optimization of urea -formaldehyde resins for the ... · and formaldehyde emission of...
TRANSCRIPT
Optimization of Urea
for the manufacture of wood
João Miguel Macias Ferra
Dissertation presented for the degree ofDoctor Engineering
by
University of Porto SupervisorsFernão Domingos de Montenegro BaptistaMalheiro de Magalhães
Luísa Maria Hora de Carvalho
Mário Rui Pinto Ferreira Nunes da Costa
LEPAE ‐‐‐‐ Laboratory of Engineering Processes, Environment and Engineering
Chemical Engineering DepartmentFaculty of Engineering
imization of Urea-Formaldehyde resins
the manufacture of wood-based panels
João Miguel Macias Ferra
Dissertation presented for the degree of Doctor of Philosophy in Chemical and Biological Engineering
University of Porto
Supervisors Fernão Domingos de Montenegro Baptista Malheiro de Magalhães
Luísa Maria Hora de Carvalho
Mário Rui Pinto Ferreira Nunes da Costa
Laboratory of Engineering Processes, Environment and Engineering
Chemical Engineering Department Faculty of Engineering – University of Porto
Porto, 2010
(Panta Rei)
Everything flows
Everything is changing
iii
ACKNOWLEDGEMENTS
First of all, I am grateful to Fundação para a Ciência e a Tecnologia for my Ph.D.
grant (SFRH/BD/23978/2005).
My sincere appreciation and thanks to my supervisors, Professor Fernão
Magalhães, Professor Luísa Carvalho and Professor Mário Rui Costa, for showing
me how to be thorough and independent and for offering their knowledge,
experience and support for this thesis.
I would also like to express my gratitude to Professor Adélio Mendes, for his
encouragement, insightful comments and hard questions. My sincere thanks also
Professor Jorge Martins, for the stimulating discussions about wood-based panels
industry and everything.
I am indebted to EuroResinas – Indústrias Químicas S.A., in particular to Eng.
Miguel Nogueira, Dr. João Paulo Liberal, Dr. Pedro Mena, and Margarida Nogueira
for the information and knowledge about UF resins and for providing the resins,
indispensable to my work.
A special thank Dr. Martin Ohlmeyer for having received me at Johann Heinrich von
Thuenen Institute (vTI), Institute of Wood Technology and Wood Biology (HTB),
Hamburg, Germany and for the supervision during my 3 months internship. To
Daniel Karspinsky which never hesitate to help me.
I thank my colleagues in the laboratory João Ranita, João Pereira, Brígida and Filipe
Silva for help during parts of my work.
I am grateful to the Chemical Engineering Department at FEUP, LEPAE-Laboratory
for Process, Environmental and Energy Engineering, and ARCP-Association
Competence Network in Polymers for the resources availability that leads this work
to a good end.
iv
Thanks very much to the colleagues from E319 (Joana, Sofia, Klara, Daniela, Filipa,
Herney, Vânia, Pedro, Filipe, Marta) for the fantastic working environment and for
friendship during the 4 years. Also, I would like to thank my friends from FEUP,
Diogo, Águia, Miguel, Bin, Pópulo, Pedro Silva, and all people of the FEUPSal team
for the nice moments in coffee breaks, lunches, football and parties.
My deepest gratitude goes to my lovely family: my precious parents Agostinho and
Julia and my brothers Cláudia and Oscar for all their love and encouragement.
Finally, always above all and above everything, thanks to my dear wife, Mayra for
supporting, for believing… for all! You have made my life more beautiful.
v
ABSTRACT
The present thesis is focused on the study of urea-formaldehyde (UF) resins, which
are widely used in the wood-based panels industry, an important sector of the
Portuguese industrial scene. In 2009, the annual production was approximately 1.3
million m3, about 70 % of which were exported. Fundamental studies on UF resins
are therefore important in order to support innovation and international
competitiveness. There is still a limited amount of information available in this area
and new studies have to deal with chemical and physical characterization of the
resins and optimization of synthesis processes. In addition, the relation between
the operating conditions of the polymerization reactors and the performance of
the resins in the bonding process is still unclear.
A new challenge has risen after the reclassification of formaldehyde as a carcinogen
agent by the International Agency for Research on Cancer (IARC), in 2006. This
forced the manufacturers of UF resins to reduce substantially the F/U molar ratio,
originating a significantly decrease in UF resin performance and therefore
evidencing the need for the study and optimization of the synthesis strategies.
The first part of this work is dedicated to the morphological study of the UF resins.
In a first approach, the resin was studied using different microscopic techniques,
which allowed the visualization of colloidal structures in the UF resins and the
formation of large aggregates during ageing. The disperse phase and the
continuous phase were separated by centrifugation in order to be studied
independently. Two mechanisms for colloidal stabilization were suggested, based
on the results obtained with flocculation test and particle size distribution. After a
fundamental study, methods for the characterization of UF resins by HPLC and
GPC/SEC were developed. These methods were validated by the analysis of several
tens of commercial resins from EuroResinas-Indústrias Químicas S.A. and other
European producers. The effects of ageing on the molecular weight distribution, as
well on the monomeric fraction of UF resins were evaluated using the developed
vi
methods in GPC/SEC and HPLC, respectively.
An experimental design methodology was applied to optimize the internal bonding
and formaldehyde emission of particleboards produced from several resin
formulations. The number of urea additions, the time between urea additions, and
the pH of the condensation step were the three factors studied. The results
obtained showed that the sequential addition of the urea in the condensation step
affects the properties of UF resins, as well as the pH of condensation, which is the
factor that most affects the rate of the condensation reaction. Then the best
results concerning the global performance of the resins were obtained conjugating
the three factors studied.
Finally, new processes for UF resins synthesis were studied and implemented. The
best resins were produced according to the so-called strongly acid process. The
curing behaviour of laboratory and commercial resins was studied using two recent
techniques: Integrated Pressing and Testing System (IPATES) and Automated
Bonding Evaluation System (ABES). The information provided by each test system is
discussed. The optimized resin obtained using the design of experiments
methodology always presented the best performances.
vii
SUMÁRIO
A presente tese é focada no estudo das resinas ureia-formaldeído (UF), que são
amplamente usadas na indústria dos painéis derivados de madeira, um importante
sector do tecido industrial português. Em 2009, a produção anual foi de
aproximadamente 1.3 milhões de m3, sendo que cerca de 70 % foi exportada.
Estudos fundamentais sobre as resinas UF são portanto importantes, para suportar
inovação e competitividade internacional. Há ainda uma limitada quantidade de
informação disponível nesta área e novos estudos tem de estar relacionados com a
caracterização química e física destas resinas, e optimização o processo de síntese.
Além disso, a relação entre as condições operatórias dos reactores de
polimerização e o desempenho das resinas no processo ainda não é clara.
Um novo desafio surgiu após a reclassificação do formaldeído como agente
carcinogénio por parte da International Agency for Research on Cancer (IARC) em
2006. Isto forçou as empresas produtoras de resinas UF a reduzir substancialmente
a razão molar F/U, originando uma significativa redução do desempenho das
resinas UF, evidenciando a necessidade de estudar a optimização das estratégias
de síntese.
A primeira parte deste trabalho é dedicada ao estudo morfológico das resinas UF.
Numa primeira abordagem, as resinas foram estudadas através de técnicas
microscópicas, que permitiu a visualização de estruturas coloidais nas resinas UF e
a formação de grandes agregados durante o envelhecimento. A fase dispersa e a
fase contínua foram separadas por centrifugação de forma a serem estudadas
independentemente. Dois mecanismos para a estabilização coloidal foram
sugeridos, fundamentados nos resultados obtidos com testes de floculação e
distribuição de tamanho de partículas. Após um estudo mais fundamental, foram
desenvolvidos métodos de caracterização de resinas UF por HPLC e GPC/SEC. Os
métodos foram validados pela análise de várias dezenas de resinas industriais
viii
provenientes da EuroResinas-Indústrias Químicas S.A. e outros produtores
Europeus. O efeito do envelhecimento na distribuição das massas moleculares bem
como na fracção monomérica das resinas UF foi avaliado com recurso aos métodos
desenvolvidos em GPC/SEC e HPLC, respectivamente.
Aplicou-se uma metodologia de planeamento de experiências para optimizar a
resistência interna e emissão de formaldeído dos painéis de aglomerado de
partículas produzidos pelas várias formulações de resinas. O número de adições de
ureia, o tempo entre adições de ureia e o pH da etapa condensação foram os três
factores estudados. Os resultados obtidos mostraram que a adição faseada da
ureia na etapa de condensação afecta muito as propriedades das resinas UF, assim
como o pH de condensação, que é o factor que mais afecta a velocidade da reacção
de condensação. Assim, os melhores resultados considerando o desempenho
global das resinas UF foram obtidos conjugando os três factores estudados.
Finalmente, novos processos de síntese de resinas UF foram estudados e
implementados. As melhores resinas foram produzidas de acordo com o
denominado processo fortemente ácido. O comportamento de cura das resinas do
laboratório e as resinas comerciais foi estudado com recurso a duas recentes
técnicas: Integrated Pressing and Testing System (IPATES) e o Automated Bonding
Evaluation System (ABES). A informação obtida por cada sistema teste é discutida.
A resina optimizada obtida através da metodologia de planeamento de
experiências apresenta sempre o melhor desempenho.
ix
RÉSUMÉ
La présente thèse porte sur l´étude des résines urée-formaldéhyde qui sont très
utilisées dans l’industrie des panneaux à base de bois, une filière très important de
l’industrie portugaise. En 2009, la production annuel a été de approximativement
1.3 million de m3, et environ 70 % de la production a été exportée. Donc, des
études fondamentales concernant les résines UF sont important pour soutenir
l’innovation et la compétitivité internationale. L’information existante dans ce
domaine est encore limitée et des nouvelles études devront porter sur la
caractérisation chimique et physique des résines UF et optimisation du procédé de
synthèse. En outre, la relation entre les conditions opératoires des réacteurs de
polymérisation et les propriétés des résines lors des procédés de collage est encore
peu claire.
Un nouveau défi a émergé après la reclassification du formaldéhyde comme
substance cancérigène par l’IARC (International Agency for Research on Cancer), en
2006. Par conséquence, les producteurs des résines UF ont été forcé à réduire
substantiellement le rapport molaire F/U, ce qui a provoqué une diminution
significative de la performance de la résine et donc attestant la nécessité d’un
étude et optimisation des stratégies de la synthèse.
