwood and hardwood formaldehyde emission values (Bohm et
al., 2012).
Young, 2004, measured the formaldehyde emission of different species harvested in one area in
New Zealand. He concluded that there was no difference in the formaldehyde emission of
different species within an area. He also measured the formaldehyde emission of Radiata Pine in
different areas and did not find any significant difference (Young, 2004).
2.3.2. Effect of wood maturity
Weigl and coworkers, 2009, found significant difference between formaldehyde emission
from juvenile wood and mature wood of different wood species. For beech, poplar, and pine
25
mature wood had higher emission than juvenile wood while for oak and spruce it was opposite
(Weigl et al., 2009).
2.3.3. Effect of wood components
Schäfer and Roffael, 2000, measured the formaldehyde emission of different wood
components. In order to measure the formaldehyde emission from cellulose, cotton linters with
about 99% alpha cellulose were subjected to thermal hydrolysis (3hr, 40⁰C, 100⁰C, and150⁰C).
They reported that increasing temperature had no significant influence on formaldehyde
emission of cellulose; the reported emission was around 5 (CH2Omg/100 g). They found that
formaldehyde emission from strach is even lower than cellulose. However for hemicellulose, the
formadlehyde emission at elevated temperature (150⁰C) was significantly higher than both
cellulose and starch. For measuring lignin formaldehyde emission, they used sodium
lignosulfonate; the results showed that at high temperature sodium lignosulfonate emitted high
quantities of formaldehyde (higher than both cellulose and hemicellulose). Regarding the
extractives, they found that some can act as formaldehyde scavengers and some can emit
formaldheyde (Schäfer and Roffael, 2000).
2.3.4. Effect of thermal treatment
The wood-bonre formaldehyde emission increases at elevated temperatures (Birkeland et al.,
2010; Schäfer and Roffael, 2000; Young, 2004). It is stated that formaldehyde emission of both
hemicellulose and lignin increase by increasing temperature while for cellulose no significant
difference was reported (Schäfer and Roffael, 2000). It is found that thermal degradation of
polysaccharides results in formaldehyde emission. Under acidic conditions the degradation of
hexoses results in oxymethyl-furfural, which then forms formaldehyde and furfural (Figure 2-14)
(Schäfer and Roffael, 2000). Another possible reaction which can result in formaldehyde
emission is the scission of carbon-oxygen bond which results in acetic acid and formaldehyde
formation (Figure 2-15) (Browne, 1958).
26
OOH
OHOH
OH-2H2O
O
OH
O
H
OxymethylfurfuralXylose
OO
H + CH2O
Furfural Formaldehyde
Figure 2-14. Xylose forms oxymethylfufural and further furfural and formaldehyde at elevated
temperature
OOH
OHOH
OH
CH2O
CH3COOH
CH3COOH
Acetic acidXyloseFormaldehyde
Figure 2-15. Thermal degradation of xylose results in acetic acid and formaldehyde
The formaldehyde generation of lignin under high tempetures is thought to occur under
acidolytic reaction. As shown in Figure 3-6 under acidic condition gamma-carbon is released as
formaldehyde (Brosse et al., 2010; Lundquist, 1976; Lundquist and Ericsson, 1970; Rousset et
al., 2009; Santos et al., 2013).
2.3.5. Effect of moisture content and particle size
It is reported that by increasing moisture content the wood-borne formaldehyde emission also
increases (Meyer and Boehme, 1997). It is stated that by decreasing wood particle size the
formaldehyde emission increases(Boruszewski et al., 2011; Roffael et al., 2012). Therefore it is
suggested that no definite formaldehyde emission value can be defined for wood (Roffael et al.,
2012).
2.4. Wood-Formaldehyde interaction
The biogenic and synthetic formaldehyde in woody matrix can form a very complicated
equilbiria as shown in Figure 3-1. Formaldehyde can react with hydroxyl groups and water
existed in woody matrix to produce hemiacetal, acetal and methylene glycol (Figure 3-1). It is
27
stated that formaldehyde can react with both lignin and cellulose (Myers, 1986b); however its
not apparent that formaldehyde favors to react with lignin or cellulose.
Formaldehyde reacts with cellulose under neutral, alkaline, and acidic conditions to form
hemiacetal and acetal. Under neutral condition, hemiacetal formation takes place by diffusion
process; however this reaction is slow and it takes 1 day at room temperature to be equilibrated
(Walker, 1975). It is also stated that under neutral condition the reaction is reversible and
hemiacetal linkages can be cleaved easily by simple water wash (Roff, 1958; Walker, 1975). The
reaction between formaldehyde and cellulose under alkaline condition also results in temporary
alterations in the physical properties of cellulose. However it seems that the resulted linkages
under alkaline condition are more stable than that forming under neutral condition. These
linkages can be cleaved by boiling the cellulose in ammonia water or sodium bisulfate solution
or even by heating it with water alone under pressure (Wood, 1931). In contrast to the neutral
and alkaline conditions, the formaldehyde-cellulose reaction under acidic condition results in
stable acetal linkages (Figure 2-16); this results in physical and chemical changes in cellulose
structure. The reaction depicted in Figure 2-16 apparently takes place primarily with cellulose
secondary hydroxyls groups (Jones, 1964; Wagner and Pacsu, 1952). It is also reported that by
increasing the formaldehyde concentration, acidity, and temperature formaldehyde can be bound
more efficiently to cellulose (cannot be removed by neutral water rinsing) (Meyer et al., 1976;
Walker, 1975).
Cell OH + OH OH
H
H
Cell O
H
H
OHH
+
H2O
H+
Cell OHCell O
H
H
O Cell
H2O
Hemiacetal Acetal
,
Figure 2-16. The reactions of formaldehyde monomer (methylene glycol) with cellulose which
results in hemiacetals and acetals.
The resulted crosslinked cellulose has more dimensional stability and water resistance
(Walker, 1975). The acid-catalyzed reaction between cellulose and formaldehyde has been
intensively studied (Cooke and Weigmann, 1982; Minato and Fujita, 1984; Roff, 1958; Stevens
and Parameswaran, 1981; Weatherwax and Caulfield, 1978).
28
In order to remove the formaldehyde from formaldehyde sorbed-southern pine, water
extraction was conducted (Myers, 1986a). Grounded sample was placed in a stoppered flask
containing water (pH 3). Results showed that after 1 or 2 hours almost all the removable
formaldehyde was extracted from the sample.
The loss of formaldehyde from formaldehyde sorbed-southern pine samples was also
investigated at different relative humidity using flask method (Myers, 1986a). At low relative
humidity, wood (at natural pH or at pH 2-3) strongly holds the formaldehyde (Myers, 1986a),
(Kubota et al., 1977). However the formaldehyde is nearly completely released in 12 days at
33%RH and in 5 days at 80%RH. The results indicated that formaldehyde is present as monomer
(methylene glycol) or oligomer dissolved in the wood’s moisture, or as hemiacetal attached to
cellulose (Myers, 1986a).
2.5. Cross-linked Poly allylamine as formaldehyde adsorbent
Using cross-linked poly allylamine beads (PAA beads) (Figure 2-17) for formaldehyde detection
purposes was introduced by Kiba and coworkers (Kiba et al., 1999; Kiba et al., 2000). PAA
beads were used in detecting drinking water formaldehyde content. Since the most common
method (DNPH) (Selim, 1977) for detecting drinking water formaldehyde is not sensitive
enough, a pre-concentration method using PAA beads was applied (Kiba et al., 1999; Kiba et al.,
2000). By adding the PAA beads to water sample, beads adsorb the formaldehyde and then by
removing the beads and transfer them into acidic solution they desorb the formaldehyde (Figure
2-18). Before using the PAA beads the adsorption and desorption reactions were tested and the
optimum conditions (temperature, time, acid concentration) were reported (Kiba et al., 1999;
Kiba et al., 2000). It was found that PAA beads can successfully be used for detecting
formaldehyde and they increased the detection sensitivity.
29
Figure 2-17. Cross-linked poly allylamine beads
+
NH2NH
HH
OH
N
HH
+ H2O
Hemiaminal Imine
O
H H
NH
HH
OH
NH2
+H H
OH
+
+ H+N
HH
and/or
Figure 2-18. PAA-formaldehyde adsorption (up) and desorption (down) reactions
In order to capture and quantify the air-formaldehyde, a highly sensitive colorimetric
detection method made of amine functionalized polymer film was introduced (Feng et al., 2010).
Due to high reactivity of primary amine groups with formaldehyde through nucleophilic addition
reaction, it was reported that the polymer film can readily react with formaldehyde. The
polymer’s color was turned into yellow after the reaction. By monitoring the color before and
after formaldehyde exposure, the air-formaldehyde concentration was reported. It was stated that
this method is highly selective for formaldehyde. They believed that the reaction between
formaldehyde and amine groups led to imine formation. However in another study it was found
that the reaction between formaldehyde gas and amine groups leads to hemiaminal instead of
imine formation (Nomura and Jones, 2013). Amine-functionalized porous Silicas were
synthesized to capture air-aldehyde gases. The adsorption capacity of primary, secondary, and
tertiary amine functionalized polymers for capturing different aldehyde and ketones was
measured. Results showed that the aminosilica polymer with primary amines had the most
efficient formaldehyde adsorption capacity while for the aminisilica polymers with secondary
NH2
NH
OH
NH
NH2
NH2
30
amines the adsorption capacity was lower. No formaldehyde adsorption was reported for the
aminisilica polymers containing tertiary amines. As mentioned above no imine formation was
monitored after the reaction between formaldehyde and aminosilica polymer however imine was
detected for acetaldehyde, hexanal, and benzaldehyde. This is due to the stability of the imine
compound which is higher for benzaldehyde and aldehyde compare to formaldehyde (Nomura
and Jones, 2013).
In this study the PAA beads were used to extract all the available free-formaldehyde from
woody matrix. Both synthetic and biogenic free-formaldehyde can exist in a complex
equilibrium in wood. Formaldehyde can react with wood hydroxyl groups and also water to form
hemiacetal, acetal, and methylene glycol. This formaldehyde sorbing ability of wood makes the
formaldehyde extraction difficult; Hence PAA beads due to having stronger nucleophiles
(Primary NH2) may react readily with formaldehyde and extract all the woody matrix free-
formaldehyde. The lower equilibrium constants of the reaction between formaldehyde-water
(Keq~2000) and formaldehyde-primary alcohol (Keq~1310) (Guthrie, 1975) compare to
formaldehyde-amine (Keq~3000) (Kallen and Jencks, 1966) support this thought.
2.6. References
Achmann, S., M. Hermann, F. Hilbrig, V. Jerome, M. Hammerle, R. Freitag and R. Moos 2008.
Direct detection of formaldehyde in air by a novel NAD(+)- and glutathione-independent
formaldehyde dehydrogenase-based biosensor. Talanta 75: 786-791. doi: DOI
10.1016/j.talanta.2007.12.015
Altshuller, A. P. and L. J. Leng 1963. Application of the 3-methyl-2-benzothiazolone hydrazone
method for atmospheric analysis of aliphatic aldehydes. Anal. Chem. 35: 1541-1542.
Altshuller, A. P., D. L. Miller and S. F. Sleva 1961. Determination of Formaldehyde in Gas
Mixtures by the Chromotropic Acid Method. Analytical Chemistry 33: 621-625.
American Society for Testing and Materials 2006. ASTM D 5582. In ASTM D 5582. West
Conshohocken.
American Society for Testing and Materials 2002. ASTM D 6007. In ASTM D 6007. West
Conshohocken.
American Society for Testing and Materials 1996. ASTM E 1333. In ASTM E 1333. West
Conshohocken.
31
Baurichtlinie, ETB 1980. Guideline for the use of particleboards with respect to the avoidance of
unacceptable formaldehyde concentration in room air. Commission for uniform technical
building specifications.: Beuth Publishers, Berlin.
Belman, S. 1963. The fluorimetric determination of formaldehyde. Anal Chim Acta 29: 120-126.
Birkeland, M. J., L. Lorenz, J. M. Wescott and C. R. Frihart 2010. Determination of native
(wood derived) formaldehyde by the desiccator method in particleboards generated during panel
production. Holzforschung 64: 429-433. doi: Doi 10.1515/Hf.2010.070
Bohm, M., M. Z. Salem and J. Srba 2012. Formaldehyde emission monitoring from a variety of
solid wood, plywood, blockboard and flooring products manufactured for building and
furnishing materials. J Hazard Mater 221-222: 68-79. doi: 10.1016/j.jhazmat.2012.04.013
Boruszewski, P., P. Borysiuk, M. Maminski, L. Szalk and L. Danecki 2011. Formaldehyde
emission from raw materials for particleboard production at the beginning of processing chain. In
Formaldehyde emission from raw materials for particleboard production at the beginning of
processing chain, 2nd International Conference on Environmental Engineering and Applications.
Braun, D., P. Gunther and S. M. Romeira 1983. Model Compounds for the Urea-Formaldehyde
Condensation .1. Methylenediurea. Angewandte Makromolekulare Chemie 118: 119-132. doi:
DOI 10.1002/apmc.1983.051180109
Brosse, N., R. El Hage, M. Chaouch, M. Petrissans, S. Dumarcay and P. Gerardin 2010.
Investigation of the chemical modifications of beech wood lignin during heat treatment. Polymer
Degradation and Stability 95: 1721-1726.
Browne, Frederick Lincoln 1958. Theories of the combustion of wood and its control.
Bunkoed, O., F. Davis, P. Kanatharana, P. Thavarungkul and S. P. Higson 2010. Sol-gel based
sensor for selective formaldehyde determination. Anal Chim Acta 659: 251-257. doi:
10.1016/j.aca.2009.11.034
Cooke, T. F. and H. D. Weigmann 1982. The chemistry of formaldehyde release from durable
press fabrics. Part 2. Textile Chemist and Colorist 14: 136-144.
De Andrade, J.B. and R.L. Tanner 1992. Determination of formaldehyde by HPLC as the DNPH
derivative following high-volume air sampling onto bisulfite-coated cellulose filters.
Atmospheric Environment, Part A: General Topics 26A: 819-825.
Dunky, M. and K. Lederer 1982. Investigations of Molecular-Weight Distributions of Urea-
Formaldehyde-Resinous Glues. Angewandte Makromolekulare Chemie 102: 199-213. doi: DOI
10.1002/apmc.1982.051020119
Ebdon, J. R., P. E. Heaton, T. N. Huckerby, W. T. S. Orourke and J. Parkin 1984.
Characterization of Urea Formaldehyde and Melamine Formaldehyde Adducts and Resins by N-
15 Nmr-Spectroscopy. Polymer 25: 821-825. doi: Doi 10.1016/0032-3861(84)90013-2
32
Emery, J. A. 1985. Formaldehyde Release from Wood Panel Products Bonded with Phenol
Formaldehyde Adhesives. Abstracts of Papers of the American Chemical Society 189: 57-Cell.
European Committee for Standardization 1992. EN 120. In EN 120. Brussels.
European Committee for Standardization 2004. EN 717-1. In EN 717-1. Brussels.
European Committee for Standardization 1994. EN 717-2. In EN 717-2. Brussels.
Feng, L., C. J. Musto and K. S. Suslick 2010. A simple and highly sensitive colorimetric
detection method for gaseous formaldehyde. J Am Chem Soc 132: 4046-4047. doi:
10.1021/ja910366p
Frihart, C.R., J.M. Wescott, T.L. Chaffee and K.M. GOonner 2012. Formaldehyde emissions
from urea-formaldehyde and no-added-formaldehyde bonded particleboard as influenced by
temperature and relative humidity. Forest Products Journal 62(7/8).
Grigoriou, A. H. 2000. Straw-wood composites bonded with various adhesive systems. Wood
Science and Technology 34: 355-365.
Guthrie, J. P. 1975. Carbonyl addition reactions. Factors affecting the hydrate-hemiacetal and
hemiacetal-acetal equilibrium constants. Canadian Journal of Chemistry 53: 898-905.
IARC 2004. Press Release-153 IARC classifies formaldehyde as carcinogenic to humans. In
Press Release-153 IARC classifies formaldehyde as carcinogenic to humans.
Japanese Standards Organization 2001. JIS A 1460. In JIS A 1460. Tokyo.
Japanese Standards Organization 2003. JIS A 1901;. In JIS A 1901;. Tokyo.
Japanese Standards Organization 2006. JIS A 1911;. In JIS A 1911;. Tokyo.
Jiang, T., D. J. Gardner and M. G. D. Baumann 2002. Volatile organic compound emissions
arising from the hot-pressing of mixed-hardwood particleboard. Forest Products Journal 52: 66-
77.
Jones, D.M. 1964. The Distribution of Oxymethylene Groups in Formaldehyde-Treated Cotton
Fabrics. Textile Research Journal 34: 257-265.
Kallen, R.G. and W.P. Jencks 1966. Equilibria for the reaction of amines with formaldehyde and
protons in aqueous solution. A re-examination of the formol titration. The Journal of biological
chemistry 241: 58645-55878.
Kiba, N., L. Sun, S. Yokose, M.T. Kazue and T.T. Suzuki 1999. Determination of nano-molar
levels of formaldehyde in drinking water using flow-injection system with immobilized
formaldehyde dehydrogenase after off-line solid-phase extraction. Anal Chim Acta 378: 169-
175.
33
Kiba, N., R. Yagi, L. M. Sun, M. Tachibana, K. Tani, H. Koizumi and T. Suzuki 2000.
Poly(allylamine) beads as selective sorbent for preconcentration of formaldehyde and
acetaldehyde in high-performance liquid chromatographic analysis. Journal of Chromatography
A 886: 83-87. doi: Doi 10.1016/S0021-9673(00)00406-4
Kim, S. and H. J. Kim 2005. Comparison of standard methods and gas chromatography method
in determination of formaldehyde emission from MDF bonded with formaldehyde-based resins.
Bioresour Technol 96: 1457-1464. doi: 10.1016/j.biortech.2004.12.003
Knop, A. and W. Scheib. 1979. Chemistry and Applications of Phenolic Resins: Springer Verlag.
Kopf, P. W. and E. R. Wagner 1973. Formation and cure of novolacs:NMR study of transient
molecules. J. Polm. Sci., Polym. Chem. ED. 11: 939-960.
Kubota, M., S. Hirata and H. Takahashi 1977. Mechanism of generating formaldehyde from
plywood glued with urea-formaldehyde resin: I. influence of formaldehyde released at plywood
manufacturing. Hoikkaido Forest Products Research Institute 303: 5-9.
Kumlin, K. and R. Simonson 1978. UF resins 1. separation of low molecular weight components
in UF resins by means of liquid chromatography. Angewandte Makromolekulare Chemie 68:
175-184.
Ludlam, P. R. and J. G. King 1984. Size Exclusion Chromatography of Urea Formaldehyde
Resins in Dimethylformamide Containing Lithium-Chloride. Journal of Applied Polymer
Science 29: 3863-3872. doi: DOI 10.1002/app.1984.070291219
Lundquist, K. 1976. Low-molecular weight lignin hydrolysis products. Applied Polymer
Symposia 28: 1393-1407.
Lundquist, K. and L. Ericsson 1970. Acid degradation of lignin. III. Formation of
formaldehyde. Acta Chemica Scandinavica (1947-1973) 24: 3681-3686.
Lv, P., Z. A. Tang, J. Yu, F. T. Zhang, G. F. Wei, Z. X. Huang and Y. Hu 2008. Study on a
micro-gas sensor with SnO2-NiO sensitive film for indoor formaldehyde detection. Sensors and
Actuators B-Chemical 132: 74-80. doi: DOI 10.1016/j.snb.2008.01.018
Maciel, G. E., I. S. Chuang and L. Gollob 1984. Solid-State C-13 Nmr-Study of Resol-Type
Phenol Formaldehyde Resins. Macromolecules 17: 1081-1087. doi: Doi 10.1021/Ma00135a018
Marcinko, J. J., S. Devathala, P.L. Rinaldi and S. Bao 1998. Investigating the molecular and bulk
dynamics of PMDI/wood and uf/wood composites. Forest Products Journal 48: 81-84.
Marcinko, J.J., P.L. Rinaldi and S. Bao 1999. Exploring the physicochemical nature of PMDI
/wood structural composite adhesion. Forest Products Journal 49: 75-78.
Meyer, B. and C. Boehme 1997. Formaldehyde emission from solid wood. Forest Products
Journal 47: 45-48.
34
Meyer, U., K. Mueller and H. Zollinger 1976. The mechanism of catalysis in the crosslinking of
cotton with formaldehyde. Textile Research Journal 46: 756-762.
Miksch, R. R., D. W. Anthon, L. Z. Fanning, C. D. Hollowell, K. Revzan and J. Glanville 1981.
Modified Pararosaniline Method for the Determination of Formaldehyde in Air. Analytical
Chemistry 53: 2118-2123. doi: Doi 10.1021/Ac00236a040
Minato, K. and M. Fujita 1984. Distribution of formaldehyde crosslinks in cellulosic fibers and
their dimensional stabilizing effect. Mokuzai Gakkaishi 30: 833-838.
Ministry of Agriculture and Forestry 2003. JAS 233. In JAS 233. Tokyo.
Myers, G. E. 1985. The Effects of Temperature and Humidity on Formaldehyde Emission from
Uf-Bonded Boards - a Literature Critique. Forest Products Journal 35: 20-31.
Myers, G. E. 1983. Formaldehyde Emission from Particleboard and Plywood Paneling -
Measurement, Mechanism, and Product Standards. Forest Products Journal 33: 27-37.
Myers, G. E. 1986a. Mechanisms of Formaldehyde Release from Bonded Wood Products. 316:
87-106. doi: 10.1021/bk-1986-0316.ch008
Myers, G. E. 1986b. Resin Hydrolysis and Mechanisms of Formaldehyde Release from Bonded
Wood Products. Forest Products Laboratory.
Myers, G. E. and M. Nagaoka 1981. Formaldehyde Emission - Methods of Measurement and
Effects of Several Particleboard Variables. Wood Science 13: 140-150.
Nair, B.R. and D.J. Francis 1983. Kinetics and mechanism of UF reaction. Polymer 24: 626-630.
Nash, T. 1953. The colorimetric estimation of formaldehyde by means of the Hantzsch reaction.
Biochem J 55: 416-421.
Nemli, G. and G. Colakoglu 2005. The influence of lamination technique on the properties of
particleboard. Building and Environment 40: 83-87. doi: DOI 10.1016/j.buildenv.2004.05.007
Nomura, A. and C. W. Jones 2013. Amine-Functionalized Porous Silicas as Adsorbents for
Aldehyde Abatement. Acs Applied Materials & Interfaces 5: 5569-5577. doi: Doi
10.1021/Am400810s
Park, Byung-Dae, Eun-Chang Kang, Sang-Bum Park and JongYoung Park 2011. Empirical
correlations between test methods of measuring formaldehyde emission of plywood,
particleboard and medium density fiberboard. European Journal of Wood and Wood Products
69: 311-316. doi: 10.1007/s00107-010-0446-6
Pizzi, A. and K.L. Mittal 2003. Handbook of Adhesive Technology, Revised and Expanded:
Taylor & Francis.
35
Que, Z. and T. Furuno 2007. Formaldehyde emission from wood products: relationship between
the values by the chamber method and those by the desiccator test. Wood Science and
Technology 41: 267-279. doi: DOI 10.1007/s00226-006-0104-7
Risholm-Sundman, M., A. Larsen, E. Vestin and A. Weibull 2007. Formaldehyde emission -
Comparison of different standard methods. Atmospheric Environment 41: 3193-3202. doi: DOI
10.1016/j.atmosenv.2006.10.079
Roff, W.J. 1958. A review of the literature relating to the interaction of cellulose and
formaldehyde. Journal of the Textile Institute Transactions 49: T646-T663.
Roffael, E. 1978. Progress in the elimination of formaldehyde from particleboards. In Progress in
the elimination of formaldehyde from particleboards, Proc. 12th Washington State Univ. Int.
Symp. on Particleboard. . Pullman, Wash.
Roffael, E. 2006. Volatile organic compounds and formaldehyde in nature, wood and wood
based panels. Holz als Roh- und Werkstoff 64: 144-149. doi: 10.1007/s00107-005-0061-0
Roffael, E., C. Behn and B. Dix 2012. On the formaldehyde release of wood particles. European
Journal of Wood and Wood Products 70: 911-912. doi: DOI 10.1007/s00107-012-0625-8
Rosenberg, G. N. 1978. Thermal Softening as a Method of Determining Resin Cure.
Holzforschung 32: 92-96. doi: DOI 10.1515/hfsg.1978.32.3.92
Rousset, P., C. Lapierre, B. Pollet, W. Quirino and P. Perre 2009. Effect of severe thermal
treatment on spruce and beech wood lignins. Annals of Forest Science 66: 110p111-110p118.
Salem, M. Z. M., M. Bohm, J. Berankova and J. Srba 2011a. Effect of some manufacturing
variables on formaldehyde release from particleboard: Relationship between different test
methods. Building and Environment 46: 1946-1953. doi: DOI 10.1016/j.buildenv.2011.04.004
Salem, M.Z.M. and M. Bohm 2013. Understanding of formaldehyde emissions from solid wood:
an overview. BioResources 8: 4775-4790.
Salem, Mohamed, Martin Böhm, Štefan Barcík and Jitka Beránková 2011b. Formaldehyde
Emission from Wood-Based Panels Bonded with Different Formaldehyde-Based Resins. Drvna
industrija 62: 177-183. doi: 10.5552/drind.2011.1102
Salem, Mohamed, Martin Böhm, Jaromír Srba and Jitka Beránková 2012. Evaluation of
formaldehyde emission from different types of wood-based panels and flooring materials using
different standard test methods. Building and Environment 49: 86-96. doi:
10.1016/j.buildenv.2011.09.011
Salthammer, T., S. Mentese and R. Marutzky 2010. Formaldehyde in the Indoor Environment.
Chemical Reviews 110: 2536-2572. doi: Doi 10.1021/Cr800399g
Santos, R.B., P.W. Hart, H. Jameel and H. Chang 2013. Wood based lignin reactions important
to the biorefinery and pulp and paper industries. BioResources 8: 22pp.
36
Schäfer, M. and E. Roffael 2000. On the formaldehyde release of wood. Holz als Roh- und
Werkstoff 58: 259-264.
Selim, S. 1977. Separation and Quantitative-Determination of Traces of Carbonyl-Compounds as
Their 2,4-Dinitrophenylhydrazones by High-Pressure Liquid-Chromatography. Journal of
Chromatography 136: 271-277. doi: Doi 10.1016/S0021-9673(00)86279-2
Slonim, I. Y., S. G. Alekseeva, Y. G. Urman, B. M. Arshava and B. Y. Akselrod 1978. Study of
Equilibriums in Urea - Formaldehyde System Using Nmr C-13 Method. Vysokomolekulyarnye
Soedineniya Seriya A 20: 1477-1485.
Standardization, European Committee for 1996. EN 717-3. In EN 717-3. Brussels.
Stevens, M. and N. Parameswaran 1981. Effects of formaldehyde-acid catalyzed reaction on
wood ultrastructure. Wood Science and Technology 15: 287-300.
Tinkelenberg, A., H. W. Vaessen, K. W. Suen and P. G. J. Leusink 1982. Combinations of Low
Formaldehyde-Aminoplast Glues and PMDI. The Journal of Adhesion 14: 219-231. doi:
10.1080/00218468208073205
U.S.A congress 2010. formaldehyde standards for composite wood products. In formaldehyde
standards for composite wood products.
Vairavamurthy, A., J.M. Roberts and L. Newman 1992. Methods for determination of low
molecular weight carbonyl compounds in the atmosphere: a review. Atmospheric Environment,
Part A: General Topics 26A: 1965-1993.
Vianello, F., L. Zennaro and A. Rigo 2007. A coulometric biosensor to determine hydrogen
peroxide using a monomolecular layer of horseradish peroxidase immobilized on a glass surface.
Biosensors & Bioelectronics 22: 2694-2699. doi: DOI 10.1016/j.bios.2006.11.007
Wagner, R.E. and E. Pacsu 1952. Cellulose studies. XVI. The reaction of cellulose with
formaldehyde. Textile Research Journal 22: 12-20.
Walker, J.F. 1975. Formaldehyde: Krieger Publishing Co., Huntington, N.Y.
Wang, J., L. Liu, S. Y. Cong, J. Q. Qi and B. K. Xu 2008. An enrichment method to detect low
concentration formaldehyde. Sensors and Actuators B-Chemical 134: 1010-1015. doi: DOI
10.1016/j.snb.2008.07.010
Wang, Song-Yung, Te-Hsin Yang, Li-Ting Lin, Cheng-Jung Lin and Ming-Jer Tsai 2007.
Properties of low-formaldehyde-emission particleboard made from recycled wood-waste chips
sprayed with PMDI/PF resin. Building and Environment 42: 2472-2479. doi:
10.1016/j.buildenv.2006.06.009
Wang, W. H., X. Q. Zang and R. S. Lu 2004. Low formaldehyde emission particleboard bonded
by UF-MDI mixture adhesive. Forest Products Journal 54: 36-39.
37
Wang, W. L. and D. J. Gardner 1999. Investigation of volatile organic compound press
emissions during particleboard production. Part 1. UF-bonded southern pine. Forest Products
Journal 49: 65-72.
Weatherwax, R. C. and D. F. Caulfield 1978. The pore structure of papers wet stiffened by
formaldehyde crosslinking. II. Results from nitrogen sorption. Journal of Colloid and Interface
Science 67: 506-515.
Weaver, F.W. 1995. Isocyanate-wood adhesive bond. Applied Spectroscopy 49: 171-176. doi:
10.1366/0003702953963751
Weigl, M., R. Wimmer, E. Sykacek and M. Steinwender 2009. Wood-borne formaldehyde
varying with species, wood grade, and cambial age. Forest Products Journal 59: 88-92.
Wiglusz, R., I. Jarnuszkiewicz, E. Sitko and L. Wolska 1990. Hygienic aspects of the use of
pressed-wood products in residential buildings. Part I. The effect of particleboards aging on
release of formaldehyde. Bulletin of the Institute of Maritime and Tropical Medicine in Gdynia
41: 73-78.
Wood, F.C. 1931. Methylene ethers of carbohydrates. I. Cellulose methylene ether. Journal of
the Society of Chemical Industry, London 50: 411T-418T.
Young, S. 2004. Formaldehyde emission from solid wood – will it become an issue? Timber Test
Laboratories. Unpublished data.