La première partie de ce travail est dédiée à l´étude morphologique des résines UF.
Dans une première approche, la résine a été étudiée par différentes techniques
microscopiques, ce qui a permet de visualiser des structures colloïdales dans la
résine UF et la formation de grands agrégats pendant le vieillissement. La phase
disperse et la phase continue ont été séparées par centrifugation pour permettre
que les deux phases d’une forme isolé. Deux mécanismes de stabilisation colloïdal
ont été suggérés, basés sur des résultats des essais de floculation et mesure des
distribution de poids moléculaires. Après l’étude fondamentale, des méthodes
pour la caractérisation des résines UF par HPLC et GPC/SEC ont été développées.
x
Ces méthodes ont été validées para l’analyse de dizaine de résines de l’entreprise
EuroResinas- Indústrias Químicas S.A. et d’autres producteurs européens. Les effets
du vieillissement sur la distribution des poids moléculaires, ainsi que sur la fraction
monomérique des résines UF ont été évalués avec les méthodes développées en
GPC/SEC et HPLC, respectivement.
Une méthodologie originale des plans d’expériences a été appliquée pour optimiser
la cohésion interne et l’émission de formaldéhyde des panneaux de particules
produits par différentes formulations. Les trois facteurs étudiés ont été: le nombre
d’additions d’urée, le temps entre additions d’urée et le pH de l’étape de
condensation. Les résultats obtenus ont montrés que l’addition séquentiel d’urée
affecte les propriétés des résines UF, ainsi que le pH de condensation, qui est le
facteur qui influence le plus la vitesse de la réaction de condensation. Alors, les
meilleurs résultats concernant la performance globale de la résine ont été obtenus
en combinant les trois facteurs étudiés.
Finalement, des nouveaux procédés pour la synthèse des résines UF ont été
étudiés et mis en œuvre. Les meilleures résines ont été produites selon le procédé
fortement acide. Le comportement lors du durcissement des résines produites au
laboratoire et des résines commerciales a été étudié par deux techniques
récentes : Integrated Pressing and Testing System (IPATES) et Automated Bonding
Evaluation System (ABES). L’information fournie par chaque système de test est
discutée. La résine optimisée obtenue moyennant l’utilisation d’une méthode de
plans d’expériences a présenté toujours la meilleure performance.
xi
TABLE OF CONTENTS
CHAPTER 1
1 INTRODUCTION .................................................................................................................... 3
1.1 Urea-Formaldehyde resins ..................................................................................... 3
1.1.1 Raw materials .................................................................................................................. 7
1.1.2 Chemical reactions ........................................................................................................ 13
1.1.3 Cure ............................................................................................................................... 15
1.1.4 Characterization ............................................................................................................ 16
1.1.5 Storage stability - Ageing ............................................................................................... 18
1.1.6 Formaldehyde emissions ............................................................................................... 19
1.2 Wood-based panels industry ................................................................................ 21
1.2.1 Manufacture of Particleboard ....................................................................................... 22
1.2.2 Market ........................................................................................................................... 23
1.2.3 Environmental Impact ................................................................................................... 25
1.3 Motivation and Outline ........................................................................................ 27
1.4 References ............................................................................................................ 29
CHAPTER 2
2 THE COLLOIDAL NATURE OF UF RESINS AND ITS RELATION WITH ADHESIVE PERFORMANCE ................ 35
2.1 Introduction .......................................................................................................... 36
2.2 Materials and Methods ........................................................................................ 38
2.2.1 Materials ....................................................................................................................... 38
2.2.2 Methods ........................................................................................................................ 39
2.3 Results and discussion .......................................................................................... 42
2.3.1 Disperse phase morphology .......................................................................................... 42
2.3.2 Particle size distribution ................................................................................................ 49
2.3.3 Molecular weight distribution ....................................................................................... 54
2.3.4 Curing behaviour and bonding strength........................................................................ 60
2.4 Conclusions ........................................................................................................... 63
2.5 References ............................................................................................................ 64
CHAPTER 3
3 EFFECT OF AGEING ON UF RESINS .......................................................................................... 69
xii
3.1 Introduction .......................................................................................................... 69
3.2 Materials and Methods ........................................................................................ 72
3.2.1 Resins preparation ........................................................................................................ 72
3.2.2 GPC/SEC analysis ........................................................................................................... 73
3.2.3 HPLC analysis ................................................................................................................. 74
3.3 Results and discussion .......................................................................................... 75
3.3.1 Characterization of UF resins ........................................................................................ 75
3.3.2 Monitoring the ageing of UF resins ............................................................................... 81
3.3.3 Determination of water tolerance ................................................................................ 87
3.4 Conclusions ........................................................................................................... 88
3.5 References ............................................................................................................ 89
CHAPTER 4
4 OPTIMIZATION OF THE SYNTHESIS OF UF RESINS USING RESPONSE SURFACE METHODOLOGY ............. 93
4.1 Introduction .......................................................................................................... 94
4.2 Materials and Methods ........................................................................................ 98
4.2.1 Experimental design ...................................................................................................... 98
4.2.2 Synthesis of UF resins .................................................................................................. 100
4.2.3 GPC/SEC analysis ......................................................................................................... 101
4.2.4 Preparation of laboratory-made particleboards ......................................................... 102
4.2.5 Particleboard Testing .................................................................................................. 102
4.3 Results and Discussion ........................................................................................ 103
4.3.1 Characteristics and performance of the UF resins produced ...................................... 103
4.3.2 Model fitting ................................................................................................................ 108
4.3.3 Effect of all three factors on the measured responses................................................ 110
4.3.4 Optimization of Operating Conditions ........................................................................ 112
4.4 Conclusions ......................................................................................................... 114
4.5 References .......................................................................................................... 115
CHAPTER 5
5 COMPARISON OF UF SYNTHESIS BY ALKALINE-ACID AND STRONGLY ACID PROCESSES ...................... 119
5.1 Introduction ........................................................................................................ 120
5.2 Materials and Methods ...................................................................................... 122
5.2.1 Resin preparation ........................................................................................................ 122
xiii
5.2.2 GPC/SEC analysis ......................................................................................................... 125
5.2.3 HPLC analysis ............................................................................................................... 125
5.3 Results and discussion ........................................................................................ 126
5.3.1 Characteristics of the produced resins ........................................................................ 126
5.3.2 Monitoring of UF synthesis ......................................................................................... 126
5.3.3 Comparison of the two resins ..................................................................................... 133
5.4 Conclusions ......................................................................................................... 136
5.5 References .......................................................................................................... 137
CHAPTER 6
6 EVALUATION OF UF ADHESIVES PERFORMANCE BY IPATES AND ABES MECHANICAL TESTS ............ 141
6.1 Introduction ........................................................................................................ 142
6.2 Materials and Methods ...................................................................................... 143
6.2.1 Raw materials .............................................................................................................. 143
6.2.2 Methods ...................................................................................................................... 145
6.3 Results and discussion ........................................................................................ 148
6.3.1 Integrated Pressing and Testing System (IPATES) ....................................................... 148
6.3.2 Automated Bonding Evaluation System (ABES)........................................................... 152
6.4 Conclusions ......................................................................................................... 158
6.5 References .......................................................................................................... 159
CHAPTER 7
7 GENERAL CONCLUSIONS AND FUTURE WORK ........................................................................ 163
7.1 General Conclusions ........................................................................................... 163
7.2 Future Work ....................................................................................................... 165
xiv
LIST OF FIGURES
Figure 1.1. Most important events in UF resin history (source [14]). ...................................... 6
Figure 1.2. Raw materials for the UF resin. ............................................................................. 7
Figure 1.3. Scheme of UF resin reactor. ................................................................................... 8
Figure 1.4. Model of chemical structure of urea. .................................................................... 8
Figure 1.5. Main countries producing urea in 2007 (adapted from [23]). ............................. 10
Figure 1.6. Model of chemical structure of formaldehyde. ................................................... 10
Figure 1.7. Product Tree for Formaldehyde (source [25]). .................................................... 11
Figure 1.8. Current formaldehyde indoor exposure limits for various countries (source [26]). ............................................................................................................................................... 12
Figure 1.9. UF resin synthesis: (a) formation of monomethylolurea; and (b) condensation reactions of methylolureas to form methylene ether bridges and methylene bridges (adapted from [32])................................................................................................................ 14
Figure 1.10. Example of structure of crosslinked UF resin. ................................................... 15
Figure 1.11. Effect of storage temperature on storage time (viscosity) of UF resin(adapted from [12]). .............................................................................................................................. 18
Figure 1.12. History of wood-based panel industry (source [73]). ........................................ 22
Figure 1.13. Particleboard process diagram (source [75]). .................................................... 23
Figure 1.14. Evolution of the production of wood-based panels in world. ........................... 24
Figure 1.15. Evolution of the production of wood-based panels in Portugal. ....................... 25
Figure 1.16. Cycle of Wood based panels (source [78])......................................................... 27
Figure 1.17. A schematic diagram of linkage between various Chapters present in this thesis. ..................................................................................................................................... 29
Figure 2.1. Optical microscopy image of resin B: a) diluted in water (1.6 wt % resin in water) and b) diluted in 10 wt % urea solution (1.3 wt % resin in urea solution). ............................ 43
Figure 2.2. Simplified representation of a UF resin heterogeneous nature: a) original resin, and b) resin after flocculation by water dilution. .................................................................. 44
Figure 2.3. TEM images of resin B diluted in water. a) fresh resin, and b) aged resin. .......... 45
Figure 2.4. Cryo-SEM image aged resin B. ............................................................................. 46
xv
Figure 2.5. SEM image of UF resin B sprayed and cured on a metal plate. ........................... 47
Figure 2.6. Elementary analysis for UF resin B cured on a metal plate. a) particle surface, and b) in continuous phase. ................................................................................................... 48
Figure 2.7. Particle size distribution for three UF resins stored at 5 °C. a) PSD in volume for 30 days, b) PSD in number for 30 days – note that the curves for samples B and C are nearly superimposed, c) PSD in volume for 170 days, and d) PSD in number for 170 days. ............ 51
Figure 2.8. Particle size distribution in volume for UF resin A-2 and supernatant. a) resin stored for 20 days at 25 °C, and b) resin stored for 50 days at 25 °C. ................................... 53
Figure 2.9. Particle size distribution in volume for UF resin D and supernatant stored for 20 days. ....................................................................................................................................... 53
Figure 2.10. Chromatogram for UF resin A-2 with 20 and 50 days, sediment and supernatant diluted 3 % in DMSO. a) normalized response of RI sensor for 20 days, b) response of RALLS sensor for 20 days, c) normalized response of RI sensor for 50 days, and d) response of RALLS sensor for 50 days. ...................................................................................................... 55
Figure 2.11. Chromatogram for UF resin D with 20 days, sediment and supernatant diluted 3 % in DMSO. a) normalized response of RI sensor, and b) response of RALLS sensor. ........ 56
Figure 2.12. DSC curves for UF resin A-2, supernatant and sediment. Resin had been stored for 50 days.............................................................................................................................. 61
Figure 3.1. Chromatograms for UF-R5 and UF-R2 diluted 3 % in DMSO and stored for 5 days at 25 °C). a) normalized weight fraction (Wt Fr), and b) RALLS response. ............................. 77
Figure 3.2. Chromatogram obtained for resin UF-R5. ........................................................... 78
Figure 3.3. Peak areas normalized by total chromatogram area for UF-R5 and UF-R2 stored at 25 °C. .................................................................................................................................. 79
Figure 3.4. Chromatograms for five UF resins from different manufactures in Europe. a) normalized weight fraction, and b) RALLS response. ............................................................. 80
Figure 3.5. Ratios of peak areas / total area of urea (U), monomethylolurea (MMU) and dimethylolurea (DMU) for five UF resins from different producers. ..................................... 81
Figure 3.6. Change of viscosity of UF-R5 and UF-R2 with storage time at 25 °C. .................. 82
Figure 3.7. Chromatograms for UF-R2 diluted 3 % in DMSO, for different storage periods at 25 °C. a) normalized weight fraction, and b) RALLS response. .............................................. 83
Figure 3.8. Chromatograms for UF-R5 diluted 3 % in DMSO, for different storage periods at 25 °C. a) normalized weight fraction, and b) RALLS response. .............................................. 84
Figure 3.9. Evolution of the ratios of peak areas / total area of the urea (U), monomethylolurea (MMU), dimethylolurea (DMU) and three other oligomeric species for
xvi
UF-R2 stored for various periods at 25 °C. ............................................................................. 86
Figure 3.10. Evolution of the ratios of peak areas / total area of the urea (U), monomethylolurea (MMU), dimethylolurea (DMU) and three other oligomeric species for UF-R5 stored for various periods at 25 °C. ............................................................................. 86
Figure 3.11. Chromatograms for UF-R5 aged for 5 days, diluted in DMSO and very diluted (flocculated) in water. ............................................................................................................ 88
Figure 4.1. Condensation reaction time versus pH of the condensation step for the resins produced. ............................................................................................................................. 103
Figure 4.2. Normalized response of RI sensor for UF resin run 5 diluted 3 % in DMSO. ..... 105
Figure 4.3. Relation between insoluble aggregates and internal bond strength. The dashed lines represent the 90 % confidence intervals. .................................................................... 107
Figure 4.4. Relation between insoluble molecular aggregates and F content. The dashed lines represent the 90 % confidence intervals. .................................................................... 108
Figure 4.5. Experimental and calculated results of the responses considered. Y1 – internal bond strength and Y2 – formaldehyde emission. ................................................................ 109
Figure 4.6. Response surface for internal bond strength as a function of: a) time span between urea additions and number of urea additions (for different pH values of condensation step), b) pH of condensation step and time span between urea additions (for different numbers of urea additions) and c) pH of condensation step and number of urea additions (for different time spans between urea additions). ............................................. 111
Figure 4.7. Response surface for formaldehyde emission as a function of: a) time span between urea additions and number of urea additions (for different pH values of condensation step), b) pH of condensation step and time span between urea additions (for different numbers of urea additions) and c) pH of condensation step and number of urea additions (for different time spans between urea additions). ............................................. 112
Figure 5.1. Reaction temperature (—) and pH (---) histories for resin UF-Exp7. These simplified history curves are based on the experimentally measured values. The urea addition (Ui) and sample collection (Si) times are also indicated in the graph. ................... 123
Figure 5.2. Reaction temperature (—) and pH (---) histories for resin UF-W6. These simplified history curves are based on the experimentally measured values. The urea addition (Ui) and sample collection (Si) times are also indicated in the graph. ................... 125
Figure 5.3. Monitoring of UF-Exp 7 synthesis by GPC/SEC: a) samples collected during methylolation step, including the chromatogram of the final resin, and b) samples during condensation step, including the chromatogram of the final resin. .................................... 127
Figure 5.4. Formation of methylolureas (mono-, di- and trimethylolurea) by the addition of formaldehyde to urea. ......................................................................................................... 128
xvii
Figure 5.5. Condensation of the methylolureas and urea to form methylene-ether bridges and methylene bridges. ....................................................................................................... 129
Figure 5.6. Monitoring of UF-W6 synthesis by GPC/SEC: a) 1ª condensation step and 1ª methylolation step, b) 2ª condensation step, and c) 2ª methylolation step and final resin. ............................................................................................................................................. 131
Figure 5.7. Condensation reactions of urea and formaldehyde to form methylolureas that form methyleneureas and urons. ........................................................................................ 132
Figure 5.8. Chromatogram for four resins (UF-W6 and UF-Exp7) with 5 days (stored at 25 °C). ........................................................................................................................................ 134
Figure 5.9. Ratios of peak areas / total area of the urea (U), monomethylolurea (MMU) and dimethylolurea (DMU) for UF-W6 and UF-Exp7 resins, with 5 days (stored at 25 °C). ........ 136
Figure 6.1. Overview and schematic of the IPATES machine. .............................................. 145
Figure 6.2. Particleboard mat during the pressing and subsequent testing. ....................... 146
Figure 6.3. Core layer temperature of the board at different pressing temperatures. ....... 147
Figure 6.4. Schematic of the ABES test procedure. ............................................................. 148
Figure 6.5. Bond strength curves for particle mats pressed at 130 °C and 160 °C for UF-W6 and UF-Exp7 (8 % resin and 3 % hardener). a) 130 °C, and b) 160 °C. ................................. 150
Figure 6.6. Bond strength curves for particle mats pressed at 130 °C and 160 °C for UF-R8 and UF-R2 (8 % resin and 3 % hardener). a) 130 °C; b) 160 °C. ............................................ 151
Figure 6.7. Bond strength curves for particle mats at 130 ºC with 8 % (•) and 10 % (○) of adhesive (3 % hardener). a) UF-W6 resin, b) UF-Exp7 resin, c) UF-R8 resin, and d) UF-R2 resin. .................................................................................................................................... 152
Figure 6.8. Shear strength evolution with time for UF-W6 resin at pressing temperature of the 80 °C (•) 90 °C (○), 100 °C (▼), 110 °C (∆) and 130 °C (■). ............................................ 153
Figure 6.9. Shear strength evolution with time of four resins at pressing temperature 80 °C (•) 90 °C (○), 100 °C (▼), 110 °C (∆) and 130 °C (■). a) UF-W6 resin; b) UF-Exp7 resin; c) UF-R8 resin and d) UF-R2 resin. ................................................................................................. 154
Figure 6.10. ABES-derived Arrhenius plots for four adhesive systems. ............................... 155
Figure 6.11. Shear strength evolution with time UF-W6 resin with two hardener ratios, 1.5 % (•) and 3.0 % (○), at 110 °C............................................................................................... 156
Figure 6.12. Breaking load in shear mode for four resins by the ABES method at 100 ºC and 80 s (ten tests were performed for each resin to evaluate the error). ................................ 158
xviii
LIST OF TABLES
Table 1.1. Advantages and disadvantages of the main adhesives used in the production of wood based panels (adapted from [2]) .................................................................................... 4
Table 1.2. Limits of formaldehyde emissions for particleboard (adapted from [71]) ........... 21
Table 2.1. Technical data for UF-resins analysed in this work ............................................... 39
Table 2.2. Solids content of original resins, sediment and supernatant phases for UF resins A-2 and D ................................................................................................................................ 52
Table 2.3. Values of Mn, Mw, polydispersity, and parameters f1 and f2, obtained by SEC for resin, sediment and supernatant of the UF resin A-2 with 20 and 50 days and UF resin D with 20 days storage at 25 °C ................................................................................................. 57
Table 2.4. Cure temperatures and enthalpies for UF resin A-2, supernatant and sediment 61
Table 2.5. Values of tensile shear strength and % of cohesive failure within wood for UF resin A-2, supernatant and sediment .................................................................................... 62
Table 3.1. Technical data of UF-resins ................................................................................... 73
Table 3.2. Technical data on UF-resins used from different producers................................. 73
Table 3.3. Values of Mn, Mw, polydispersity (Mw/Mn), and parameters f1 and f2, obtained by SEC for UF-R5 and UF-R2 stored for 5 days at 25 °C ......................................................... 77
Table 3.4. Values of Mn, Mw, polydispersity (Mw/Mn), and parameters f1 and f2, obtained by GPC/SEC for UF-R2 and UF-R5 stored for different days at 25 °C ..................................... 85
Table 4.1. Current classification of formaldehyde by some organizations ............................ 94
Table 4.2. Experimental levels of the three factors ............................................................... 98
Table 4.3. Central composite design matrix of experiments generated by the DoE tool ...... 99
Table 4.4. Characteristics of UF resins produced ................................................................. 104
Table 4.5. Experimental results for the three measured responses .................................... 106
Table 4.6. Operating conditions that produce the minimum formaldehyde emission ....... 113
Table 4.7. Characteristics of the UF resin optimized for minimizing the formaldehyde emission ............................................................................................................................... 113
Table 4.8. Predicted and experimental values for the three responses .............................. 114
Table 5.1. Technical data of UF-resins ................................................................................. 126
xix
Table 5.2. Identification of different stages during the synthesis (UF-Exp 7) ...................... 127
Table 5.3. Values of Mn, Mw, polydispersity (Mw/Mn) obtained by GPC/SEC for samples collected during the synthesis of UF-Exp7 ........................................................................... 129
Table 5.4. Identification of different stages during the synthesis ....................................... 130
Table 5.5. Values of Mn, Mw, polydispersity (Mw/Mn) obtained by GPC/SEC for samples collected during the synthesis of UF-W6 ............................................................................. 133
Table 5.6. Values of Mn, Mw and parameters f1 and f2, obtained by GPC/SEC for resins (UF-W6 and UF-Exp7) ................................................................................................................. 135
Table 6.1. Technical data of UF-resins ................................................................................. 144
Table 6.2. Parameters evaluated by IPATES ........................................................................ 146
Table 6.3. Parameters evaluated on ABES ........................................................................... 148
Table 6.4. Values of the activation energy (Ea) for each adhesive system .......................... 156
CHAPTER 1
Introduction
3
1 Introduction
1.1 Urea-Formaldehyde resins
Urea-formaldehyde (UF) polymers have been for decades the most widely used
adhesives in the manufacture of wood-based panels, such as particleboard and
medium density fiberboard (MDF) (both consuming 68 % of the world’s UF resins
production) and plywood (consuming 23 %) [1]. Melamine-urea-formaldehyde
(MUF), phenol-formaldehyde (PF) and polymeric 4, 4’-diphenylmethane
diisocyanate (pMDI) polymers are also used in production of wood-based panels,
but their use in particleboard and MDF is relatively small when compared with UF
resins. pMDI is used mostly in the manufacture of moisture resistant panels. Table
1.1 summarizes the advantages and disadvantages of the four main wood
adhesives concerning a few significant parameters [2].