Zhou, K. W., X. L. Ji, N. Zhang and X. R. Zhang 2006. On-line monitoring of formaldehyde in
air by cataluminescence-based gas sensor. Sensors and Actuators B-Chemical 119: 392-397. doi:
DOI 10.1016/j.snb.2005.12.038
Zhou, X. and C.E. Frazier 2001. Double labeled isocyanate resins for the solid-state NMR
detection of urethane linkages to wood. International Journal of Adhesion and Adhesives 21:
259-264. doi: 10.1016/S0143-7496(01)00013-6
38
3. Novel Liquid extraction method for detecting Native-wood Formaldehyde
3.1. Introduction
Formaldehyde was described by the Russian scientist (Alexander Michailowitsch) in the year
1885 and the technical formaldehyde synthesis was achieved in 1867 by German chemist August
Wilhelm von Hafmann. Between 1900 and 1930, the formaldehyde-based resins were started to
be used in producing wood based composites. The first adverse health effects of formaldehyde
(mostly eye irritation) were reported in 1960s which after that German Federal Agency of Health
proposed a guideline value of 0.1 ppm for formaldehyde emission in 1977. The first
formaldehyde emission regulations for wood-based composites were established in Germany and
Denmark in 1981 (Salthammer et al., 2010). In 1992, the California Air Resources Board
(CARB) classified formaldehyde as a toxic air contaminant and in 2009 they placed limits on
formaldehyde emission from wood based composites (Salem et al., 2011). In July 2010, a new
regulation was signed which was extended to all 50 states in the United States and the
Environmental Protection Agency implemented the regulations on January 1st, 2013. Based on
the new regulations, which vary by product, the limits (not to be exceeded) of wood based
composites formaldehyde emission are in the range of 50 to 100 ppb which is the most rigorous
standard in the world (U.S.A congress, 2010) and it’s close to wood-borne formaldehyde
emission (Bohm et al., 2012; Meyer and Boehme, 1997; Weigl et al., 2009).
The new regulations target the formaldehyde emission of non-structural wood based
composites, such as particleboard, medium density fiberboard, and other materials commonly
used for interior residential applications. Due to the use of hydrolytically unstable urea
formaldehyde (UF) and melamine-urea formaldehyde (MUF) adhesives in making non-structural
wood based composites, their formaldehyde emission is much more higher than hydrolytically
stable structural wood based composites, such as structural particleboard and OSB (Dunky,
1998; Meyer and Hermanns, 1986; Myers, 1983, 1986; Myers and Nagaoka, 1981; Park et al.,
2011; Salem et al., 2011). Efforts have been put into decreasing the formaldehyde emission of
the non-structural boards by manipulating adhesives (UF, MUF) formulation; for instance by
using reduced formaldehyde to urea mole ratio UF resins and formaldehyde scavengers in
39
manufacturing non-structural wood based composites (Meyer and Hermanns, 1986). However,
based on the new rigorous regulations, questions arise about the formaldehyde emission of wood
itself.
Several studies have been conducted on measuring the formaldehyde emission of non-
structural wood based composites (Bohm et al., 2012; Myers, 1985, 1983; Myers and Nagaoka,
1981; Park et al., 2011; Que and Furuno, 2007; Roffael et al., 2010; Salem et al., 2011; Salem et
al., 2012; Xiong and Zhang, 2010) but wood-borne formaldehyde emission has received little
attention and there is confliction among the reports (Birkeland et al., 2010; Bohm et al., 2012;
Boruszewski et al., 2011; Meyer and Boehme, 1997; Roffael et al., 2012; Salem and Bohm,
2013; Schäfer and Roffael, 2000; Weigl et al., 2009; Young, 2004). Native-wood formaldehyde
emission varies substantially and with influence from several parameters such as, wood species
(Bohm et al., 2012; Meyer and Boehme, 1997; Weigl et al., 2009; Young, 2004), tissue maturity
(Weigl et al., 2009), wood components (Schäfer and Roffael, 2000), temperature and process
treatments employed in the industry (Birkeland et al., 2010; Schäfer and Roffael, 2000; Young,
2004), moisture content (Meyer and Boehme, 1997) and particle size (Boruszewski et al., 2011;
Roffael et al., 2012).
Wood-borne formaldehyde exists in a complex equilibrium within the woody matrix as
crudely depicted in Figure 3-1. The existence of hydroxyl groups and water in woody matrix
results in formaldehyde diffusion into the woody matrix and getting trapped there
(physically/chemically). This ability of wood in sorbing formaldehyde suggests that in order to
extract the wood-borne formaldehyde from the woody matrix, effective extraction methods such
as, chemi-sorption strategies with large equilibrium constants should be employed. Free amines
(primary/secondary) are powerful nucleophiles and react readily with formaldehyde (Walker,
1975); the equilibrium constant (Keq) for primary amines and formaldehyde is much higher
(hemiaminal Keq ~ 3000) (Kallen and Jencks, 1966) than for the reaction of formaldehyde with
alcohols (primary alcohol hemiacetal keq ~ 1310) (Guthrie, 1975).Therefore free amines can be
used as strong formaldehyde scavengers to remove all the available formaldehyde from within
the woody matrix.
40
Figure 3-1. Complex equilibria of the Biogenic formaldehyde exists in woody matrix
Crosslinked poly (allylamine) beads were used to capture the drinking water
formaldehyde (Kiba et al., 1999; Kiba et al., 2000) ; hydrated formaldehyde was sorbed by the
beads in form of imine. In order to quantify the sorbed formaldehyde, the beads were hydrolyzed
using HCl or Sulfuric acid and the released formaldehyde was analyzed using flow-injection
system (Kiba et al., 1999) or HPLC (Kiba et al., 2000). The effect of amine-functionalized silica
containing primary, secondary and tertiary amines on capturing air formaldehyde was also
investigated (Nomura and Jones, 2013). It was found that the primary amines are the most
effective amines in sorbing formaldehyde. (Feng et al., 2010) developed a colorimetric method
for detecting air formaldehyde using primary amines polymer which was highly selective for
formaldehyde.
In this study crosslinked poly (allylamine) beads (PAA beads) were synthesized and used
to extract all the available formaldehyde within the woody matrix. The PAA beads and wood
sample were stirred in aqueous solution to make all the wood-borne formaldehyde sorbed by the
beads; then the wood sample was removed and beads were hydrolyzed using sulfuric acid and
the released formaldehyde was quantified. Before using the PAA beads with wood samples, the
formaldehyde sorption and desorption efficiency of the beads were determined with standardized
formaldehyde solutions. Novel formaldehyde capture system developed in this study was
employed to help measuring the wood-borne formaldehyde. Finally a comparison between the
wood-borne formaldehyde measured with and without using the PAA beads was conducted and
the efficiency of using PAA beads in extracting the native-wood formaldehyde was determined.
41
3.2. Experimental
3.2.1. Materials
A single Virginia pine (P. Viginiana) was harvested in August 2011 from Fishburn forest in
Blacksburg Virginia (Figure 3-2), and the log was cut into sections (~10-12 cm thick in the
longitudinal direction) which were stored in a freezer for one year. The sections contained
compression wood as shown in Figure 3-3. Tree sections were processed into rectangular
specimens suitable for flaking; specimens were soaked in water for 5 days prior to processing
into flakes with dimensions of 85mm (Radial) 75mm (Longitudinal) 0.7mm (Tangential).
Mature wood (greater than 20 yrs age) was sampled but the precise location with respect to the
compressions wood was not recorded.
Figure 3-2. Harvesting Virginia pine (P.Viginiana), Fishburn forest Blacksburg Virginia
Figure 3-3. Cross-section of the Virginia Pine (P.Viginiana) log containing compression wood
Ammonium Acetate( ), Acetic acid glacial (aldehyde free), Sodium sulfite anhydrous,
Sodium hydroxide, Formaldehyde 37% wt, HPLC water and Toluene were all obtained from
42
Fisher scientific. Acetyl acetone ( ) and Poly (allylamine hydrochloride) were obtained
from Fluka and Alfa Aesar, respectively. Deionized water was obtained from a water purification
system (Miliipore MilliQ D3-UV with 20 liter reservoir).
Adjustable Pipette (Eppendorf, 0.5-5 ml) was used for transferring the solutions.
Pre-cleaned vials (40 ml, PTFE septa, polypropylene cap) and serum bottles (50 ml, silicone
septa, aluminum seal) were purchased from Sigma-aldrich. Test tubes (Kimax, rubber-lined
screw cap, O.D. L: 16 125mm), glass shell vials (O.D. H: 12mm 35mm, plastic stopper)
and 0.45 µm PTFE syringe filters were obtained from Fisher scientific.
3.2.2. Methods
3.2.2.1. Formaldehyde analysis
Formaldehyde standard solutions were prepared daily using sodium sulfite standardization
(Walker, 1975) and working solutions were obtained by suitable dilutions from those standard
solutions. Acetylacetone reagent was prepared weekly and stored in amber bottles in the
refrigerator (Nash, 1953). Calibration standards were derived from formaldehyde standard
solutions and quantified using the Hantzsch reaction (Acetylacetone method). The concentration
of the calibration standards was below 0.4 μg/ml, which is reported to be the maximum
formaldehyde concentration for linear fluorimetric calibration curves (Belman, 1963).
Unknowns subjected to formaldehyde determination were passed through 0.45 µm PTFE syringe
filters, producing 4 ml specimens that were mixed and reacted with the acetylacetone reagent
(4ml) in sealed tubes (60°C, 10 min). Specimens were allowed to cool to room temperature and
then analyzed fluorimetrically (Belman 1963) (Perkin Elmer LS55, 30°C, excitation: 410nm,
emission: 510nm, scan range: 490-530 nm, scan speed: 200 nm/min, excitation and emission slit
width 10nm; the final fluorimetric intensity was averaged between 505nm and 515nm).
Pre-cleansing the glassware: before transferring a solution to a vial/tube, the vial/tube was first
rinsed with the solution and then completely emptied (by vigorously shaking the vial/tube); this
protocol was applied to all the sampled solutions, and this helped improve the reproducibility of
the fluorimetric determinations.
43
3.2.2.2. Preparation of poly(allylamine), PAA, beads
Poly(allylamine hydrochloride) (20g, 0.215 moles primary NH2) was mixed with sodium
hydroxide (9g, 0.22 moles) in HPLC water (500 ml). Epichlorohydrin (8g, 86 mmoles) and
toluene (150 ml) were added to the PAA solution (pre-chilled by overnight storage in the
refrigerator) and the mixture was stirred under N2 (200 rpm, 3hrs, 60°C). The mixture was
cooled to room temperature; toluene was decanted and the crosslinked beads (10.75 mmole total
free NH2/beads g, 2.1 mmole free primary NH2/ beads g,) were centrifuged (500rpm,
3min)/washed with isopropanol (50 ml) and deionized water (50 ml) for five times respectively.
Thereafter the beads were added to 2M sulfuric acid (750ml) and stirred under N2 (700 rpm,
4hrs, 70°C). This removes all adventitiously sorbed formaldehyde and also minimizes additional
formaldehyde contamination by converting to the much less nucleophilic acid bisulfate form.
The beads were then centrifuged (500rpm, 3min)/washed using deionized water (50 ml, for 5
times), vacuum dried (0.15 mmHg) overnight and stored in the refrigerator. The yield for PAA
beads preparation was ~54% which was calculated based upon the free amine form.
Before each set of experiments, beads (~ 2g) were added to 3M sodium hydroxide (100 ml) and
stirred under N2 (1hr, 25°C). The beads (free amine form) were centrifuged (500rpm,
3min)/washed with deionized water (50 ml, for 5 times), vacuum dried (0.15 mmHg) overnight
and transferred to a pre-cleaned vial. Certain masses of beads (5mg or 35mg) were then
accurately weighed and transferred to sealed glass shell vials and used.
3.2.2.3. PAA beads: sorption and desorption using formaldehyde standard solutions
3.2.2.3.1. Formaldehyde sorption
The free amine form of the PAA beads (5 mg) was added to standard formaldehyde solutions (15
ml of 0.1 μg/ml) (Primary NH2/CH2O=200) in sealed pre-cleaned vials. The sample was stirred
under N2 (25°C, 2 hrs) and then a 4ml specimen was removed and analyzed for formaldehyde
concentration using the fluorimetric acetylacetone method.
The corresponding blank was HPLC water subjected to the fluorimetric acetylacetone method.
44
3.2.2.3.2. Formaldehyde desorption
The free amine form of the PAA beads (5 mg) was added to standard formaldehyde solutions (15
ml of 0.1 μg/ml) (Primary NH2/CH2O=200) in sealed pre-cleaned vials. The sample was stirred
under N2 (25°C, 2 hrs).Thereafter the formaldehyde/amine adduct was hydrolyzed using an acid-
wash as follows: sulfuric acid (15 ml, 0.1M) was added and the sample stirred under N2 (25°C,
60 min). Allowing for the beads to completely settle, the specimen supernatant (15 ml) was
removed using a pipette and 4 ml of that was analyzed for formaldehyde using the fluorimetric
acetylacetone method. The acid-wash was conducted 4 times sequentially (60 min reaction
period for wash #1; 30 min for washes 2-4). The total desorbed formaldehyde mass was
calculated as the sum of formaldehyde determined from all four acid washings.
The corresponding blank was a mixture of beads (free amine form) (5 mg) with HPLC water (15
ml), which was subjected to the same 2hr reaction followed by the same sequential acid washing
as described above and then subjected to the fluorimetric acetylacetone method.
3.2.2.4. PAA beads: Native-wood formaldehyde analysis
Wood flakes were initially dried by cycling between vacuum (0.15 mm Hg) and N2 gas, three
times consecutively; thereafter the flakes were stored over anhydrous P2O5 and N2 for 48 hours
prior to use. From the dry wood flakes, small pieces (2.5mm (Radial) 5mm (Longitudinal)
0.7mm (Tangential)) were cut and used for native wood formaldehyde analysis.
3.2.2.4.1. Control specimens (flakes with no PAA beads):
Flakes were placed in sealed serum bottles and subjected to vacuum (0.15-0.2 mm Hg, 1 min, by
inserting a needle, connected to a vacuum line, through the serum bottle septum). The specimen-
containing serum bottle was heat treated (1hr, 25°C or 200°C) and then allowed to cool to room
temperature for 15 min (flake mass was 70mg and 5mg for 25°C and 200°C heat treatments
respectively). Via syringe and without releasing the serum bottle vacuum, HPLC water (15 ml)
was added to solvate formaldehyde (1hr, 25°C); the serum bottles were occasionally shaken to
facilitate dissolution of formaldehyde sorbed onto the bottle walls. Vacuum on the serum bottle
was then released to N2 at atmospheric pressure (needle connected to N2 line and N2 bubbler),
45
the serum bottle septum was removed and the specimen water (10 ml) was sampled using a
pipette whereby 4 ml was subjected to analysis via the fluorimetric acetylacetone method.
The corresponding blank was HPLC water (15 ml) (with no flakes), placed into a similar serum
bottle and subjected to the same procedure described above including the fluorimetric
acetylacetone method.