The properties of wood-based panels depend on three factors: the wood species
and origins, particularly the interface between the wood surface and the adhesive;
the adhesive; and operating conditions and process of the production of boards
[3].The adhesive is the most expensive component in the panel cost structure.
According with SRI Consulting [1] the global production of UF resins in 2008 was
approximately 14 million ton. Their consumption increased 2.8 % in 2008, and is
expected to grow an average 3.2 % per year from 2008 to 2013, and just under 2 %
per year from 2013 to 2018.
Chapter 1
4
Table 1.1. Advantages and disadvantages of the main adhesives used in the production of wood based panels (adapted from [2])
Proprieties Adhesive
UF MUF PF pMDI
Price Low Medium to
high Medium High
Cure temperature Low Medium High Low
Press time Short Medium Medium to
long Medium
Susceptibility against wood species
High Medium Low Low
Efficiency Low Medium to
high Medium to
high High
Manipulation Easy Easy Easy Difficult
Resistant against hydrolyzed
No Medium to
high High High
Use in wet conditions No Partly yes Yes Yes
Formaldehyde emission
E1, Carb I possible
E1, Carb II possible
Very low emission
No
The main reasons for the wide utilization of UF resin in wood based panels are their
high reactivity, low cost and excellent adhesion to wood [4-5]. One the other hand,
the most important drawbacks are the low moisture resistance and the
formaldehyde emission during panel manufacture and service life [4-5]. Although
the free formaldehyde content on these resins has been decreased during the last
decades, the recent reclassification of formaldehyde by International Agency for
Research on Cancer (IARC) as “carcinogenic to humans”, is forcing resin producers
to develop systems that lead to a decrease in its emissions to levels as low as the
present in natural wood [6-7]. This imposition has been a driving force for a
considerable research effort, not only in the engineering of UF resins, but also in
the development of all sort of alternative resins. In 2007, Dynea AS Company
started the commercialization of AsWood™ resin (formaldehyde based resin)
reaching a formaldehyde emission level in wood based panel products at the same
level as found in solid wood. However, the price of this resin is too high for
Introduction
5
producing standard particleboard and MDF [8].
Until now, the decrease on free formaldehyde emissions has been obtained by
decreasing the molar ratio F/U and/or by the addition of formaldehyde scavengers
(UF resin with an ultra-low molar ratio F/U well below 1.0, MUF scavenger resin,
solid urea, urea solution, starch, lignin, tannin, rice husk, etc.). Both lead to a
decrease on reactivity and degree of curing, harming the formation of adhesive
bonds. Moreover, currently used hardeners are adapted to high F/U molar ratios
and high levels of free formaldehyde in solution. Therefore, the decrease in F/U
molar ratio can result in panels with low internal bond strength. The experience of
wood-based panel producers is that resins with lower molar ratio F/U have lower
flexibility in the control of the process variables. This is a very important factor,
since producers are nowadays using mixtures made of recycled wood and wood
from different origins.
Dynea AS Company, one of the most important wood adhesives producers in the
world, claims that, currently, the key topics in the development and production of
adhesives are: cost reduction with effective bonding solutions; broadening the
operating window in terms of production processes; reduction of the formaldehyde
emissions; improvement of adhesive behavior in humid conditions; and finding
alternative raw materials, namely bio-based adhesives [3].
History
The first published studies about the reaction between urea and formaldehyde
were the works by Tollens [9] in 1884. However, Goldschmidt authored the
currently most cited work in the open literature, in 1896 [5, 10]. This researcher
described the formation of precipitates from solution due to the reaction of urea
and formaldehyde under acidic conditions. Later, Carlson [11] related the primary
precipitate referred by Goldschmidt [10] with cyclic structures (now called urons)
and empirically identified it as C5H10O3N4.
Chapter 1
6
In 1931 IG-Farbenindustrie (now BASF)
production of UF resins for wood based industry
development of wood based panels industry, namely particleboard and MDF
forced the growth of UF resins industries
Still later, formaldehyde was designated in 1978
giving rise to a major change in the synthesis process. The reduction of
formaldehyde emissions in the wood based panel was achieved with a large
decrease on molar ratio F/U. This change had a great impact on
process (higher press times and higher press temperatures) and
properties (lower bonding strength and lower moisture resistance
based panels. The decreasing of
formaldehyde emission, but increases
absorption (WA), as well as leads to a decrease
(internal bond strength (IB) and modulus
led to an 80 % decrease in formaldehyde
methods to measure the formalde
developed and the first rules and regulation were established.
Figure 1.1. Most important events in UF resin history
Farbenindustrie (now BASF) Company in Germany started the industrial
production of UF resins for wood based industry [5]. After Second World War, the
development of wood based panels industry, namely particleboard and MDF,
growth of UF resins industries [5, 12].
formaldehyde was designated in 1978 as a possible carcinogenic agent,
change in the synthesis process. The reduction of
the wood based panel was achieved with a large
io F/U. This change had a great impact on the manufacturing
and higher press temperatures) and on the physical
gth and lower moisture resistance) of the wood-
of F/U molar ratio leads to a decrease in
aldehyde emission, but increases the thickness swelling (TS) and water
as well as leads to a decrease in mechanical performance
(internal bond strength (IB) and modulus of rupture (MOR)) [2, 13]. This strategy
ormaldehyde emissions. In addition, new testing
methods to measure the formaldehyde emissions from wood-panels were
developed and the first rules and regulation were established.
Most important events in UF resin history (source [14]).
In the 1980s, novel synthesis processes were developed by several com
15-17] and the number of patents on UF resins has grown rapidly.
1.1.1 Raw materials
As the name indicates, the two monomers used in manufacture of UF resins are
urea and formaldehyde. The two compounds are derived from natural gas (
Figure 1.2)[18-19].
Figure 1.2. Raw materials for the UF resi
Typically the production is carried out with a batch reactor. These reactors have
normally volumes about 20 to 40 m
and heating and cooling systems to control the temperature. The reactors have
also a continuous addition of raw materials, namely urea, formaldehyde, acid, base,
and others additives. Figure 1.3 presents
manufacture of UF resins.
Introduction
7
In the 1980s, novel synthesis processes were developed by several companies [11,
and the number of patents on UF resins has grown rapidly.
s the name indicates, the two monomers used in manufacture of UF resins are
urea and formaldehyde. The two compounds are derived from natural gas (see
the UF resin.
Typically the production is carried out with a batch reactor. These reactors have
normally volumes about 20 to 40 m3 and are equipped with a mechanical stirrer
and heating and cooling systems to control the temperature. The reactors have
ous addition of raw materials, namely urea, formaldehyde, acid, base,
presents a scheme of a typical reactor for the
Chapter 1
8
Figure 1.3. Scheme of UF resin reactor.
Urea
Urea is an organic compound with the chemical
The synthesis of this organic compound by heat
discovered by Friedrich Wöhler in
obtained from inorganic chemicals.
soluble in water.
The main applications of urea are in agriculture (used in
source of nitrogen) and chemical industry (used in the production of amino resins)
[19, 21].
Figure 1.4. Model of chemical structure of urea.
Scheme of UF resin reactor.
Urea is an organic compound with the chemical formula (NH2)2CO (see Figure 1.4).
The synthesis of this organic compound by heating ammonium cyanate was
in 1828 [20-21]. It is the first organic compound
obtained from inorganic chemicals. Urea is a solid, colorless, odorless, and highly
are in agriculture (used in fertilizers as a convenient
source of nitrogen) and chemical industry (used in the production of amino resins)
Model of chemical structure of urea.
Introduction
9
Production and manufacturing process of urea
Urea was first produced industrially by the hydration of calcium cyanamide but the
easy availability of ammonia led to the development of ammonia/carbon dioxide
technology. This is a two step process where the ammonia and carbon dioxide
react to form ammonium carbamate, which is then dehydrated to urea [19, 21-22].
4223COONHNHCO2NH −→+ (1.1)
OHCO)(NHCOONHNH 22242 +→− (1.2)
In the process, ammonia and carbon dioxide are fed to the synthesis reactor which
operates around 180-210 °C and 150 bar. The reaction mixture containing
ammonia, ammonium carbamate and urea is first stripped of the ammonia and the
resultant solution passes through a number of decomposers operating at
progressively reduced pressures. Here, the unconverted carbamate is decomposed
back to ammonia and carbon dioxide and recycled to the reactor. The urea solution
is concentrated by evaporation or crystallization, and the crystals can be melted to
yield pure urea in the form of pills or granules. Pills are made by spraying molten
urea from the top of a high tower through a counter current air stream. Granular
urea is formed by spraying molten urea into a mixture of dried urea particles and
fines in a rotating drum [19, 22].
Urea is produced worldwide at large scale. The world production in 2007 was
approximately 144 million ton [23]. Figure 1.5 shows the production of the main
countries producers of urea. China and Russia produce more than 50 % of urea.
Chapter 1
10
Figure 1.5. Main countries producing urea in 2007 (adapted from
Formaldehyde
Formaldehyde is one of the most abundant organ
with the chemical formula HCHO (see
Hofmann, with Alexander Butlerov in 1867
isolated and purified until 1892. This was achieved by Friedrich Von Stradonitz, who
also introduced the concept of chemical bonds. Formaldehyde is also called
methanal and is formed by oxidizing methanol.
original state because it has a short half
easily dissolves, hydrates and oligomerizes
transported commercially.
Figure 1.6. Model of chemical structure of formaldehyde.
0
10
20
30
40
50
60
China Russia
10
6x
Ton
urea in 2007 (adapted from [23]).
Formaldehyde is one of the most abundant organic compounds in the Universe,
HCHO (see Figure 1.6). It was discovered by Wilhelm Von
with Alexander Butlerov in 1867 [18, 22, 24]. Despite this, it was not
isolated and purified until 1892. This was achieved by Friedrich Von Stradonitz, who
also introduced the concept of chemical bonds. Formaldehyde is also called
ing methanol. Formaldehyde is rarely found in its
original state because it has a short half-life in air. In water, it is very unstable and it
, hydrates and oligomerizes in water, which is the form in which is
Model of chemical structure of formaldehyde.
India USA Indonesia Trinidad
Formaldehyde is an important chemical for the global economy, widely used in t
production of thermosetting resin
material in the synthesis of the
methylenebis (4-phenyl isocyanate) or MDI, and pentaerythritol), and
preservation and disinfection [25-26]
million ton. Figure 1.7 shows the consumption of formaldehyde and the
for several end uses. The substitution
would suffer large losses in performance
capital investments would be required to produce or utilize the substitutes. Data of
Formaldehyde Council Inc. show the great impact of formaldehyde industry to the
U.S. and Canadian economies [25]. In 2003 the value of sales of formaldehyde and
derivative products achieved approximately $145 billion
workers is 4.2 million, which represents nearly 3.4 % of employment in private,
nonfarm establishments in North America
Figure 1.7. Product Tree for Formaldehyde
Introduction
11
Formaldehyde is an important chemical for the global economy, widely used in the
resins (UF, MUF, MF, PF), as an intermediate raw
s of the chemicals (polyacetal resins, 1,4-butanediol,
phenyl isocyanate) or MDI, and pentaerythritol), and for
26]. The annual world production is about 21
the consumption of formaldehyde and the percentage
ubstitution of these products is very difficult; consumers
performance using alternative materials, and new
investments would be required to produce or utilize the substitutes. Data of
Formaldehyde Council Inc. show the great impact of formaldehyde industry to the
. In 2003 the value of sales of formaldehyde and
derivative products achieved approximately $145 billion [25]. The number of
workers is 4.2 million, which represents nearly 3.4 % of employment in private,
in North America [25].
Product Tree for Formaldehyde (source [25]).
Chapter 1
12
It is highly toxic, and its exposure affects eye
Organization (WHO) fixed 0.1 mg/m
concentration. However, some countries adopted national regulations with higher
limits (see Figure 1.8). In 2006 the
carcinogen that causes nasopharyngeal cancer and probably leukemia
Figure 1.8. Current formaldehyde indoor exposure limits for various countries
Production and Manufacturing Process of the Formaldehyde
Industrially, formaldehyde is produced by the catalytic oxidation of methanol. The
most commonly used catalysts are silver metal or a mixture of an iron oxide with
molybdenum and vanadium [18, 24]
used in the world. The methanol and oxygen react at 250
formaldehyde according to the chemical
2H2HCHOOOH2CH 23 +→+
xposure affects eyes, nose and throat. The World Health
WHO) fixed 0.1 mg/m3 as a limit for formaldehyde indoor-air
. However, some countries adopted national regulations with higher
In 2006 the IARC classified formaldehyde as a human
carcinogen that causes nasopharyngeal cancer and probably leukemia [27].
Current formaldehyde indoor exposure limits for various countries (source [26]).
Production and Manufacturing Process of the Formaldehyde
duced by the catalytic oxidation of methanol. The
most commonly used catalysts are silver metal or a mixture of an iron oxide with
[18, 24]. Formox process, using iron oxide is the more
anol and oxygen react at 250 °C – 400 °C to produce
formaldehyde according to the chemical equation [24]:
O2H2 (1.3)
Introduction
13
The silver-based catalyst is usually operated at a higher temperature, about 650 °C.
On it, two chemical reactions simultaneously produce formaldehyde: the one
shown above, and the dehydrogenation reaction [24]:
23 HHCHOOHCH +→ (1.4)
Further oxidation of formaldehyde during its production usually yields formic acid,
which is found in formaldehyde commercial solutions.
1.1.2 Chemical reactions
In conventional production, UF resin synthesis has been established as a two step
process, involving methylolation and condensation reactions, respectively [4, 28-
29]. The methylolation reaction is an addition reaction and it is performed in
neutral or slightly alkaline medium. Urea reacts with formaldehyde to form
methylolureas (monomethylolurea, dimethylolurea, trimethylolurea and
tetramethylolurea) (see Figure 1.9). Tetramethylolurea could not be observed
experimentally [30]. In the condensation reaction, the system is acidified to achieve
the condensation of methylolureas by the reactions between methylol groups and
their primary and secondary amides. The growth of polymer is obtained by the
formation of methylene ether bridges (CH2-O-CH2) and/or methylene bridges (-CH2-
). In Figure 1.9, a simplified scheme of the reactions that occurred in the UF resin
synthesis is presented [31].
Chapter 1
14
Figure 1.9. UF resin synthesis: (a) formation of monomethylolurea; and (b) condensation reactions of methylolureas to form methylene ether bridges and methylene bridges (adapted from [32]).
During manufacture, progress of synthesis reaction is followed by viscosity
measurement; the reactions proceed until the desired viscosity is reached. At this
point, the reactions are blocked by neutralization and cooling, resulting in a
complex mixture of molecules with different sizes and different condensation
degrees [4].
The most important parameters in the synthesis of UF resins are: purity of raw
materials (formaldehyde and urea); F/U molar ratio; and production process (i.e.,
pH program, temperature program, kind and quantities of alkaline and acid
catalysts, sequential addition of raw materials and duration of the each step).
1.1.3 Cure
During UF resin synthesis, the polymer condensation is stopped by neutralization
and cooling. In order to reactivate it and complete the crosslinking process, it is
needed to add an acid catalyst and increase the temperature (up to 80
After curing, UF resins become an insoluble, thermoset, three
(see Figure 1.10).
Figure 1.10. Example of structure of crosslinked UF resin.
In general, the acid conditions are adjusted by the addition of latent hardeners
(ammonium chloride, ammonium sul
ammonium nitrate, the latent hardener
in the resin to generate nitric acid (Eq. 1.5), which
4NHO6HCHONO4NH 34 ↔+
The production of low formaldehyde UF resins, with low levels of free
formaldehyde originated a decrease
latent hardeners were originally selected and optimized
high levels of free formaldehyde. So, latent hardeners are insufficient to promote
the acid medium favorable to the complete cure of UF resins
The curing of UF resins can also be
Introduction
15
synthesis, the polymer condensation is stopped by neutralization
and cooling. In order to reactivate it and complete the crosslinking process, it is
needed to add an acid catalyst and increase the temperature (up to 80 °C – 120 ºC).
an insoluble, thermoset, three-dimensional network
Example of structure of crosslinked UF resin.
In general, the acid conditions are adjusted by the addition of latent hardeners
ammonium sulfate and ammonium nitrate). In the case of
ammonium nitrate, the latent hardener reacts with the free formaldehyde present
acid (Eq. 1.5), which decreases the pH [4]:
O6HN)(CH4NHO 24623 ++ (1.5)
The production of low formaldehyde UF resins, with low levels of free
formaldehyde originated a decrease in the performance of latent hardeners. The
selected and optimized to be used with resins with
maldehyde. So, latent hardeners are insufficient to promote
complete cure of UF resins [33].
be catalyzed by the direct addition of acids (maleic,
Chapter 1
16
acetic, oxalic, formic, hydrochloric, nitric, and phosphoric, and others) [12, 29, 33-
34]. However, these hardeners originate corrosion problems in the equipments,
wood degradation, and reduce considerably the pot life of the resin (stability time
of the catalyzed resin) [35].
Several works [36-38] reported that the thermal curing reactivity of UF resins is
significantly affected by the molar ratio F/U. The decrease of molar ratio F/U
increases the gel time and cure temperature. Siimer et al. [36] reported also that
UF resins with higher amount of reactive methylolureas groups present a higher
reactivity.
1.1.4 Characterization
The main problem of the analysis of UF resin is related to the fact that these resins
are a very complex mixture with: 1) reversible reactions, namely the methylolation
reaction (release of formaldehyde), 2) structural rearrangements, 3) different types
of linkages (methylene-methylene ether), and 4) different monomers
(methylolation degree: monomethylolurea, dimethylolurea and trimethylolurea)
[39-43].
To characterize the liquid UF resin, several techniques have been used, including
13C-NMR (Carbon Nuclear Magnetic Resonance) [44-45] and FTIR (Fourier
Transform Infrared) [46-47] to investigate the structure of the resin, and GPC/SEC
(Gel permeation Chromatography/Size Exclusion Chromatography) [48-49] to
determine the molecular weight (MW) and molecular weight distribution (MWD).
More recently, the capabilities of FT-NIR spectroscopy (Fourier Transform Near-
Infrared) have been exploited. This technique has demonstrated to be useful for
on-line monitoring the consumption of NH2 groups during the early stages of resin
synthesis [50-51].
The study of resin curing is more difficult. The insolubility of the cured product is a
restriction to the use of some techniques. Recently, several techniques such as
Introduction
17
solid-state C13-NMR [52-53] FTIR [46] allowed a better understanding of the acid
cure at high temperature of UF resin. The chemical cure reaction can be monitored
by DSC (Differential Scanning Calorimetry) [37, 54] allowed to estimate the degree
of chemical cure as well as the heat of cure reaction. The healing mechanics can be
monitored by TMA (Thermal Mechanical Analysis) [55-56], DMA (Dynamic
Mechanical Analysis) [57] or ABES (Automatic bonding Evaluation System) [29, 58].
Raman spectroscopy was used by Hill et al. [59] to assess the structure of cured UF
resin and by Carvalho et al. [60-61] to study the cure reaction. This technique
allows the analysis of the liquid resin, cured resin or the cured resin in wood based
panels. The Raman spectra of several model compounds have permitted to clarify
the presence of certain characteristic groups and their quantitative determination.