3.2.2.4.2. Experimental specimens (flakes with PAA beads):
3.2.2.4.2.1. Sorption analysis:
Flakes were placed in sealed serum bottles and subjected to vacuum (0.15-0.2 mm Hg, 1 min, by
inserting a needle, connected to a vacuum line, through the serum bottle septum). The specimen-
containing serum bottle was heat treated (1hr, 25°C or 200°C) and then allowed to cool to room
temperature for 15 min (flake mass was 70mg and 5mg for 25°C and 200°C heat treatments
respectively). Via syringe and without releasing the serum bottle vacuum, HPLC water (15 ml)
was added to solvate formaldehyde (1hr, 25°C); the serum bottles were occasionally shaken to
facilitate dissolution of formaldehyde sorbed onto the bottle walls. Vacuum on the serum bottle
was then released to N2 at atmospheric pressure (needle connected to N2 line and N2 bubbler);
the serum bottle septum was removed and beads (free amine form) (35 mg) were added to the
serum bottle containing the heat-treated wood flake. The serum bottle was re-sealed and allowed
to stir under N2 (1hr, 25°C). The septum was then removed and specimen water (10 ml) was
sampled using a pipette whereby 4 ml was subjected to analysis via fluorimetric acetylacetone
method. The blank was HPLC water subjected to the fluorimetric acetylacetone method.
3.2.2.4.2.2. Desorption analysis:
Flakes were placed in sealed serum bottle and subjected to vacuum (0.15-0.2 mm Hg, 1 min, by
inserting a needle, connected to a vacuum line, through the serum bottle septum). The specimen-
containing serum bottle was heat treated (1hr, 25°C or 200°C) and then allowed to cool to room
temperature for 15 min (the flakes mass was 70mg and 5mg for 25°C and 200°C heat treatments
respectively). Via syringe and without releasing the serum bottle vacuum, HPLC water (15 ml)
was added to solvate formaldehyde (1hr, 25°C); the serum bottles were occasionally shaken to
facilitate dissolution of formaldehyde sorbed onto the bottle walls. Vacuum on the serum bottle
46
was then released to N2 at atmospheric pressure (needle connected to N2 line and N2 bubbler);
serum bottle septum was removed and beads (free amine form) (35 mg) were added to the serum
bottle containing the heat-treated wood flake. The serum bottle was re-sealed and allowed to stir
under N2 (1hr, 25°C). The septum was then removed and the flakes were taken out using
tweezers; before removing the flakes, they were slightly shaken in the specimen supernatant to
dislodge beads attached to the flake, returning them to the solution. The formaldehyde/amine
adduct was hydrolyzed using an acid-wash as follows: sulfuric acid (15 ml, 0.1M) was added and
the sample stirred under N2 (25°C, 60 min). Allowing for the beads to completely settle, the
specimen supernatant (15 ml) was removed and 4 ml of that was analyzed for formaldehyde
using the fluorimetric acetylacetone method. The acid-wash was conducted 3 times sequentially
(60 min reaction period for wash #1; 30 min for washes 2-3). The total desorbed formaldehyde
mass was calculated as the sum of formaldehyde determined from all four acid washings.
The corresponding blank was a mixture of beads (free amine form) (5 mg) with HPLC water (15
ml), placed into a similar serum bottle and subjected to the same procedure described above
including the fluorimetric acetylacetone method.
3.3. Results and discussion
The objective of this effort is to develop a highly efficient extraction of formaldehyde from solid
wood. From sources external and internal to wood, formaldehyde will diffuse among wood
polymers where it is expected to adsorb and/or chemisorb (chemically react) with wood and also
with wood-water. Within the woody matrix, formaldehyde/wood/water reaction equilibria could
be expected to hinder the complete extraction of formaldehyde using a simple water wash.
Consequently one might need to water-wash a wood specimen several times in order to
completely remove all formaldehyde from the woody matrix. In contrast, the PAA beads are
expected to serve as a more efficient formaldehyde sorption media because they contain primary
amines. In order to avoid extensive water-washing, it is anticipated that wood-sorbed
formaldehyde could be removed from the woody matrix more effectively by water-washing
wood specimens in the presence of PAA beads. The greater nucleophilicity of the PAA primary
amines is expected to promote rapid formaldehyde desorption from the woody matrix because of
the more favorable formation of the hemiaminal, which could perhaps form the imine (Feng et
al., 2010; Kiba et al., 1999; Kiba et al., 2000; Nomura and Jones, 2013) as shown in Figure 3-4.
47
H
OHOH
H
+
NH2NH
HH
OH
N
HH
+ H2O
Hemiaminal Imine
H2O+
Figure 3-4. Formaldehyde reaction with PAA beads to form the hemiaminal, where after
dehydration leads to the imine.
Subsequently, the wood specimen could be removed from the PAA bead/water suspension where
after acid treatment should cause hydrolysis of the imine/hemiaminal, releasing formaldehyde
and making it available for quantitative analysis.
3.3.1. PAA beads: Sorption and desorption using formaldehyde standard solutions
3.3.1.1. Formaldehyde sorption
Prior to testing the above hypothesis with solid wood specimens, trials were conducted by
placing PAA beads into standardized formaldehyde solutions (15 ml of 0.1 μg/ml). The simple
PAA-formaldehyde sorption was determined by stirring the beads in the standard solution (25°C,
2 hr), and then analyzing the supernatant using the fluorimetric acetylacetone determination. For
determining the formaldehyde mass of all samples, the obtained fluorimetric intensity of
corresponding blank was subtracted from the sample’s fluorimetric intensity and the resulted
number was divided by the slope of the calibration curve (Figure 3-5). The detection limit was
defined by the lowest fluorimetric intensity point (~3) of the calibration curve.
48
Figure 3-5. Calibration curve used for this study (slope=376.26).
Table 3-1 indicates that within experimental error the formaldehyde sorption was
complete; from the bead/water suspension no formaldehydes was detected in the water phase. It
is reported that imine formation generates a color change detectable by the eye (Bandi et al.,
2005; Feng et al., 2010) however no such coloration was observed during this simple sorption
experiment. It is likely that imine formation was not promoted under these conditions with
excess water and no heating.
After reacting aqueous formaldehyde solution with primary/secondary/tertiary amine-
functionalized silica (25°C, 1-6 hr), (Nomura and Jones, 2013) reported no detection of the
imine by Raman spectroscopy. On the other hand, (Feng et al., 2010) reported a color change due
to imine formation after reacting gaseous formaldehyde with primary amine-functionalized
polymer films (25°C, 1-10 min). They indicated that the color change was specific to
formaldehyde and other aldehydes (acetaldehyde, butyraldehyde, and benzaldehyde) did not
cause a color change.
y = 376.26x
R² = 0.9998
0
20
40
60
80
100
120
140
160
0 0.1 0.2 0.3 0.4 0.5
Flu
ori
met
ric
inte
nsi
ty
CH2O (ug)
49
Table 3-1. Formaldehyde sorption efficiency using PAA beads in formaldehyde solution
(Primary NH2/CH2O=200; n=3).
Specimens Fluorimetric intensity Blank subtraction Remaining CH2O
(μg/ml) CH2O Sorption (%)
Blank (H2O) 37.39
1.63 (0.80) Below detection limit ~100
Experimental 39.06 (0.83)
For the results given in table 1, the NH2/CH2O mole ratio was 200. This optimum ratio
was determined from a similar analysis covering a mole ratio range of 10-1000 as shown in
Table 3-2.
Table 3-2. Formaldehyde sorption efficiency using PAA beads in formaldehyde solution with
NH2/CH2O = 10, 100, and 1000; n=3.
Specimens Fluorimetric
intensity
Blank
Subtraction
Remaining CH2O
(μg/ml) CH2O Sorption (%)
Blank
(H2O) 34.33 - - -
10 (
279.50
(40.98) 245.17 0.16 (0.026) 98.09 (0.31)
100 (
54.08
(2.78) 19.75 0.012 (0.002) 98.49 (0.24)
1000
(
35.76
(0.96) 1.43
Below detection
limit ~100
Table 3-2 indicates that NH2/CH2O = 100 provided nearly complete sorption, and that complete
sorption occurred at NH2/CH2O = 1000. From this information, a mole ratio of 200 was selected
50
and subsequently confirmed as effective (Table 1). (Kiba et al., 2000) employed an approximate
mole ratio of 4500 (primary and secondary amines/CH2O) and they reported that the optimum
time and temperature for PAA bead-formaldehyde sorption was 15min and 60°C respectively.
They also used alkaline solution (pH 10.5) to promote formaldehyde sorption. In this study
alkaline conditions were not employed because the beads had been converted to the free amine
form through prior NaOH treatment.
3.3.1.2. Formaldehyde desorption
Having demonstrated effective formaldehyde sorption onto the PAA beads, the
desorption efficiency was tested by subjecting the PAA/formaldehyde adduct to
separate/sequential acid washings. Sequential acid washings were imposed since imine formation
was not expected; consequently the hemiaminal equilibrium was manipulated through sequential
washing.
After each acid washing, supernatant specimens were withdrawn and analyzed for
formaldehyde. Summation of formaldehyde recoveries from each acid wash demonstrated that
the desorption efficiency was about 94%, Table 3. (Kiba et al., 1999) employed a one-step acid-
wash (hydrolysis) (70°C, 10min) using 1 molar HCl to completely remove all sorbed
formaldehyde (primary and secondary NH2/formaldehyde=4500). In this study, the
sorption/desorption conditions of (Kiba et al., 1999) were attempted, but the results were highly
variable. The lower temperature sorption/desorption conditions described here were subsequently
developed, and they proved to be more reproducible.
51
Table 3-3. Sequential and cumulative formaldehyde desorption efficiency over 4 sequential acid-
washes; n=3.
Acid-wash
sequence Specimens
Fluorimetric
intensity
Blank
subtraction
Removed
CH2O
(μg)
Sum of
CH2O (μg)
CH2O
Desorption
(%)
First
Blank 41.55 (1.26)
59.5 (2.19) 0.59
(0.021)
1.40
(0.04)
94.06
(2.74)
Experimental 101.03 (1.54)
Second
Blank 46.61 (2.20)
34.27 (6.8) 0.34
(0.068) Experimental 81.49 (6.5)
Third
Blank 39.31 (0.84) 23.99
(4.06)
0.23
(0.040) Experimental 63.31(3.98)
Fourth
Blank 37.83 (0.72) 11.90
(3.10)
0.23
(0.06)* Experimental 49.73 (2.58)
* Calculations were appropriately adjusted for volumetric changes caused by sampling and acid
washing
Table 3-3 indicates that the greatest quantity of formaldehyde was released in the first two acid
washings, and that residual formaldehyde leveled off to a constant value. Note that even if imine
formation was promoted, this sequential acid washing would nevertheless be required since the
hemiaminal Keq is so high. Consequently it is acknowledged that the sequential wood washing
that is avoided (by using PAA beads) becomes effectively transferred to the PAA beads.
3.3.2. PAA beads: Native-wood formaldehyde analysis
3.3.2.1. Control specimens (Flakes with no PAA beads)
In this study, the PAA beads were devised to sorb wood-borne (biogenic) formaldehyde;
but baseline (control) biogenic formaldehyde levels were first determined. Two types of baseline
specimens were analyzed 1) never-heated control wood specimens, and 2) wood specimens
52
heated for 1 hour (200°C). The analytical procedure involved sealing small wood specimens
(flake form, 5 – 70mg) into a 50ml serum bottle that was subjected to vacuum via a needle-
tipped vacuum-line inserted through the septum. Since the serum bottles were to be heated at
200°C, vacuum was applied as a safety measure (Later it was determined that the serum bottles
could withstand 200°C at atmospheric pressure, but the vacuum procedure was maintained for
this study). After heating, the serum bottles were allowed to cool and water was injected into the
bottle to initiate a room temperature extraction. After 1 hr extraction, supernatant specimens
were withdrawn and subjected to formaldehyde determination.
Table 3-4 shows that never-heated wood specimens contain detectable formaldehyde, and
a 60 min 200°C treatment generates considerably more formaldehyde, 36 times over. For
specimens heated at 200°C, only 5mg of wood is required to easily detect formaldehyde;
whereas reliable detection in never-heated specimens requires a larger wood mass, ~ 70mg.
Wood-derived biogenic formaldehyde is well known, including that heat treatments cause greater
emissions as from drying and hotpressing (McDonald et al., 2004; Meyer and Boehme, 1997;
Salem and Bohm, 2013; Schäfer and Roffael, 2000; Young, 2004). All organic wood
components have some potential to form formaldehyde upon heating, but lignin appears to have
the greatest potential (Schäfer and Roffael, 2000). Thermally induced formaldehyde generation
in lignin is thought to occur via an acidolytic reaction where the gamma-carbon is released as
formaldehyde due to activation by the alpha benzyl carbon, as shown in Figure 3-6 (Brosse et al.,
2010; Lundquist, 1976; Lundquist and Ericsson, 1970; Rousset et al., 2009; Santos et al., 2013):
Figure 3-6. Acidolytic cleavage of the β-O-4 lignin linkage leading to formaldehyde release.
53
Besides the β-O-4 shown in Figure 3-6, other lignin linkages containing activated gamma
hydroxyl groups should behave similarly as depicted in Figure 3-7.
Figure 3-7. Depiction of how other lignin linkages can also lead to formaldehyde release through
acid catalysis.
The reactions in Figures 3-6 and 3-7 have practical significance for determining the
impact of biogenic formaldehyde formation in the forest products industry. Likewise the
formaldehyde capture system developed in this work (small wood flake sealed in the serum
bottle) is very practical in determining formaldehyde emission of very small pieces of wood; this
shows that in future studies there will be no need to cut down a tree in order to study the native
wood formaldehyde emission; using an increment borer and getting small wood samples from
trees will be enough. Also serum bottle method could prove to be very convenient for
determining the potential for biogenic formaldehyde formation in various wood tissues. Since the
specimen is so small one could easily determine the biogenic formaldehyde potential for highly
localized tissue specimens, i.e. earlywood vs latewood for instance.
dibenzodioxocin
OMe
O
O
O
OMe
MeOO
H
O
OMe
OMe
O
HO
MeO
vinyl ether
C
O
HHH+
stilbene
OMe
O
OH
OMe
-5 (-O-4)
OMe
O
O
OH
OMeH
+
C
O
HH
54
Table 3-4. Measured quantities of formaldehyde water-extracted from Virginia pine specimens,
never-heated and heated at 200°C, 60 min. n = 3.
Samples specimens Fluorimetric
Intensity Blank subtraction μgCH2O/g dry wood
Never
heated
Blank 28.66 (0.84)
14.74 (0.95) 9.82 (0.64)
Experimental 43.54 (0.64)
200°C
Blank 40.64 (1.15)
91.76 (4.24) 363.76 (14.63)
Experimental 132.40 (4.65)
3.3.2.2. Experimental specimens (flakes with PAA beads):
3.3.2.2.1. Sorption and desorption analyses:
Having demonstrated the baseline formaldehyde production from never-heated and
200°C heated wood flakes, the efficacy of the beads were again tested, and in two stages,
sorption and desorption. Water and PAA beads added to the bottles after the respective heat
treatments and allowed to stir for 1 hr (25°C). Thereafter a supernatant specimen was withdrawn
and analyzed for formaldehyde. Table 3-5 shows that sorption efficiency was essentially
complete.
Table 3-5. Sorption efficiency of the PAA beads in the presence of wood specimens, never-
heated and 200°C heat treated
Samples Specimens Fluorimetric
Intensity
Blank
subtraction
CH2O
(μg/ml)
Never
heated
Blank 28.66 (0.84)
1.96 (1.78) Below
detection
limit
Experimental 29.74 (2.45)
200°C
Blank 23.87 (0.13)
0.15 (0.13)
Experimental 23.89 (0.45)
55
In contrast to all prior observations it was noted that the PAA beads took on a yellow color in the
case of the never-heated specimen, Figure 3-8. However in the case of the 200°C heated
specimen, no color change was observed during the formaldehyde sorption phase. Recall that
yellow color has been associated with imine formation. It is calculated that in case of never-
heated specimen the ratio of primary NH2 to formaldehyde increases significantly compare to
heated specimen (around 2000 more); this results the formaldehyde-beads reaction favors the
products and the imine formation (Figure 3-4) can occur.