It was possible to observe the relative growth of the concentration of methylol and
methylene groups. Raman spectroscopy was found to be very interesting for the
study of the resin cure and permitted to obtain kinetic data as the basis for a
simple empirical model, considering a homogeneous irreversible reaction of a
single kind of methylol group and ureas with rate constants depending on their
degree of substitution [61]. However, the existing fluorescence on the spectra of
cured resin may affect the band relative intensities and therefore the quantitative
results. An internal standard, if provided, could enable more precise results.
Another technique, very recent, which has not been used for the analysis of UF
resin, is MALDI-TOF (Matrix Assisted Laser-desorption Ionization Time of Flight)
mass spectroscopy, which can be used to investigate the distribution of molecular
weights. In this technique the polymer is dispersed in a matrix - consisting of an UV
absorber - and then bombarded by a laser. The absorbed energy is able to vaporize
some of the molecules between two high voltage electrodes. The electric field
between the electrodes accelerates the molecules, which will hit the detector with
an acceleration inversely proportional to molecular weight. MALDI-TOF was used to
analyze melamine-urea-formaldehyde (MUF) resins by Zanetti et al. [62].
Chapter 1
18
1.1.5 Storage stability - Ageing
During storage, uncured urea/formaldehyde resins undergo reactions that result in
structural changes. Methylene groups adjacent to secondary amino groups are
formed by condensation. This reaction proceeds during storage between the free
terminal hydroxymethyl and amino groups.
Physical processes also take place during ageing, with the formation of colloidal
particles followed by clustering, especially in the UF, melamine-formaldehyde (MF)
and MUF [62-64].
During the storage time, that is about 30 days the UF resins must be maintained
constant the viscosity [12, 65]. Figure 1.11 shows the evolution of viscosity of UF
resin during the storage time for two temperatures [12]. It is evident that for higher
temperature the storage time is lower. The problem of storage stability is very
important in Iberian Peninsula due to the high temperature prevailing in summer.
Figure 1.11. Effect of storage temperature on storage time (viscosity) of UF resin(adapted from [12]).
0
500
1000
1500
2000
0 5 10 15 20 25 30
visc
osi
ty /
cP
s
Time / days
24 ℃
32 ℃
Introduction
19
1.1.6 Formaldehyde emissions
The issue of formaldehyde emission from wood based panels is related to the use
of UF resins as adhesives. The emission is originated from: unreacted free
formaldehyde, formaldehyde in the form of methyIenegIycoI, retained in the
panels moisture, and formaldehyde released from the slow hydrolysis of methylol
groups and methylene ether linkages present in UF resins [29, 66]. The emission
level depends of different factors, namely the resin, the species and origins of the
wood, the pressing time, the press temperature and the moisture content of wood
before and after pressing [66].
In 1979, the Chemical Industry Institute of Toxicology reported that formaldehyde
caused nasal cancer in rats exposed at high levels during longs time. IARC in 2006
concluded from several case-studies that there is sufficient evidence for the
carcinogenicity of formaldehyde in humans, namely nasal cancer and leukemia
[27]. FormaCare, which represents key European producers of formaldehyde,
aminoplast glues and polyols, disagrees with this conclusion and stated that the
“weight of scientific evidence does not support such a determination” [67]. In
2005, it started an independent research in order to provide a solid database about
the effect of formaldehyde in humans [6, 67]. In 2007, FormaCare organized in
Barcelona, Spain the International Formaldehyde Science Conference in order to
discuss the new reclassification of formaldehyde by IARC, as well as gather all
newly available scientific research and study results about formaldehyde [6, 67].
The bottom line of this conference was that the common use of formaldehyde in
consumer products and other applications does not pose a risk to human health
[67].
The determination of formaldehyde emission can be done according to several
methods. In Europe, four test methods have been approved as European
standards: EN 120 (perforator method), EN 717-1 (chamber method), EN 717-2 (gas
analysis method), and EN 717-3 (flask method) [6].
Chapter 1
20
The perforator method is the most common and is used in particleboard and MDF
industrial plants [6, 28-29, 66, 68-69]. In this process the formaldehyde is extracted
from test pieces by means of boiling toluene and then is absorbed by water. The
formaldehyde content of this aqueous solution is determined by photometrical
detection or fluorescence spectroscopy and expressed in weight (mg) per 100 g of
oven dried board. The perforator values apply to boards with moisture contents of
6.5 %. In the case of boards with different moisture content (in the range of 3 % ≤ H
≤ 10 %) the perforator value shall be multiplied by a factor F = − 0,133 H + 1,86. The
value of formaldehyde content obtained can be used to estimate the actual
formaldehyde emission using correlations available in open literature. The mostly
used nowadays was developed by Risholm-Sundman et al. [70] to convert the EN
120 test values (mg HCHO / 100 g) to EN 717-1 values (mg HCHO / m3). The
disadvantages of the perforator method are the use of toluene and the low
efficiency/reproducibility for very low formaldehyde emissions. In the future, the
perforator method should be replaced by the gas analysis method due to the
shorter analysis times and simpler procedures. Since 2009, gas analysis method is
an approved alternative, small-scale, quality control test accepted by California Air
Resources Board (CARB).
In recent years, national regulations for formaldehyde were established and/or
reformulated in some countries limiting the formaldehyde emission from wood-
based panels. Table 1.2 presents the limits of formaldehyde emissions for
particleboard applied in Japan, Europe and EUA according to different regulations
and the equivalent values for different test methods [71-72]. The test methods
used as reference in each national regulation are different and require the use of
correlations for establishing relationships between the existents regulations. The
major customers of wood based panels, such as IKEA, have required the “E0.5”
regulation in 2009, which is significantly below European requirement [71].
Introduction
21
Table 1.2. Limits of formaldehyde emissions for particleboard (adapted from [71])
Method Japan Europe IKEA EUA
F*** F**** E1 E0.5 CARB P1 CARB P2
EN 120 (mg / 100 g o.d board)
≤ 4.51
≤ 2.71
≤ 8.0 4.0
max 11.3
1
max 5.6
1
max
EN 717-1 (mg / m
3 air)
≤ 0.0541
≤ 0.0541
≤ 0.124 ≤ 0.050
max 0.176
1
max 0.088
1
max
ASTM E1333 (ppm)
≤ 0.0551
≤ 0.0351
≤ 0.1271 0.051
1
max 0.180 max
0.090 max
JIS A 1460 (mg / L)
≤ 0.5 ≤ 0.3
≤ 0.91 0.4
1
max 1.3
1
max 0.6
1
max 1The values are estimating using correlations
1.2 Wood-based panels industry
Wood based panels are manufactured from wood materials having various
geometries (e.g., fibers, particles, strands, flakes, veneers, and lumber) combined
with an adhesive and bonded in a press. The press applies heat (if needed) and
pressure to activate (cross-link) the adhesive resin and bond the wood material into
a solid panel, lumber, or beam having good mechanical and physical properties
(strength, stiffness, form, dimensional stability, etc.
The most used wood-based panels are plywood, particleboard, MDF and oriented
strand board (OSB). Plywood, made by gluing together several hardwood veneers
or plies, was the first type of wood-based panel produced in the world. Only 60
years later particleboard panels were produced. Figure 1.12 shows the main
milestones in the development of the wood-based panels industry.
Chapter 1
22
Figure 1.12. History of wood-based panel industry
1.2.1 Manufacture of Particleboard
Particleboard is manufactured from
recycled woodchips [74]. Typically, it is made in three laye
layers consist of finer particles and sawdust, while the core layer is made of coarser
material.
The manufacture of particleboard has five main steps: (1) furnish preparation, (2)
resin application, (3) mat formation, (4) hot pressing,
is prepared by refining the raw materials into small particles and drying them to
achieve a desired moisture content, about 2 to 7 %
particleboard depends of the characteristics desired, but normally UF resin is used.
The resin/wood ratio, based on resin
usually 6 to 9 % [4, 74]. Additives to enhance characteristics
moisture resistance can be applied at this stage.
particles and the adhesive system, the
forming system and is then hot-pressed under pressures between two and thr
MPa and temperatures between 140 °C and 220 °C
complete, the panel is transported
sawing into finished panel sizes and sanding.
particleboard is represented in Figure
based panel industry (source [73]).
Manufacture of Particleboard
om wood chips, sawdust, waste materials and
. Typically, it is made in three layers. The two external
layers consist of finer particles and sawdust, while the core layer is made of coarser
The manufacture of particleboard has five main steps: (1) furnish preparation, (2)
resin application, (3) mat formation, (4) hot pressing, and (5) finishing. The furnish
prepared by refining the raw materials into small particles and drying them to
achieve a desired moisture content, about 2 to 7 % [74]. The type of resin used in
particleboard depends of the characteristics desired, but normally UF resin is used.
based on resin dry solids content, and particle dry weight , is
Additives to enhance characteristics like fire retardancy or
moisture resistance can be applied at this stage. After mechanically mixing the
, the material goes through a continuous mat-
pressed under pressures between two and three
MPa and temperatures between 140 °C and 220 °C [4, 74]. After the press cycle is
complete, the panel is transported to a board cooler, and then hot-stacked to wait
sawing into finished panel sizes and sanding. The typical production process of
Figure 1.13.
Figure 1.13. Particleboard process diagram (
1.2.2 Market
The increase of the world demand for wood
of the tree role in the global ecosystem are driving the use of recycled wood and
wood from different sources/species in the formulation of wood composites
The variability of available wood
compatibility/adequacy of the resin (bi
resin can be appropriate to glue a certain wood species, but not appropriated for
others [3].
Food and Agriculture Organization of the United Nations (FAO) reported that in
2007, approximately 106 million m
China manufacture 20 %, 10 %, 9 % and 8 % respectively), 56 million m
(China, Germany and EUA manufa
Introduction
23
Particleboard process diagram (source [75]).
The increase of the world demand for wood-based composites and the awareness
global ecosystem are driving the use of recycled wood and
wood from different sources/species in the formulation of wood composites [60].
available wood creates difficulties concerning the
compatibility/adequacy of the resin (binding agent) with the wood [3]. One UF
a certain wood species, but not appropriated for
Agriculture Organization of the United Nations (FAO) reported that in
approximately 106 million m3 of particleboard (EUA, Germany, Canada and
hina manufacture 20 %, 10 %, 9 % and 8 % respectively), 56 million m3 of MDF
(China, Germany and EUA manufacture 45 %, 8 % and 6 % respectively) and 76
Chapter 1
24
million m3 of plywood (EUA, Germany, Canada and china manufacture 20 %, 10 %,
9 % and 8 % respectively) were manufactured in the world (see Figure 1.14) [76]).