Figure 3-8. Demonstration of PAA bead color change observed during the sorption phase for
never-heated specimens.
After the sorption phase above, wood flakes were removed from the serum bottle and
separate/sequential acid washes were conducted as before to recover the sorbed formaldehyde.
Table 3-6 shows the formaldehyde recoveries and Figure 3-9 compares the recovery with and
without the PAA beads. Regarding never-heated specimens Figure 3-9 indicates that the PAA
beads were less efficient than was the simple water extraction. This might be explained that for
never-heated specimens imine formation made the hydrolysis difficult and some of the
formaldehyde molecules were still in form of imine. It might be possible to recover all the
formaldehyde if the hydrolysis took place at higher temperature. In contrast for the 200°C
heated specimens, the PAA beads appear to have recovered the same quantity of formaldehyde
as was recovered in the simple water extraction, Figure 3-9.
56
Figure 3-9. Comparison of the recovered formaldehyde from Virginia pine specimens with and
without using the PAA beads, never-heated and 200°C heat treated
Ignoring the never-heated specimens, the results summarized in Figure 3-9 suggest that the PAA
beads offer no advantage for recovering free formaldehyde from the woody matrix. While this
was initially disappointing, the practical conclusion is perhaps more important- simple water
extraction is apparently highly efficient. This conclusion suggests that the excess water forces the
equilibrium in favor of hydrated formaldehyde as depicted in Figure 3-10. Extracting hemiacetal
from formaldehyde treated-cellulose has been reported by using water wash at room temperature
(Roff, 1958; Walker, 1975).
Figure 3-10. Depiction of lignocellulose hemiacetal formation and subsequent
hydration/hydrolysis that favors formaldehyde release from the woody matrix.
Biogenic formaldehyde forms hemiacetal linkages (center-top Figure 3-10) which can be cleaved
by simple water extraction. It is also reported the hemiacetal extraction from the formaldehyde
treated-cellulose by using simple water wash at room temperature (Roff, 1958; Walker, 1975). It
must also be feasible that biogenic formaldehyde forms acetal linkages (center-down Figure 3-
Never-heated 200C Never-heated 200C1
10
100
Fo
rma
ldeh
yd
e r
eco
vere
d (
g/g
woo
d)
No beads With beads
OHOHHO
HO
lignocellulose
OH
heat
O
CH2
OH
OHO
CH2
OH
lignocellulose
OCH2
O
H2OOHOHHO
lignocellulose
OCH2
O
CH2
HO OH
CH2
OH
OH
OH H
O
H H
OH H O
H H
O
H H O
H H
57
10) under the anhydrous and high temperature heat treatment employed in this work. Such
lignocellulose acetal linkages are expected to remain intact during PAA bead desorption.
Lignocellulose acetal linkages could be cleaved using external acid catalysis and heat(Roff,
1958; Walker, 1975), but that is also expected to generate more biogenic formaldehyde via lignin
acidolysis (Lundquist and Ericsson, 1970).
Table 3-6. Measured quantities of formaldehyde extracted using PAA beads from Virginia pine
specimens, never-heated and heated at 200°C, 60 min. n = 3.
Samples Acid-wash
sequence Specimens
Fluorimetric
intensity
Blank
subtraction
Removed
CH2O
(μg)
Sum of
CH2O
(μg)
μg
CH2O/g
flake
Never
heated
First
Blank 39.66 (1.61)
6.5 (3.03) 0.12
(0.06) 0.23
(0.09)
3.38
(1.39)
Experimental 46.16 (1.46)
Second
Blank 35.33 (0.45)
2.7 (0.94) 0.10
(0.03)* Experimental 38.05 (0.55)
200°C
First
Blank 47.27 (20.1) 38.61
(1.21)
0.051
(0.001)
1.85
(0.14)
359.80
(24.78)
Experimental 85.89 (2.17)
Second
Blank 38.79 (1.46) 26.90
(3.46)
0.035
(0.004) Experimental 65.7 (4.00)
Third
Blank 33.96 (0.49) 13.89
(1.3)
0.018
(0.001)* Experimental 47.85 (1.13)
* Calculations were appropriately adjusted for volumetric changes caused by sampling and acid
washing
58
3.4. Conclusions
The synthesized PAA beads were tested using standardized formaldehyde solution for
determining the sorption and desorption efficiencies. The formaldehyde emission of never-
heated and heat treated wood specimens were determined with and without using the PAA beads.
For heat treated wood specimen, there was no advantage of using the PAA beads for
formaldehyde extraction. Whereas for the never-heated wood specimen, using PAA beads was
even less efficient than water extraction. This shows that by using simple water extraction the
free-formaldehyde in woody matrix can be completely extracted. Another remarkable finding
was the simple and practical serum bottle technique for capturing native-wood formaldehyde
emission. By using this method, formaldehyde emission of very small pieces of wood can be
determined; this proves that in future studies there will be no need to cut down a tree in order to
study the native wood formaldehyde emission; instead, using an increment borer and getting
small wood samples from trees will be enough. This method can also be used in measuring
formaldehyde emission of specific small wood tissues such as early wood or late wood.
3.5. References
Bandi, S., S. Mehta and D. A. Schiraldi 2005. The mechanism of color generation in
poly(ethylene terephthalate)/polyamide blends. Polymer Degradation and Stability 88: 341-348.
doi: DOI 10.1016/j.polymdegradstab.2004.11.009
Belman, S. 1963. The fluorimetric determination of formaldehyde. Anal Chim Acta 29: 120-126.
Birkeland, M. J., L. Lorenz, J. M. Wescott and C. R. Frihart 2010. Determination of native
(wood derived) formaldehyde by the desiccator method in particleboards generated during panel
production. Holzforschung 64: 429-433. doi: Doi 10.1515/Hf.2010.070
Bohm, M., M. Z. Salem and J. Srba 2012. Formaldehyde emission monitoring from a variety of
solid wood, plywood, blockboard and flooring products manufactured for building and
furnishing materials. J Hazard Mater 221-222: 68-79. doi: 10.1016/j.jhazmat.2012.04.013
Boruszewski, P., P. Borysiuk, M. Maminski, L. Szalk and L. Danecki 2011. Formaldehyde
emission from raw materials for particleboard production at the beginning of processing chain. In
Formaldehyde emission from raw materials for particleboard production at the beginning of
processing chain, 2nd International Conference on Environmental Engineering and Applications.
59
Brosse, N., R. El Hage, M. Chaouch, M. Petrissans, S. Dumarcay and P. Gerardin 2010.
Investigation of the chemical modifications of beech wood lignin during heat treatment. Polymer
Degradation and Stability 95: 1721-1726.
Dunky, M. 1998. Urea-formaldehyde (UF) adhesive resins for wood. International Journal of
Adhesion and Adhesives 18: 95-107. doi: Doi 10.1016/S0143-7496(97)00054-7
Feng, L., C. J. Musto and K. S. Suslick 2010. A simple and highly sensitive colorimetric
detection method for gaseous formaldehyde. J Am Chem Soc 132: 4046-4047. doi:
10.1021/ja910366p
Guthrie, J. P. 1975. Carbonyl addition reactions. Factors affecting the hydrate-hemiacetal and
hemiacetal-acetal equilibrium constants. Canadian Journal of Chemistry 53: 898-905.
Kallen, R.G. and W.P. Jencks 1966. Equilibria for the reaction of amines with formaldehyde and
protons in aqueous solution. A re-examination of the formol titration. The Journal of biological
chemistry 241: 58645-55878.
Kiba, N., L. Sun, S. Yokose, M.T. Kazue and T.T. Suzuki 1999. Determination of nano-molar
levels of formaldehyde in drinking water using flow-injection system with immobilized
formaldehyde dehydrogenase after off-line solid-phase extraction. Anal Chim Acta 378: 169-
175.
Kiba, N., R. Yagi, L. M. Sun, M. Tachibana, K. Tani, H. Koizumi and T. Suzuki 2000.
Poly(allylamine) beads as selective sorbent for preconcentration of formaldehyde and
acetaldehyde in high-performance liquid chromatographic analysis. Journal of Chromatography
A 886: 83-87. doi: Doi 10.1016/S0021-9673(00)00406-4
Lundquist, K. 1976. Low-molecular weight lignin hydrolysis products. Applied Polymer
Symposia 28: 1393-1407.
Lundquist, K. and L. Ericsson 1970. Acid degradation of lignin. III. Formation of
formaldehyde. Acta Chemica Scandinavica (1947-1973) 24: 3681-3686.
McDonald, A. G., J. S. Gifford, D. Steward, P. H. Dare, S. Riley and I. Simpson 2004. Air
emission from timber drying: high temperature drying and re-drying of CCA treated timber. Holz
Als Roh-Und Werkstoff 62: 291-302. doi: DOI 10.1007/s00107-004-0494-x
Meyer, B. and C. Boehme 1997. Formaldehyde emission from solid wood. Forest Products
Journal 47: 45-48.
Meyer, B. and K. Hermanns 1986. Formaldehyde Release from Wood Products - an Overview.
Acs Symposium Series 316: 1-16.
Myers, G. E. 1985. The Effects of Temperature and Humidity on Formaldehyde Emission from
Uf-Bonded Boards - a Literature Critique. Forest Products Journal 35: 20-31.
60
Myers, G. E. 1983. Formaldehyde Emission from Particleboard and Plywood Paneling -
Measurement, Mechanism, and Product Standards. Forest Products Journal 33: 27-37.
Myers, G. E. 1986. Mechanisms of Formaldehyde Release from Bonded Wood Products. 316:
87-106. doi: 10.1021/bk-1986-0316.ch008
Myers, G. E. and M. Nagaoka 1981. Formaldehyde Emission - Methods of Measurement and
Effects of Several Particleboard Variables. Wood Science 13: 140-150.
Nash, T. 1953. The colorimetric estimation of formaldehyde by means of the Hantzsch reaction.
Biochem J 55: 416-421.
Nomura, A. and C. W. Jones 2013. Amine-Functionalized Porous Silicas as Adsorbents for
Aldehyde Abatement. Acs Applied Materials & Interfaces 5: 5569-5577. doi: Doi
10.1021/Am400810s
Park, Byung-Dae, Eun-Chang Kang, Sang-Bum Park and JongYoung Park 2011. Empirical
correlations between test methods of measuring formaldehyde emission of plywood,
particleboard and medium density fiberboard. European Journal of Wood and Wood Products
69: 311-316. doi: 10.1007/s00107-010-0446-6
Que, Z. and T. Furuno 2007. Formaldehyde emission from wood products: relationship between
the values by the chamber method and those by the desiccator test. Wood Science and
Technology 41: 267-279. doi: DOI 10.1007/s00226-006-0104-7
Roff, W.J. 1958. A review of the literature relating to the interaction of cellulose and
formaldehyde. Journal of the Textile Institute Transactions 49: T646-T663.
Roffael, E., C. Behn and B. Dix 2012. On the formaldehyde release of wood particles. European
Journal of Wood and Wood Products 70: 911-912. doi: DOI 10.1007/s00107-012-0625-8
Roffael, E., B. Johnsson and B. Engstrom 2010. On the measurement of formaldehyde release
from low-emission wood-based panels using the perforator method. Wood Science and
Technology 44: 369-377. doi: DOI 10.1007/s00226-010-0355-1
Rousset, P., C. Lapierre, B. Pollet, W. Quirino and P. Perre 2009. Effect of severe thermal
treatment on spruce and beech wood lignins. Annals of Forest Science 66: 110p111-110p118.
Salem, M. Z. M. and M. Bohm 2013. Understanding of Formaldehyde Emissions from Solid
Wood: An Overview. BioResources 8: 4775-4790.
Salem, Mohamed, Martin Böhm, Štefan Barcík and Jitka Beránková 2011. Formaldehyde
Emission from Wood-Based Panels Bonded with Different Formaldehyde-Based Resins. Drvna
industrija 62: 177-183. doi: 10.5552/drind.2011.1102
Salem, Mohamed, Martin Böhm, Jaromír Srba and Jitka Beránková 2012. Evaluation of
formaldehyde emission from different types of wood-based panels and flooring materials using
61
different standard test methods. Building and Environment 49: 86-96. doi:
10.1016/j.buildenv.2011.09.011
Salthammer, T., S. Mentese and R. Marutzky 2010. Formaldehyde in the Indoor Environment.
Chemical Reviews 110: 2536-2572. doi: Doi 10.1021/Cr800399g
Santos, R.B., P.W. Hart, H. Jameel and H. Chang 2013. Wood based lignin reactions important
to the biorefinery and pulp and paper industries. BioResources 8: 22pp.
Schäfer, M. and E. Roffael 2000. On the formaldehyde release of wood. Holz als Roh- und
Werkstoff 58: 259-264.
U.S.A congress 2010. formaldehyde standards for composite wood products. In formaldehyde
standards for composite wood products.
Walker, J.F. 1975. Formaldehyde: Krieger Publishing Co., Huntington, N.Y.
Weigl, M., R. Wimmer, E. Sykacek and M. Steinwender 2009. Wood-borne formaldehyde
varying with species, wood grade, and cambial age. Forest Products Journal 59: 88-92.
Xiong, J. and Y. Zhang 2010. Impact of temperature on the initial emittable concentration of
formaldehyde in building materials: experimental observation. Indoor Air 20: 523-529. doi: DOI
10.1111/j.1600-0668.2010.00675.x
Young, S. 2004. Formaldehyde emission from solid wood – will it become an issue? Timber Test
Laboratories. Unpublished data.
62
4. Appendices
Appendix A : The effect of HMR on the rheological and physical properties of
wood
Abstract
Hydroxy methylated resorcinol (HMR) is an adhesive promoter which when it is applied on
wood bonding surface it results in a highly durable bond-line. The reaction between HMR and
wood which results in this durable bond-line is not clearly understood. In this study the
rheological properties of HMR treated specimens immersed in NMP and DMF plasticizers were
studied using dynamic mechanical analysis (DMA), also the effect of HMR treatment on
volumetric swelling of wood after water soaking was studied. The volume change percentage
and weight gain percentage of wood samples after HMR treatment was also determined. Two
wood species including yellow poplar and southern yellow pine were used in this study. Results
showed that HMR treatment increased wood stiffness and decreased mobility of wood
amorphous phase (lignin) in both wood species; however the effect of HMR on pine was more
significant compared to poplar. Based on DMA results, the HMR-wood reaction seemed more
likely to be secondary interactions than covalent bonding. It was found that the detection of
HMR effect on wood is plasticizer independent. It was also revealed that HMR treatment
decreased the wood volumetric swelling.
Introduction
Hydroxy methylated resorcinol (HMR) was introduced in 1995 by Vick and coworkers (Vick et
al., 1995). HMR is an aqueous alkaline solution which is applied as a primer on wood bonding
surface and results in a highly durable bond-line. The HMR effect was successful on different
adhesives including: epoxy (Vick et al., 1995), phenol-resorcinol-formaldehyde, polymeric(Vick,
1995), diphenylmethane diisocyanate (pMDI) (Vick, 1996), vinyl ester, polyvinyl acetate
(Lopez-Anido et al., 2000), melamine formaldehyde, urea-formaldehyde, and phenol-
formaldehyde (Kurt et al., 2008). Several studies were conducted to understand the reason of
63
durable bond-line caused by HMR treatment. Vick and coworkers suggested that HMR acts as
coupling agent and forming covalent bond with both adhesive and wood (Vick et al., 1995).
Later Christiansen proved that HMR does not form covalent bond with epoxy adhesives hence he
concluded that HMR does not act as coupling agent (Christiansen, 2005). However in another
study (Szczurek et al., 2010), the presence of covalent bond between HMR and polyurethane
adhesives was proved. Christiansen (2005) also studied the effect of HMR crosslink density on
bond-line durability; he replaced some resorcinol groups with 2-methylresorcinol to decrease the
crosslink density and it resulted in lower bond-line durability (Christiansen, 2005). However
Hosen found that (Hosen, 2010) the effect of HMR is more related to forming an effective
network in wood matrix rather than having high crosslink density.
The effect of HMR solution highly depends on its molecular weight distribution. HMR
consists of methylolated monomers, dimers, and higher molecular weight oligomers; and all of
these compounds are required in order to have an efficient HMR effect (Vick et al., 1995). It is
suggested that HMR diffuses into the cell wall and crosslinks with lignin (Gardner et al., 2005).
The increase of wood stiffness after HMR treatment, which can be due to crosslinking, is
reported (Chowdhury, 2011; Sun and Frazier, 2005). Chowdhury studied the viscoelastic
behavior of HMR-treated wood specimens immersed in different solvents using dynamic
mechanical analysis (DMA) and he found that the detection of HMR effect is solvent dependent
(Chowdhury, 2011). Studies were also conducted on investigating the effect of HMR treatment
on physical properties of wood. Dimensional stability of wood veneer was measured after HMR
treatment; results showed that HMR treatment improved the dimensional stability of wood and
decreased the water uptake (Son and Gardner, 2004). However Moon and coworkers found that
HMR treatment had no effect on dimensional stability of wood (Moon et al., 2010).
Although several studies were conducted on HMR subject, HMR-wood reaction is still not
clear and there are contradictions among literatures. In this effort we tried to discover more on
the reaction between HMR and wood. The effect of HMR treatment on two wood species (pine
and poplar) was investigated using DMA. HMR-treated specimens were immersed in plasticizing
solvents (DMF, and NMP) and the effect of plasticizer soaking time on HMR-treated specimens
was also studied using DMA. For better understanding the behavior of HMR part of HMR-
treated wood, the neat-HMR samples were produced and immersed in DMF and NMP and their
64
volumetric swelling was measured as a function of swelling time. The weight gain percentage
and volume expansion of wood specimens caused by HMR treatment were also measured.
Finally the effect of HMR treatment on volumetric swelling of wood specimens after water
soaking was also studied.
Experimental
Materials and methods
HMR-wood DMA
Cylindrical specimens (8 mm dia., 4 mm thickness) were acquired from single pieces of yellow-
poplar (Liriodendron tulipifera) and southern yellow-pine (Pinus spp.) sapwood lumbers, using a
“plug cutter”. Two growth rings per centimeter appeared in the cross section of each specimen,
and the positioning of growth rings was random within specimens. All specimens were tested
with parallel-plate compressive-torsion DMA using TA instruments AR2000 rheometer. The
position of specimen in rheometer was such that the specimen’s thickness was parallel to the
torsional axis and the specimen ends were in contact with the parallel plates. Specimens were
precisely machined with the grain orientation of RT; the first letter indicates the plane which
torsion occurs, and the second letter indicates the grain direction parallel to torsional axis (T,
tangential; R, Radial). Two different plasticizing solvents were used: N, N-dimethylformamide
(DMF, >99.8%), and N-methyl-2-pyrrolidone (NMP, 99.5%). Both solvents were obtained from
Fisher scientific.
Drying wood specimens
All wood specimens were initially dried by cycling between vacuum (0.15 mm Hg) and N2 gas,
three times consecutively; thereafter the specimens were stored over P2O5 and N2 for 48 years
prior to any treatment.
HMR treatment
HMR solution (5% conc.) was prepared based on the procedure determined by Vick and
coworkers (Vick et al., 1995); aqueous mixture of resorcinol and formaldehyde was reacted
(25⁰C, 4hr) and then impregnated in wood specimens using vacuum-pressure treatment (4.5
65
mmHg, 25⁰C, 45min) followed by atmospheric pressure (25⁰C, 15min). Thereafter the
impregnated specimens were cured (25⁰C, 12hr) and then cured again in oven (150⁰C, 1hr); the
cured specimens were dried in desiccator by cycling between vacuum (0.15 mm Hg) and N2 gas,
three times consecutively; the specimens were then stored over anhydrous P2O5 and N2 for 48
hours prior to analysis. For control, a separate set of wood specimens were heat treated (150⁰C,
1hr) and dried as described before.
Effect of wood species on HMR-treated specimens
Control and HMR-treated yellow poplar and southern yellow-pine specimens were saturated
with plasticizer using vacuum-pressure treatment (4.5 mmHg, 25⁰C, 1hr) followed by
atmospheric pressure for 24 hr prior to DMA analysis. DMF and NMP were used as plasticizers.
Effect of plasticizer soaking time on HMR-treated poplar
Three sets of HMR-treated yellow poplar specimens were saturated with plasticizer using
vacuum-pressure treatment (4.5 mmHg, 25⁰C, 1hr) followed by atmospheric pressure for 1 day,
1 week, and 1 month prior to DMA analysis. DMF and NMP were used as plasticizers.
Compressive-torsion DMA
Compressive-torsion solvent-submersion DMA was performed generally as described in
previous studies (Chowdhury et al., 2010; Chowdhury and Frazier, 2013).
Specimens which were immersed in DMF were subjected to compressive-torsion DMA as
follows: Under a 20 N clamping (normal) force, impose sequential heating and cooling ramps: 1)
equilibrate, 25°C, 5 min, 2) heat 25°C - 80°C, 3°C min-1
, 1 Hz, 3) equilibrate, 80°C, 5 min, 4)
cool 80°C - -10°C, 3°C min-1
, 1 Hz.
Specimens which were immersed in NMP were subjected to compressive-torsion DMA as
follows: Under a 20 N clamping (normal) force, impose sequential heating and cooling ramps: 1)
equilibrate, 25°C, 5 min, 2) heat 25°C - 130°C, 3°C min-1
, 1 Hz, 3) equilibrate, 130°C, 5 min, 4)
cool 130°C - -20°C, 3°C min-1
, 1 Hz.
66
DMA Analysis was conducted such that the specimen was surrounded by a stainless steel cup
that maintained solvent immersion. An aluminum cover was used to reduce evaporative losses.
All analyses were conducted within the linear viscoelastic response, as determined using stress
sweep experiments at the temperature extremes. Three specimens were analyzed for each test.
Average graphs were created using Originpro software version 8.0951 (OriginLab, Northampton,
MA, USA).
Neat-HMR and HMR-wood Physical properties study
Neat-HMR in DMF and NMP
HMR solution (50% conc.) was prepared (Table 4-1); aqeous mixture of rescorcinol and
formaldehyde was reacted (25⁰C, 1hr) and transferred into vials (17mm O.D. 63 mm H) and
cured (25⁰C, 12hr). The cured specimens were cured again in oven (70⁰C, 20hr) and formed
cylindrical shapes which were taken out of the vials. In order to have similar dimensions,
surfaces of specimens were sanded. The dimensions of specimens were measured carefully using
caliper; for every thickness and diameter 2 measurements were conducted. Thereafter specimens
were saturated with plasticizer using vacuum-pressure treatment (4.5 mmHg, 25⁰C, 1hr).
Saturated samples were stored in plasticizer for 9 days and dimensions were measured daily
using caliper as mentioned before. The volumetric swelling (%) (Equation 1) was then calculated
for each specimen. DMF and NMP were used as plasticizers.
Table 4-1. HMR 50% conc. formulation
HMR ingredient Parts by weight
Water, deionized 51.43
Resorcinol, crystalline 16.34
Formaldehyde, 37 percent 16.79
Sodium hydroxide, 3M 15.44
Total 100
67
Equation 1:
HMR-wood in water
Rectangular wood blocks (4 cm radial 4cm tangential 1cm longitudinal) were cut from
single pieces of yellow-poplar (Liriodendron tulipifera) and southern yellow-pine (Pinus spp.)
sapwood lumbers. Blocks were initially dried by cycling between vacuum (0.15 mm Hg) and N2
gas, three times consecutively; thereafter the blocks were stored over P2O5 and N2 for 48 years
prior to any treatment. Blocks dimensions were carefully measured using caliper; 2
measurements in radial and tangential directions and 4 measurements in longitudinal direction.
HMR solution (5% conc.) was prepared based on the procedure determined by Vick and
coworkers (Vick et al., 1995). Wood blocks were immersed in beaker (300ml) filled with HMR
solution; screen wire was used to hold the blocks immersed in HMR. The HMR filled beaker
containing blocks were placed in a steel pressure vessel (5 L) and pressurized (20 psi, 1hr) using
N2 gas. The pressure was then released and the specimens were removed from the HMR solution
and transferred into desiccator. Using water aspirator vacuum, specimens were vacuumed (19
mm Hg, 15min); vacuum was released and blocks were taken out. One set of blocks were cured
in oven (70⁰C, 24hr) and the other set cured at room temperature (25⁰C, 24hr). Both sets were
then dried by cycling between vacuum (0.15 mm Hg) and N2 gas, three times consecutively;
thereafter the blocks were stored over P2O5 and N2 for 48 hours. The weight and dimensions of
blocks before and after HMR treatment were measured for calculating the HMR volume change
percentage (equation 2) and weight gain percentage (equation 3). Blocks were soaked in water
for 48 hours. The dimensions were also measured carefully using caliper before and after water
soaking. The control was initially dried yellow poplar, and southern yellow-pine blocks with the
same size as experimental blocks soaked in water for 48 hours. The control dimensions were
carefully measured before and after water soaking.
Equation 2:
Equation 3:
Volumetric swelling (%) =
1
HMR weight gain (%) =
1
Volume change (%) =
1
68
Results and discussion
For better understanding the reaction between HMR and wood, two different analyses were
conducted: 1) the DMA studies of HMR-treated wood specimens immersed in DMF and NMP
and 2) the physical properties study of both neat-HMR soaked in DMF and NMP and HMR-
treated wood in water.
DMA study
The effect of wood species and also plasticizer soaking time on solvent plasticized HMR-treated
specimens were investigated using compressive torsion solvent-submersion DMA.
Effect of wood species on solvent-plasticized HMR-treated specimens
Figures 4-1 and 4-2 show 1st heat and 1
st cool DMA scans and demonstrate the effect of wood
species on DMF-plasticized HMR-treated specimens. Figures 4-3 and 4-4 demonstrate the same
effect but on NMP-plasticized HMR-treated specimens. In both 1st heat and 1
st cool scans
(Figures 4-1 - 4-4), the stiffness of pine and poplar HMR-treated specimens was higher than
control; decrease in tan δ intensity also occurred for both poplar and pine after HMR treatment.
In other words, for both NMP and DMF plasticizers, HMR treatment increased the stiffness and
decreased the tan δ intensity. This supports that HMR can crosslink within the woody matrix and
decrease the mobility of wood amorphous phase (lignin). This finding was reported in other
studies as well (Chowdhury, 2011; Sun and Frazier, 2005). Literature suggested that HMR can
penetrate into the cell wall and speculated to associate with lignin (Chowdhury, 2011; Sun and
Frazier, 2005) .
69
Figure 4-1. The effect of wood species on dynamic mechanical responses of DMG-plasticized
HMR-treated specimens. Average first heating scans of pine (top) and poplar (bottom)
specimens. Error bars represent ±1 standard deviation (n=3).
Figure 4-2. The effect of wood species on dynamic mechanical response of DMF-plasticized
HMR-treated specimens. Average first cooling scans of pine (top) and poplar (bottom)
specimens. Error bars represent ±1 standard deviation (n=3).
0 20 40 60 8010
6
107
108
Loss
Modulu
s (
Pa)
Temperature (°C)
Storage
106
107
108
Control
HMR
Modulu
s (
Pa)
Storage
Loss
40 60 80
0.05
0.10
0.15
0.20
Tan
Temperature (°C)
0.05
0.10
0.15
0.20
Tan
106
107
108
CONTROL
HMR
Modulu
s (
Pa)
Storage
Loss
1st Heat
Pine
Poplar
40 60 8010
6
107
108
Modulu
s (
Pa)
Temperature (°C)
Storage
Loss
0.05
0.10
0.15
0.20
Tan
0 20 40 60 80
0.05
0.10
0.15
0.20
Tan
Temperature (°C)
1st Cool
Pine
Poplar
70
Figure 4-3. The effect of wood species on dynamic mechanical response of NMP-plasticized
HMR-treated specimens. Average first heating scans of pine (top) and poplar (bottom)
specimens. Error bars represent ±1 standard deviation (n=3).
Figure 4-4. The effect of wood species on dynamic mechanical response of NMP-plasticized
HMR-treated specimens. Average first cooling scans of pine (top) and poplar (bottom)
specimens. Error bars represent ±1 standard deviation (n=3).
0.05
0.10
0.15
0.20
Ta
n
-20 0 20 40 60 80 100 120
0.05
0.10
0.15
0.20
Ta
n
Temperature (°C)
40 60 80 100 120
0.05
0.10
0.15
0.20
Tan
Temperature (°C)
0.05
0.10
0.15
0.20
Tan
106
107
108
Control
HMR
Loss
Modulu
s (
Pa)
Storage
-20 0 20 40 60 80 100 12010
6
107
108
Modulu
s (
Pa)
Temperature (°C)
Storage
Loss
106
107
108
Control
HMR
Modulu
s (
Pa)
Storage
Loss
40 60 80 100 12010
6
107
108
Modulu
s (
Pa)
Temperature (°C)
Storage
Loss
1st Heat
Pine
Poplar
1st Cool
Pine
Poplar
71
Regarding the DMF-plasticized specimens glass transition temperatures (Tg) , table 4-2 shows
that HMR treatment resulted in significant Tg (defined as temperature of the tan δ maximum)
increase for both pine and poplar specimens. This Tg increase can be seen in both 1st heat and 1
st
cool scans. Table 4-2 also indicates that for both 1st heat and 1
st cool scans, the ΔTg values
(difference between the Tg of control and HMR-treated specimens) of pine were significantly
higher than poplar. It also shows that there was no difference between the 1st heat and 1
st cool
ΔTg values for poplar; however this difference was very significant for pine. Regarding the Tg
values of NMP-plasticized specimens, table 4-3 shows that HMR treatment also resulted in
significant Tg increase for both pine and poplar specimens. However this Tg increase was only
significant for 1st heat scans. On 1
st cool scan, the ΔTg values for both poplar and pine decreased
significantly.
Table 4-2. Glass transition temperatures (Tg: temp at tan δ max), and ΔTgs of DMF-plasticized
control and HMR-treated specimens, for poplar and pine species. Average values are shown
(n=3), and numbers in parenthesis are standard deviations.