Figure 1.14. Evolution of the production of wood-based panels in world.
European Panel Federation (EPF) reported that the wood-based panels industry
was affected by the economic crisis in 2008 [77], in particular the production of
particleboard and MDF, that decreased in 2008 by, 8.7 % and 8 % respectively [77].
Production of wood-based panels in Portugal has been approximately stable in the
last decade. In 2007, the production volume was 850 000 m3 for particleboard, 330
000 m3 for MDF and 21 000 m3 for plywood (see Figure 1.15) [76]. The production
of plywood in Portugal is small due to the lack of raw material (veneers must be
produced from large diameter logs).
0
20
40
60
80
100
120
1961 1965 1970 1975 1980 1985 1990 1995 2000 2005
10
6x
m3
Particleboard
Plywood
MDF
Introduction
25
Figure 1.15. Evolution of the production of wood-based panels in Portugal.
1.2.3 Environmental Impact
The European woodworking industry stands for about 100 000 companies, two
million employees and an annual turnover of 150 billion € [78]. Furthermore,
forests and forest-based industries provide direct employment to three million
people throughout the EU, especially in remote areas [78]. They represent 10 % of
the total production value of the EU manufacturing industry [78]. According to
European woodworking industry, these businesses invest continuously in
sustainable forest management, deflorestation and reforestation activities to
ensure reliable wood availability.
Wood plays a major role in fighting climate changes. This was the conclusion of a
multi-disciplinary working group of the European Commission. The better use of
wood sources stimulates forest expansion and reduces greenhouse gas emissions.
The recycling process has a great paper in future of wood-based panels industry. In
2004 the proportion of recycled wood used in manufacturing of particleboard was
0
100
200
300
400
500
600
700
800
900
1000
10
6x
m3
Particle Board
Plywood
MDF
Chapter 1
26
23 % [79]. Roffael et al. in 2009 [80] studied the use of recycled fiberboards with
raw material to making MDF. They concluded that the use of waste fiberboards up
to 33 % does not have effect on the mechanical properties of the panels.
Wood is formed by photosynthesis of CO2 and water, thereby blocking carbon in a
durable way. During growth a tree absorbs, through photosynthesis, approximately
the equivalent of 1 ton of CO2 for every m3 growth, while producing the equivalent
of 0.7 ton of oxygen [81].
Wood products require less energy for manufacturing (up to 6000 MJ/m³) than
alternative raw materials, hence contributing even more to the reduction of fossil
fuel consumption. By using the full potential of wood (sink and substitution effects)
in buildings, Europe could reduce emissions of CO2 with 300 million ton or 15 to
20% [78].
In 2003 the European Woodworking Industries, Pulp and Paper Industries and the
European Commission created a work group for discuss the use of the wood
sources with energy and wood products [81]. The main recommendation was to
consider “wood-based products as carbon sinks under the Kyoto Protocol, thereby
acknowledging the contribution of wood-based products to climate change
mitigation and the carbon cycle (see Figure 1.16), and recognize their superior eco-
efficiency versus other materials, as well as their outstanding properties in
recycling with minimal energy use” [81].
Figure 1.16. Cycle of Wood based panels
1.3 Motivation and Outline
This work started within the scope of the
Optimização de Resinas de Ureia
Compósitos de Madeira de Diferentes Espécie
Inovação. The goals of this project
characterization of UF resin and optimize the synthesis process, focusing in three
key variables, temperature, pH, and sequential addition of raw materia
project has an academic partner LEPAE (Porto, Portugal) and an industrial partner
Euroresinas - Indústrias Químicas, S.A. (Sines, Portugal).
The present thesis is divided into seven chapters, including this introduction.
Chapter II, “A study on the colloidal nature of
of UF resins was exhaustively studied
implications of the colloidal phase. Industrial resins with different molar ration F/U
were studied during long time spans in order to
Introduction
27
Cycle of Wood based panels (source [78]).
the scope of the project entitled “UFMadeira -
Optimização de Resinas de Ureia-Formaldeído para a Produção de Materiais
Compósitos de Madeira de Diferentes Espécies/Origens” funded by Agência de
The goals of this project were to develop competences in the
characterization of UF resin and optimize the synthesis process, focusing in three
variables, temperature, pH, and sequential addition of raw materials. This
project has an academic partner LEPAE (Porto, Portugal) and an industrial partner
Indústrias Químicas, S.A. (Sines, Portugal).
The present thesis is divided into seven chapters, including this introduction. In
the colloidal nature of UF resins”, the two- phase character
of UF resins was exhaustively studied in order to understand the formation and
colloidal phase. Industrial resins with different molar ration F/U
spans in order to assess the changes of the colloidal
Chapter 1
28
particles.
Chapter III, “Characterization of UF resins”, discusses the use of two techniques
(GPC/SEC and HPLC) as mutually complementary methods for characterizing
industrially produced resins. The differences between the resins and the effects of
ageing are analyzed.
Chapter IV, “Optimization of UF resins”, explores the use of a design of experiments
(DoE) methodology to optimize the synthesis process of UF resins. Focusing on the
conditions of the condensation step on an alkaline-acid process, response surface
methodology (RSM) was employed to optimize the 3 selected factors (number of
urea additions, time span between urea additions, and pH of condensation
reaction), targeting maximum internal bond strength and minimum formaldehyde
release of the final particleboards.
Chapter V, “Monitoring of UF resin synthesis by GPC/SEC”, investigates a recently
developed process for producing UF resins. The potential of the GPC/SEC technique
in monitoring the resin synthesis is demonstrated in this chapter.
Chapter VI, “Evaluation of UF resins performance”, describes the experimental
evaluation of the curing behaviour of four resins produced within this work, by two
different processes, using the Integrated Pressing and Testing System (IPATES) and
the Automated Bonding Evaluation System (ABES).
Concluding remarks and future work can be found in chapter VII.
Figure 1.17 shows a schematic diagram of the structure of this thesis.
Introduction
29
Figure 1.17. A schematic diagram of linkage between various Chapters present in this thesis.
1.4 References
[1] SRI, Urea-Formaldehyde (UF) Resins, SRI Consulting, 2009. [2] M. Dunky, Macromolecular Symposia 217, 417 (2004). [3] T. Sigvartsen and M. Dunky, New Adhesive Development to Meet the
Challenges of Tomorrow. In Wood Adhesives, San Diego, California, USA, 2005.
[4] A. Pizzi and K. L. Mittal, Handbook of Adhesive Technology, Marcel Dekker, New York (2003).
[5] M. Dunky, Urea-Formaldehyde Glue Resins. In Polymeric Materials
Encyclopedia, J. C. Salamone, Ed. CRC Press: 1996; Vol. 11, pp 8502. [6] E. Athanassiadou and M. Ohlmeyer, Emissions of formaldehyde and VOC from
wood-based panels. In Performance in Use and New Products of Wood
Based Composites, M. Fan, M. Ohlmeyer, M. Irle, H. Haelvoet, E. Athanassiadou and I. Rochester, Eds. Brunek University Press: 2009.
[7] E. Athanassiadou, S. Tsiantzi and C. Markessini, Towards composites with formaldehyde emission at natural wood levels. In COST Action E49
"Measurement and Control of VOC Emissions from Wood-Based Panels", Braunschweig, Germany, 2007.
[8] C. Durkic, J. Bueso and E. Fliedner, Dynea AsWood™ - Technologies for Composite Boards with formaldehyde emissions as defined by nature. In International Conference on Wood Adhesives, C. Frihart, Ed. Lake Tahoe, Nevada, USA, 2009.
[9] B. Tollens, Ber Disch Chem Ges 17, 659 (1884).
Study on the colloidal nature of
UF resins
Optimization of the synthesis of UF
resins
Monitoring of UF resin synthesis by
GPC/SEC
Evaluation of UF resins
performance
Characterization of UF resins
Chapter 1
30
[10] C. Goldschmidt, Ber Disch Chem Ges 29, 2438 (1896). [11] J. E. Carlson, US Patent 4429075 (1984). [12] G. Graves, Urea-formaldehyde resins: Yesterday, today, and tomorrow (1999). [13] Z. Que, T. Furuno, S. Katoh and Y. Nishino, Building and Environment 42, 1257
(2007). [14] R. Schmidt and T. Holloway, The Evolution of Composite Panel Resin
Technology In International Conference on Wood Adhesives, C. Frihart, Ed. Lake tahoe, Nevada, USA, 2009.
[15] H. Spurlock, US Patent 4381368 (1983). [16] S. Vargiu, S. Giovanni, G. Mazzolen and U. Nistri, US Patent 3842039 (1974). [17] J. H. Williams, US Patent 4410685 (1983). [18] ICIS, Formaldehyde Production and Manufacturing Process.
http://www.icis.com/v2/chemicals/9076014/formaldehyde/process.html (2009).
[19] ICIS, Urea Production and Manufacturing Process. http://www.icis.com/v2/chemicals/9076560/urea/process.html (2009).
[20] K. C. Nicolaou and T. Montagnon, Molecules That Changed the World, Wiley-VCH (2008).
[21] J. H. Meessen and H. Petersen, Urea. In Ullmann's Encylopedia of Industrial
Chemistry, 6th ed.; Wiley-VCH, Ed. 2002. [22] B. Meyer, Urea-Formaldehyde Resins, Addison-Wesley, London (1979). [23] IFA, http://www.fertilizer.org/ [24] G. Reuss, W. Disteldorf, A. O. Gamer and Albrecht Hilt, Formaldehyde. In
Ullmann's Encylopedia of Industrial Chemistry, 6th ed.; Wiley-VCH, Ed. 2002. [25] F. C. Inc., Economic Primer on Formaldehyde, Global Insight, 2006. [26] X. Tang, Y. Bai, A. Duong, M. T. Smith, L. Li and L. Zhang, Environment
International 35, 1210 (2009). [27] IARC, Monographs Vol 88: Formaldehyde, 2-Butoxyethanol and 1-tert-
Butoxypropan-2-2-ol, 2006. [28] A. Pizzi, Wood adhesives : Chemistry and Technology, Marcel Dekker, New
York (1983). [29] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998). [30] B. Tomita and Y. Hirose, Journal of Polymer Science: Polymer Chemistry Edition
14, 387 (1976). [31] A. H. Conner, Urea-Formaldehyde Adhesive resins. In Polymeric Materials
Encyclopedia, J. C. Salamone, Ed. CRC Press: 1996. [32] J. K. Fink, Urea/formaldehyde Resins. In Reactive Polymers Fundamentals and
Applications, J. K. Fink, Ed. William Andrew, Inc.): 2005. [33] F. Dongbin, L. Jianzhang and M. An, Journal of Adhesion and Interface 7, 45
(2006). [34] W. K. Motter and D. M. Harmon, WO Patent 2006-015331 (2006). [35] B. Stefke and M. Dunky, Journal of Adhesion Science and Technology 20, 761
(2006).