Species Samples 1st Heat 1
st Heat ΔTg 1
st Cool 1
st Cool ΔTg
Pine
Control 52 (1.5)
16.2
47.6 (1.36)
32.4
HMR 68.2 (1.84) 80
Poplar
Control 48.2 (0.8)
8.4
46.6 (1.9)
8.2
HMR 56.6 (0.46) 54.8 (1.27)
Solvent plasticizer : DMF
72
Table 4-3. Glass transition temperatures (Tg: temp at tan δ max), and ΔTgs of NMP-plasticized
control and HMR-treated specimens, for poplar and pine species. Average values are shown
(n=3), and numbers in parenthesis are standard deviations.
Species Samples 1st Heat 1
st Heat ΔTg 1
st Cool 1
st Cool Δtg
Pine
Control 72.2 (1.65)
20.5
49.2 (1.6)
5.4
HMR 92.7 (1.65) 54.6 (2.00)
Poplar
Control 53.8 (1.32)
16.8
51.8 (1.03)
1.5
HMR 70.6 (4.02) 53.3 (2.06)
Solvent plasticizer: NMP
Based on the above results, for NMP-plasticized HMR-treated specimens, the decrease of ΔTg,
which occurred on 1st cool, may be due to the disruption of HMR-wood secondary interactions.
This disruption was not seen for DMF-plasticized HMR-treated specimens. This can be due to
the conditioning step which was applied before running the 1st cool scan. Since the conditioning
step (130⁰C, 5min) applied to NMP-plasticized specimens had higher temperature than that
(80⁰C, 5min) for DMF-plasticized specimens, HMR-wood secondary interactions might not
withstand in NMP but remained unchanged in DMF. The existence of secondary interactions
between HMR and wood can be very likely. For resorcinol formaldehyde (RF) adhesive,
considerable amount of secondary forces between adhesive and wood was well known (Pizzi and
Mittal, 2003); since HMR is also a low molecular weight RF, HMR-wood secondary
interactions can be highly possible. The hypothesis of the existence of HMR-wood secondary
interactions was also reported in another study (Vick, 1996). These interactions might be due to
the hydrogen bonding between HMR methylols and hydroxyl groups of woody matrix. However
secondary interactions are not the only linkages between HMR and wood and covalent
crosslinking can also occur. The higher stiffness and lower tan delta intensity of HMR treated
samples in NMP (1st cool scans) compared to control (Figure 4-4) might be a proof for that.
73
The higher ΔTg values of pine compare to poplar (Table 4-2, and 4-3) shows that the effect of
HMR treatment is more significant on pine. This can be due to the different lignin structures of
pine and poplar (Figure 4-5). In softwood lignin (Guaiacyl), the ortho position to the OH group is
available to react with HMR. However in hardwood lignin, which is a combination of Guaiacyl
and Syringyl lignins, Syringyl lignin can hinder this reaction. Hence the HMR can react more
efficient with softwoods than hardwoods.
Figure 4-5. Softwood and hardwood lignin monomers
The effect of HMR treatment on wood in DMF and NMP plasticizers was also investigated in
another study (Chowdhury, 2011). Chowdhury, 2011 found that the detection of HMR treatment
effect on wood is plasticizer dependent. Based on only the 1st cool scans data, he found that
HMR treatment only increased the Tg for DMF-plasticized specimens. However in our study 1st
heat scans showed that HMR treatment increased the Tg for both DMF and NMP specimens.
This shows that the detection of HMR effects is not plasticizer dependent.
Effect of plasticizer soaking time on solvent-plasticized HMR-treated poplar
Figure 4-6, shows the change of the Tg and the storage modulus at Tg of DMF-plasticized HMR-
treated poplar as a function of soaking time. Both 1st heat and 1
st cool scans show that by
increasing soaking time both storage modulus and Tg decreased. Regarding the NMP-plasticized
HMR-treated poplar (Figure 4-7), the storage modulus at Tg decreased by increasing soaking
OH
OH
OCH3
OH
OH
OCH3H3CO
OH
OH
OCH3
Guaiacyl lignin Syringyl l ignin Guaiacyl lignin
Softwood Hardwood
74
time for both 1st heat and 1
st cool scans, while the Tg values were decreased in 1
st heat scans and
remained unchanged in 1st cool scans.
The decrease of Tg and storage modulus by increasing soaking shows the plasticizing effect of
solvents on HMR-treated poplar samples.
Figure 4-6. The effect of DMF soaking time on the storage modulus at Tg (up) and the Tg
(bottom) of HMR-treated poplar. Error bars represent ±1 standard deviation (n=3).
1 7 3045
50
55
60
Tan
1 7 3045
50
55
60
Tan
2.4x107
3.0x107
3.6x107
4.2x107
4.8x107
Modu
lus (
Pa
)
2.4x107
3.0x107
3.6x107
4.2x107
4.8x107
Modu
lus (
Pa
)
1st Cool 1st Heat
Storage modulus at Tg
Tg
Soaking time (days)
75
Figure 4-7. The effect of NMP soaking time on the storage modulus at Tg (up) and the Tg
(bottom) of HMR-treated poplar. Error bars represent ±1 standard deviation (n=3).
Physical properties study
Neat-HMR in DMF and NMP
Figure 4-8, shows the change of volumetric swelling (%) of neat-HMR samples immersed in
DMF and NMP solvents by increasing soaking time. There was no volumetric swelling (%)
change for neat-HMR samples immersed in NMP over 9 days of soaking. However a significant
volumetric swelling (%) increase occurred for neat-HMR samples immersed in DMF. As it is
shown in Figure 4-9, DMF specimens almost fell apart after 9 days while there was no change in
NMP dimensions. It can be stated that DMF plasticized the neat-HMR while NMP showed no
plasticizing effect. However by looking at the DMF plasticized HMR-wood results (Figures 4-1
and 4-2) since both Tg and stiffness increased after HMR treatment, the plasticizing effect of
DMF cannot be seen on the HMR section of HMR-wood sample. This suggests that the
interaction between HMR and wood is more significant than the plasticizing effect of DMF on
the HMR section of HMR-wood. However by increasing solvent soaking time, the plasticizing
effect of DMF can be seen (Figure 4-6).
1 7 3045
50
55
60
65
70
75
Ta
n
1 7 3045
50
55
60
65
70
75
Ta
n
2.4x107
3.0x107
3.6x107
4.2x107
Modulu
s (
Pa)
2.4x107
3.0x107
3.6x107
4.2x107
Mo
du
lus (
Pa
)
1st Cool 1st Heat
Storage modulus at Tg
Tg
Soaking time (days)
76
Figure 4-8. The volumetric swelling (%) of neat-HMR samples immersed in DMF and NMP
plasticizers as a function of soaking time. Error bars represent ±1 standard deviation (n=3).
Figure 4-9. Neat-HMR samples after 9 days soaked in DMF and NMP solvents
HMR weight gain (%) and volumetric expansion (%) caused by HMR treatment
Table 4-4 shows the volumetric expansion (%), and weight gain (%) of poplar and pine after
HMR treatment (25⁰C, and 70⁰C). Results show that for all samples the weight gain (%) is
around 5%. This result is consistent with other studies which reported the weight gain (%)
between 2-5% (Moon et al., 2010; Sun and Frazier, 2005). The volumetric expansion (%) for all
samples was around 2-3 (%). It is well known that the swelling of wood is related to the cell wall
and not lumen (Hill, 2007). Hence the volume increase of samples caused by HMR treatment is
evidence for HMR penetrating into the cell wall.
Neat HMR in NMP
Neat HMR in DMF
1 2 3 4 5 6 7 8 9
0
20
40
60
80
100
Vo
lum
etr
ic s
we
llin
g (
%)
Time (day)
HMR-DMF
HMR-NMP
77
Volumetric swelling (%) of HMR-wood soaked in water
Figure 4-10 shows the effect of HMR treatment (25⁰C, and 70⁰C) on the volumetric swelling (%)
of poplar and pine blocks after 48 hours soaking in water. It shows that HMR treatment slightly
decrease the volumetric swelling (%) for both poplar and pine samples. The less volumetric
swelling (%) of HMR-treated samples may also be evidence of HMR penetrating into the cell
wall and occupying free volume.
Table 4-4. The volume change (%) and weight gain (%) of poplar and pine after HMR treatment
(25⁰C, 70⁰C). Average values are shown (n=4), and numbers in parenthesis are standard
deviations.
Wood
species
HMR
treatment (⁰C)
Weight gain
(%)
Volume change
(%)
Poplar
70 5.67 (0.104) 2.13 (0.240)
25 5.10 (0.183) 1.67 (0.119)
Pine
70 4.68 (0.133) 1.76 (0.183)
25 5.77 (0.135) 2.89 (0.140)
78
Figure 4-10. Comparison between the volumetric swelling (%) of HMR-treated (25⁰C, and
70⁰C) poplar and pine after 48 hr soaking in water. Error bars represent ±1 standard deviation
(n=4).
Conclusions
HMR treatment increased wood stiffness, Tg and decreased the tan δ intensity for both poplar
and pine species. This suuports that HMR can crosslink within the woody matrix and decrease
the mobility of wood amorphous phase (lignin). Based on DMA results, the HMR-wood reaction
seemed more likely to be secondary interactions but covalent bonding may also exist. It was
found that the effect of HMR treatment was more significant on pine than on poplar, which
might be due to their different lignin structures. Results also showed that the detection of HMR
effect on wood is plasticizer independent. It was revealed that HMR can penetrate into the cell
wall and decrease the volumetric swelling of wood.
0
4
8
12
16
Pine
Vo
lum
etr
ic s
we
llin
g (
%)
48 hr water soaking
Control
HMR25
HMR70
Poplar
79
References
Chowdhury, S. 2011. Advancing characterization techniques for structure-property
determination of in-situ lignocelluloses. Macromolecular Science and Engineering, Virginia
Tech.
Chowdhury, S., J. Fabiyi and C. E. Frazier 2010. Advancing the dynamic mechanical analysis of
biomass: comparison of tensile-torsion and compressive-torsion wood DMA. Holzforschung 64:
747-756. doi: Doi 10.1515/Hf.2010.123
Chowdhury, S. and C.E. Frazier 2013. Compressive-torsion DMA of yellow-poplar wood in
organic media. Holzforschung 67: 161-168.
Christiansen, A. W. 2005. Chemical and mechanical aspects of HMR primer in relationship to
wood bonding. Forest Products Journal 55: 73-78.
Gardner, D. J., C. E. Frazier and A. W. Christiansen 2005. Characteristics of the Wood Adhesion
Bonding Mechanism Using Hydroxymethyl Resorcinol. Wood adhesives, San Diego, California,
USA, Madison, WI: Forest Products Society: 93-97.
Hill, Callum AS 2007. Wood modification: chemical, thermal and other processes: John Wiley &
Sons.
Hosen, J.C. 2010. Fundamental Analysis of Wood Adhesion Primers. Sustainable Biomaterials,
Virginia tech.
Kurt, R., A. Krause, H. Militz and C. Mai 2008. Hydroxymethylated resorcinol (HMR) priming
agent for improved bondability of wax-treated wood. Holz Als Roh-Und Werkstoff 66: 333-338.
doi: DOI 10.1007/s00107-008-0265-1
Lopez-Anido, R., D.J. Gardner and J.L. Hensley 2000. Adhesive bonding of eastern hemlock
glulam panels with E-glass/vi-nyl ester reinforcement. Forest Products Journal 50: 43-47.
Moon, R.J., J. Wells, D.E. Kretschmann, J. Evans, A.C. Wiedenhoeft and C.R. Frihart 2010.
Influence of chemical treatments on moisture-induced dimensional change and elastic modulus
of earlywood and latewood. Holzforschung 64: 771-779.
Pizzi, A. and K.L. Mittal 2003. Handbook of Adhesive Technology, Revised and Expanded:
Taylor & Francis.
Son, J. and D. J. Gardner 2004. Dimensional stability measurements of thin wood veneers using
the Wilhelmy plate technique. Wood and Fiber Science 36: 98-106.
Sun, N. J. and C. E. Frazier 2005. Probing the hydroxymethylated resorcinol coupling
mechanism with stress relaxation. Wood and Fiber Science 37: 673-681.
Szczurek, A., A. Pizzi, L. Delmotte and A. Celzard 2010. Reaction Mechanism of
Hydroxymethylated Resorcinol Adhesion Promoter in Polyurethane Adhesives for Wood
80
Bonding. Journal of Adhesion Science and Technology 24: 1577-1582. doi: Doi
10.1163/016942410x500954
Vick, C. B. 1995. Coupling Agent Improves Durability of Prf Bonds to Cca-Treated Southern
Pine. Forest Products Journal 45: 78-84.
Vick, C.B. 1996. Hydroxymethylated resorcinol coupling agent for enhanced adhesion of epoxy
and other thermosetting adhesives to wood. Wood adhesives, Portland OR. Proc.47-55, Madison,
WI: Forest Products Society.
Vick, C.B., K. Richter, B.H. River and A.R.Jr. Fried 1995. Hydroxymethylated resorcinol
coupling agent for enhanced durability of bisphenol-A epoxy bonds to Sitka spruce. Wood and
Fiber Science 27: 2-12.
81
Appendix B : The effect of Ionic liquid on lignocellulose rheology
Abstract
Ionic liquid is gaining interest for its use in fractionating and processing lignocellulose for
producing biofuels and biomaterials. However little is known about the effect of ionic liquid on
lignocellulose rheology. In this study the rheological properties of yellow poplar immersed in
ionic liquid was studied using DMA. The effect of ionic liquid treatment on yellow poplar
immersed in water was also studied using DMA. Results showed that ionic liquid partially
dissolved yellow poplar at elevated temperatures during DMA study. It was also found that only
10 min ionic liquid treatment of wood sample at 80⁰C resulted in significant softening and Tg
decrease of wood sample.
Introduction
Ionic liquids are gaining increasing interest for use in processing and fractionating
lignocellulose in order to produce biofuels, biomaterials, paper, fibers, membranes, etc. Ionic
liquids are salts that are in liquid state at room or slightly warmer temperatures. They are also
known as RTIL (room temperature ionic liquids), molten salts, organic ionic liquids, etc. They
consist of cations and anions (Earle and Seddon, 2000; El Seoud et al., 2007; Gordon, 2001;
Sheldon, 2001; Swatloski et al., 2002; Wasserscheid and Keim, 2000; Welton, 1999). Ionic
liquids have unique characteristics which make them interesting to work on. They are green
solvents due to their very low (or zero) vapor pressure; they can be synthesized to be
biodegradable; they are are nonflammable and thermally stable and they can be produced to have
many different structures and properties since different cations and counter ions can be selected
(Laus et al., 2005). There are different opinions on recovering ionic liquids but many researchers
(El Seoud et al., 2007; Lateef et al., 2009; Zhu et al., 2006) believe that it is highly recyclable.
Presently, little is known about lignocellulose rheology in ionic liquids and it will be helpful
to begin documenting how such liquids impact the lignocellulose glass transition. This study is
an initial investigation of lignocellulose rheology in ionic liquids, and also the rheology of ionic
liquid treated lignocellulose in water. Tests were all conducted using dynamic mechanical
analysis, DMA; and the change of stiffness and glass transition temperature were studied.
82
Experimental
Materials
Cylindrical specimens (8 mm dia., 4 mm thickness) were machined from a single piece of
yellow-poplar (Liriodendron tulipifera) sapwood lumber, using a “plug cutter”. Wood density
was not measured, but on average, two growth rings per centimeter appeared in each specimen
cross section. Growth ring positioning within specimens was random. All Specimens were tested
in parallel-plate compressive-torsion using TA instruments AR1000 and AR2000 rheometers.
The specimen position in rheomter was such that the thickness was parallel to the torsional axis
with the ends in contact with the parallel plates. Specimens were precisely machined with the
three typical grain orientations named here as RT, TR and XL; the first letter indicates the plane
in which torsion occurs, (the wood surface contacting the parallel plates: R, radial; T, tangential;
and X, cross sectional); and the second letter indicates the grain direction parallel to torsional
axis (T, tangential; R, radial; and L, longitudinal).
1-ethyl-3- methyl imidazolium acetate ([C2mim]OAc) (95% purity) was obtained from Lolitec
and used as plasticizing agent.
Methods
Wood specimens were initially dried by cycling between vacuum (0.15 mm Hg) and N2 gas,
three times consecutively; thereafter the specimens were stored over anhydrous P2O5 and N2 for
at least 48 hours prior to any treatment.
Two different analyses were conducted: 1) Ionic liquid DMA (IL-DMA); specimens saturated
with ionic liquid and then subjected to DMA in the ionic liquid, and 2) Water-DMA; specimens
saturated with ionic liquid, subjected to thermal treatment, water extracted to remove the ionic
liquid, and then subjected to DMA in water.
IL-DMA
Specimens with all grain orientations were saturated in the ionic liquid (1 hr vacuum, 4.5 mm
Hg; then soak at atmospheric pressure 24 hr), and subsequently subjected to compressive-torsion
DMA while immersed in the ionic liquid as follows. Under a 10 N clamping (normal) force,
83
equilibrate (25°C, 5 min) and then heat (25-160°C; 3°C min-1
; 1 Hz) until the specimen lost
mechanical integrity.
Water-DMA
Only TR grain orientation specimens were used in water-DMA analysis. Two kinds of specimens
were prepared, IL-H2O-80 and IL-H2O-25; for IL-H2O-80, specimens were saturated in the ionic
liquid (1 hr vacuum, 4.5 mm Hg; then soak at atmospheric pressure 15 min) and then transferred
to, and soaked in, a preheated container of the ionic liquid (80°C, 10 min); thereafter the ionic
liquid was removed by water extraction (soak in 1 L of stirring water, 24 hr; exchange with fresh
water; continue soaking, 48 hr), and the specimens were dried (as described in methods), and
saturated in water (1 hr vacuum, 4.5 mm Hg; then soak at atmospheric pressure 24 hr). The
preparation procedure for IL-H2O-25 specimens was exactly same as IL-H2O-80 specimens
except for the thermal treatment step which is omitted for IL-H2O-25 specimens. Control
specimens were only saturated in water (1 hr vacuum, 4.5 mm Hg; soak at atmospheric pressure
24 hr). IL-H2O-80, IL-H2O-25 and control specimens were then subjected to compressive-torsion
DMA while immersed in water as follows. Under a 20 N clamping (normal) force, impose
sequential heating and cooling ramps: 1) equilibrate, 25°C, 5 min, 2) heat 25°C - 95°C, 3°C min-
1, 5 Hz, 3) equilibrate, 95°C, 5 min, 4) cool 95°C - 25°C, 3°C min
-1, 5 Hz.
Compressive-torsion, solvent-submersion DMA was performed generally as described in
previous studies (Chowdhury et al., 2010; Chowdhury and Frazier, 2013). Analysis was
conducted such that the specimen was surrounded by a stainless steel cup that maintained solvent
immersion. An aluminum cover was used to reduce evaporative losses. All analyses were
conducted within the linear viscoelastic response, as determined using stress sweep experiments
at room temperature for IL-DMA, and at the temperature extremes for Water-DMA. Three
specimens were analyzed for each treatment. Average graphs were created using Originpro
software version 8.0951 (OriginLab, Northampton, MA, USA).
84
Results and discussion
To see the effect of ionic liquid on rheological behavior of lignocellulose, two different DMA
analyses were conducted, including IL-DMA and water-DMA. In IL-DMA, the DMA was
conducted while the wood specimen was immersed in ionic liquid. While in Water-DMA, ionic
liquid treatment was applied to wood specimen and after removing all the ionic liquid, DMA was
conducted while the wood specimen was immersed in water.
IL-DMA
Figure 4-11, shows the DMA scans of three grain orientations poplar wood specimens immersed
in ionic liquid. By looking at the storage moduli, in a range between 25⁰C to 95⁰C, the stiffness
of XL and TR specimens were very similar and both were higher than RT specimen. This shows
different behavior of various grain orientations in ionic liquid. Same result is also reported in
other studies using different solvents (Chowdhury and Frazier, 2013; Hoglund et al., 1976).
20 40 60 80 100 120 140
106
107
108
20 40 60 80 100 120 140
0.1
0.2
0.3
0.4
RT
XL
TR
Modulu
s (P
a)
Loss Tan
Temperature (°C)
Storage
Figure 4-11. DMA scans of IL-DMA specimens as a function of grain orientation. Average first
heating scans are shown with error bars representing ±1 standard deviation (n=3).
Figure 4-11 also shows an abrupt softening which started at around 60⁰C for XL and TR and
around 80⁰C for RT samples. This abrupt softening might be related to wood Tg (glass transition
85
temperature), however by looking at tan δ graphs no distinct peak (which corresponds to glass
transition temperature) was observed. As can be seen in Figure 4-11, the tan δ error bars are
getting bigger after around 100⁰C especially for XL and TR samples; by looking at the gap
change of specimens during the same temperature ramp (Figure 4-12), a significant gap decrease
can be seen for XL and TR at around 130⁰C and for RT at around 160⁰C. This gap decrease
along with the big tan δ error bars (Figure 4-11) shows that the specimens lost their integrity and
fell apart. The disintegrated specimens after temperature ramp are shown in Figure 4-13. It is
well known that ionic liquids are capable of dissolving wood (both cellulose and lignin) (Feng
and Chen, 2008; Pu et al., 2007; Remsing et al., 2006; Sun et al., 2009; Swatloski et al., 2002).
Since ionic liquid had a very strong degradation effect on wood specimens, it seemed that
instead of immersing and heating wood specimen directly in ionic liquid, a short ionic liquid
treatment can be enough to see its effect on wood. Hence next series of experiments (Water-
DMA) are focusing on the effect of ionic liquid treatment on wood rheological property while
the sample is immersed in water.
20 40 60 80 100 120 140 160
2800
3200
3600
4000
4400
4800
XL
Ga
p (
Mic
rom
ete
r)
Temperature (°C)
RT
TR
Figure 4-12. The gap changing of ionic liquid saturated specimens (IL-DMA) as a function of
grain orientation. Average first heating scans are shown with error bars representing ±1 standard
deviation (n=3).
86
Figure 4-13. Disintegrated RT and TR specimens after DMA temperature ramp
Water-DMA
Figure 4-14, shows the effect of different ionic liquid treatments on wood rheological
property. As it was described in methods, the only difference between IL-H2O-80 and IL-H2O-
25, is the thermal treatment of specimens in ionic liquid which was omitted for IL-H2O-25
samples. Figure 4-14 shows that only 10 minutes thermal treatment at 80⁰C in ionic liquid can
result in a significant decrease of stiffness, which can be seen in both 1st heat and 1
st cool scans.
On the other hand there was no difference between the stiffness of control and IL-H2O-25. By
looking at 1st heat and 1
st cool tan δ graphs (Figure 4-14), tan δ intensity of IL-H2O-80 was
significantly higher than both IL-H2O-25 and control. This shows that the mobility of wood
amorphous phase (lignin) was more intense in IL-H2O-80 than in IL-H2O-25 and control. There
was a slight increase in tan δ intensity of IL-H2O-25 compare to control in 1st heat but no
difference was seen in 1st cool. All these indicate that thermal treatment, even just 10 minutes, is
necessary in order to see the wood softening effect of ionic liquid.
There is also a shoulder in all 1st heat tan δ graphs which was gone in 1
st cool. This shoulder
which was also reported in other studies (Chowdhury et al., 2010; Furuta et al., 1998; Ishimaru et
al., 2001) was confirmed to be related to drying history of the sample (Furuta et al., 1998).
Table 4-5 also shows that thermal treatment of specimens in ionic liquid resulted in a
decrease of wood Tg (defined as temperature of the tan δ maximum) around 4 degrees; however
TR RT
87
there was no significant difference between control and IL-H2O-25 Tg values. It seems that ionic
liquid at room temperature did not have significant effect on plasticizing wood.
10 20 30 40 50 60 70 80 90
107
108
10 20 30 40 50 60 70 80 90
0.08
0.12
0.16
0.20
0.24
10 20 30 40 50 60 70 80 90
107
108
10 20 30 40 50 60 70 80 90
0.08
0.12
0.16
0.20
0.24
IL-H2O-80
IL-H2O-25
Control
First Heat
Temperature (°C)
Ta
n
Storage
Loss
Loss
Storage
Modulu
s (
Pa)
First Cool
Figure 4-14. The dynamic mechanical response of water-DMA specimens; First heat and first
cool scans are shown. Average Tan δ, storage and loss modulus of first heat and cool scans, error
bars representing ±1 standard deviation (n=3).
Table 4-5. The glass transition temperatures of water-DMA specimens (Tg: temp at tan δ max).
Average (n=3) values with standard deviation in parenthesis.
Treatment First Heat Tg First Cool Tg
IL-H2O-80
85.5 (1.9) 85.9 (1.2)
IL-H2O-25 89.4 (1.3) 89.8 (0.11)
Control 90.3 (1.1) 91.5 (0.49)
88
Considering all the water-DMA results, one can say that the effect of ionic liquid in wood
softening and increasing its amorphous phase mobility can only be seen at elevated temperatures.
In other words ionic liquid treatment at room temperature does not have any effect on
plasticizing wood. This statement is in consistent with other studies (Pu et al., 2007; Swatloski et
al., 2002). They found that the solubility of both lignin (Pu et al., 2007) and cellulose (Swatloski
et al., 2002) in ionic liquid is dependent upon the type of ionic liquid and the applied
temperature. They found that some of the ionic liquids had no effect on lignin or cellulose at
room temperature. They tested different types of ionic liquids with different anions and
concluded that the nature of counter anions has effect on lignin and cellulose ionic liquid
solubility (Pu et al., 2007; Swatloski et al., 2002). It is also found that the ionic liquids containing
longer chain cations are less reactive in dissolving cellulose (Swatloski et al., 2002).
Conclusions
The ionic liquid that was used in this study (1-ethyl-3- methyl imidazolium acetate) disintegrated
lignocellulose at high temperature during DMA temperature ramp analysis. This shows the
partially dissolving effect of ionic liquids on lignocellulose. It was also revealed that the
lignocellulose plasticizing effect of ionic liquid can only be seen when it is combined with
thermal treatment. Results showed that only 10-minute thermal treatment (80⁰C) of
lignocellulose in ionic liquid can significantly decrease lignocellulose stiffness and Tg.
References
Chowdhury, S., J. Fabiyi and C. E. Frazier 2010. Advancing the dynamic mechanical analysis of
biomass: comparison of tensile-torsion and compressive-torsion wood DMA. Holzforschung 64:
747-756. doi: Doi 10.1515/Hf.2010.123
Chowdhury, S. and C.E. Frazier 2013. Compressive-torsion DMA of yellow-poplar wood in
organic media. Holzforschung 67: 161-168.
Earle, M. J. and K. R. Seddon 2000. Ionic liquids. Green solvents for the future. Pure and
Applied Chemistry 72: 1391-1398. doi: DOI 10.1351/pac200072071391
89
El Seoud, O. A., A. Koschella, L. C. Fidale, S. Dorn and T. Heinze 2007. Applications of ionic
liquids in carbohydrate chemistry: A window of opportunities. Biomacromolecules 8: 2629-
2647. doi: Doi 10.1021/Bm070062i
Feng, L. and Z. I. Chen 2008. Research progress on dissolution and functional modification of
cellulose in ionic liquids. Journal of Molecular Liquids 142: 1-5. doi: DOI
10.1016/j.molliq.2008.06.007
Furuta, Y., M. Norimoto and H. Yano 1998. Thermal-softening properties of water-swollen
wood V. The effects of drying and heating histories. Mokuzai Gakkaishi 44: 82-88.
Gordon, C.M. 2001. New developments in catalysis using ionic liquids. Applied Catalysis, A:
General 222: 101-117.
Hoglund, H., U. Sohlin and G. Tistad 1976. Physical-Properties of Wood in Relation to Chip
Refining. Tappi 59: 144-147.
Ishimaru, Y., S. Narimoto and I. Iida 2001. Mechanical properties of wood swollen in organic
liquids with two or more functional groups for hydrogen bonding in a molecule. Journal of Wood
Science 47: 171-177.
Lateef, H., S. Grimes, P. Kewcharoenwong and B. Feinberg 2009. Separation and recovery of
cellulose and lignin using ionic liquids: a process for recovery from paper-based waste. Journal
of Chemical Technology and Biotechnology 84: 1818-1827. doi: Doi 10.1002/Jctb.2251
Laus, G., G. Bentivoglio, H. Schottenberger, V. Kahlenberg, H. Kopacka, H. Roeder, T. Roeder
and H. Sixta 2005. Ionic liquids: current developments, potential and drawbacks for industrial
applications. Lenzinger Berichte 84.
Pu, Y. Q., N. Jiang and A. J. Ragauskas 2007. Ionic liquid as a green solvent for lignin. Journal
of Wood Chemistry and Technology 27: 23-33. doi: Doi 10.1080/02773810701282330
Remsing, R. C., R. P. Swatloski, R. D. Rogers and G. Moyna 2006. Mechanism of cellulose
dissolution in the ionic liquid 1-n-butyl-3-methylimidazolium chloride: a C-13 and Cl-35/37
NMR relaxation study on model systems. Chemical Communications: 1271-1273. doi: Doi
10.1039/B600586c
Sheldon, R. 2001. Catalytic reactions in ionic liquids. Chem Commun (Camb): 2399-2407.
Sun, N., M. Rahman, Y. Qin, M. L. Maxim, H. Rodriguez and R. D. Rogers 2009. Complete
dissolution and partial delignification of wood in the ionic liquid 1-ethyl-3-methylimidazolium
acetate. Green Chemistry 11: 646-655. doi: Doi 10.1039/B822702k
Swatloski, R. P., S. K. Spear, J. D. Holbrey and R. D. Rogers 2002. Dissolution of cellose with
ionic liquids. J Am Chem Soc 124: 4974-4975. doi: Doi 10.1021/Ja025790m
90
Wasserscheid, P. and W. Keim 2000. Ionic liquids - New "solutions" for transition metal
catalysis. Angewandte Chemie-International Edition 39: 3772-3789. doi: Doi 10.1002/1521-
3773(20001103)39:21<3772::Aid-Anie3772>3.0.Co;2-5
Welton, T. 1999. Room-temperature ionic liquids. Solvents for synthesis and catalysis. Chemical
Reviews 99: 2071-2083. doi: Doi 10.1021/Cr980032t
Zhu, S. D., Y. X. Wu, Q. M. Chen, Z. N. Yu, C. W. Wang, S. W. Jin, Y. G. Ding and G. Wu
2006. Dissolution of cellulose with ionic liquids and its application: a mini-review. Green
Chemistry 8: 325-327. doi: Doi 10.1039/B601395c
91