Introduction
31
[36] K. Siimer, T. Kaljuvee and P. Christjanson, Journal of Thermal Analysis and
Calorimetry 72, 607 (2003). [37] S. Chow and P. R. Steiner, Holzforschung 29, 4 (1975). [38] B. D. Park, E. C. Kang and J. Y. Park, Journal of Applied Polymer Science 101,
1787 (2006). [39] J. I. Dejong, Recueil Des Travaux Chimiques Des Pays-Bas-Journal of the Royal
Netherlands Chemical Society 69, 1566 (1950). [40] J. I. Dejong and J. Dejonge, Recueil Des Travaux Chimiques Des Pays-Bas-
Journal of the Royal Netherlands Chemical Society 71, 643 (1952). [41] J. I. Dejong and J. Dejonge, Recueil Des Travaux Chimiques Des Pays-Bas-
Journal of the Royal Netherlands Chemical Society 71, 890 (1952). [42] J. I. Dejong and J. Dejonge, Recueil Des Travaux Chimiques Des Pays-Bas-
Journal of the Royal Netherlands Chemical Society 72, 207 (1953). [43] J. I. Dejong and J. Dejonge, Recueil Des Travaux Chimiques Des Pays-Bas-
Journal of the Royal Netherlands Chemical Society 72, 139 (1953). [44] M. G. Kim and L. W. Amos, Industrial & Engineering Chemistry Research 29,
208 (1990). [45] J. R. Ebdon and P. E. Heaton, Polymer 18, 971 (1977). [46] G. E. Myers, Journal of Applied Polymer Science 26, 747 (1981). [47] S. S. Jada, Journal of Macromolecular Science-Chemistry A27, 361 (1990). [48] J. Billiani, I. Amtmann, T. Mayr and K. Lederer, Journal of Liquid
Chromatography 13, 2973 (1990). [49] S. Katuscak, M. Tomas and O. Schiessl, Journal of Applied Polymer Science 26,
381 (1981). [50] E. Minopoulou, E. Dessipri, G. D. Chryssikos, V. Gionis, A. Paipetis and C.
Panayiotou, International Journal of Adhesion and Adhesives 23, 473 (2003).
[51] E. Dessipri, E. Minopoulou, G. D. Chryssikos, V. Gionis, A. Paipetis and C. Panayiotou, European Polymer Journal 39, 1533 (2003).
[52] G. E. Maciel, N. M. Szeverenyi, T. A. Early and G. E. Myers, Macromolecules 16, 598 (1983).
[53] S. Tohmura, C. Y. Hse and M. Higuchi, Journal of Wood Science 46, 303 (2000). [54] M. Szesztay, Z. Laszlohedvig, E. Kovacsovics and F. Tudos, Holz als Roh- und
Werkstoff 51, 297 (1993). [55] R. O. Ebewele, B. H. River and G. E. Myers, Journal of Applied Polymer Science
52, 689 (1994). [56] S. Yin, 1994. Charactérisation par analyses thermiques de la polycondensation
d’adhésifs aminoplastes et du durcissement de composites modèles bois-adhésif, Ph.D. Thesis, Université de Nancy I, Nantes, France.
[57] K. Umemura, S. Kawai, Y. Mizuno and H. Sasaki, Mokuzai Gakkaishi 42, 489 (1996).
[58] P. E. Humphrey, US Patent 5176028 (1993). [59] C. G. Hill, A. M. Hedren, G. E. Myers and J. A. Koutsky, Journal of Applied
Chapter 1
32
Polymer Science 29, 2749 (1984). [60] L. M. H. Carvalho, 1999. Estudo da Operação de Prensagem do Aglomerado de
Fibras de Média Densidade (MDF): Prensa Descontínua de Pratos Quentes, Ph.D. Thesis, University of Porto, Porto.
[61] L. M. H. Carvalho, M. R. P. F. N. Costa and C. A. V. Costa, Journal of Applied
Polymer Science 102, 5977 (2006). [62] M. Zanetti, A. Pizzi, M. Beaujean, H. Pasch, K. Rode and P. Dalet, Journal of
Applied Polymer Science 86, 1855 (2002). [63] A. Despres and A. Pizzi, Journal of Applied Polymer Science 100, 1406 (2006). [64] M. Zanetti and A. Pizzi, Journal of Applied Polymer Science 91, 2690 (2004). [65] P. Christjanson, K. Siimer, T. Pehk and I. Lasn, Holz Als Roh-Und Werkstoff 60,
379 (2002). [66] B. Meyer and K. Hermanns, Formaldehyde Release from Wood Products: An
Overview. In Formaldehyde Release from Wood Products, B. Meyer, B. A. K. Andrews and R. M. Reinhardt, Eds. American Chemical Society: Washington, DC, 1986.
[67] FormaCare, http://www.formaldehyde-europe.org/ [68] F. Bulian, R. Battaglia and S. Ciroi, European Journal of Wood and Wood
Products 61, 213 (2003). [69] CARB, Airborne Toxic Control Measure to Reduce Formaldehyde Emissions
from Composite Wood Products, California Environmental Protection Agency, 2007.
[70] M. Risholm-Sundman, A. Larsen, E. Vestin and A. Weibull, Atmospheric
Environment 41, 3193 (2007). [71] D. M. Harmon, CHANGE – Its Challenges and Opportunities. In Technical
Formaldehyde Conference, Hannover, Germany, 2008. [72] H. Schwab, R. Marutzky and B. Meyer, European Regulations for
Formaldehyde. In Technical Formaldehyde Conference, Hannover, Germany, 2008.
[73] APA, http://www.apawood.org/ [74] J. A. Youngquist, Wood-based Composites and Panel Products. In Wood
Handbook: wood as an engineering material, F. P. Laboratory, Ed. 1999. [75] Sonae, http://www.sonae-industria-tafisa.com/ [76] FAO, http://www.fao.org/ [77] K. Wijnendaele, Panel Industry Struggles Through Economic Crises towards
Bright Future for Carbon-Storing Panels. In International Panel Products
Symposium, M. Spear, Ed. Nantes, France, 2009. [78] EPF, http://www.europanels.org [79] EPF, Annual Report 2004-2005, 2005. [80] E. Roffael, C. Behn, B. Dix and G. Bar, Recycling UF-Bonded Fibreboards. In
International Panel Products Symposium, M. Spear, Ed. Nantes, France, 2009.
[81] CEI-Bois, Tackle Climate Change: Use Wood, 2007.
CHAPTER 2
The colloidal nature of UF resins and its relation with adhesive performance
35
2 The colloidal nature of UF resins and its relation with
adhesive performance1
Abstract
UF resins present a swollen colloidal phase dispersed within a continuous water
phase containing soluble oligomers. The main goal of the present investigation is to
clarify the physical and chemical nature of those two phases and elucidate their
impact on the bonding process.
Optical and electronic microscopy has provided information on the morphology of
the colloidal phase, showing primary particles and particle agglomerates.
Mechanisms are suggested for the colloidal stabilization and dilution-induced
flocculation.
Three commercial UF resins with different F/U molar ratios were studied using
particle size distribution (PSD) analysis. The results showed the influence of the
resin degree of condensation and the ageing status on the size of the colloidal
structures.
Gel permeation chromatography analysis was performed on samples of different
resins and of the respective continuous and dispersed phases, separated by
centrifugation. The quantified fraction of insoluble molecular aggregates present in
the chromatograms was related to the resins synthesis conditions and age.
Differential scanning calorimetry and tensile shear strength tests were carried out
to evaluate the reactivity and adhesive performance of each phase. It is suggested
that the colloidal phase acts as a reactive filler at the wood joint interfaces,
1 J. M. M. Ferra, A. M. Mendes, M. R. N. Costa, L. H. Carvalho and F. D. Magalhães, Journal of Applied Polymer
Science 118, 1956 (2010).
Chapter 2
36
contributing for the resins bonding performance.
2.1 Introduction
Urea-formaldehyde (UF) resins are still the most widely used adhesives in the
manufacture of wood-based panels, mostly due to their high reactivity, low cost
and excellent adhesion to wood. Moreover, the impact of the resin limitations
(potential formaldehyde emissions and low resistance to moisture due to the
reversibility of aminomethylene link) has been considerably alleviated by
improvements in the resin production technology [1].
The multiplicity of chemical and physical structures present in UF resins has made
their study a complex task. In addition, these highly reactive chemical systems are
prone to change during sample preparation for analysis or during the analysis itself.
The difficulty to find suitable solvents is another serious problem [2].
This paper is mainly concerned with the physical structure of these resins and its
influence on the resins performance. Amino resins have been perceived for a long
time to possess a colloidal structure [3]. Pratt et al. [4] were the first to admit the
formation of an intermediate lyophobic colloidal sol which subsequently coalesces
to cause gelation during resin cure, concurrently with the conventional chemical
gelation mechanism caused by step-growth polymerization. Dunker and co-workers
[5] suggested that the formation of insoluble aggregates is originated by the loss of
hydrogen bonding potential between oligomers and water. This is caused by the
decrease in the number of possible hydrogen bonding locations as the
condensation reaction proceeds, together with the formation of intra- and inter-
molecular bonds.
Motter [6] studied the mechanism of nucleation and growth of the colloidal phase
in UF resins. The continuous and disperse phases were separated by centrifugation
and the relative molecular weight distribution (MWD) determined by Gel
The colloidal nature of UF resins and its relation with adhesive performance
37
Permeation Chromatography/Size Exclusion Chromatography (GPC/SEC). The
results obtained indicate that the primary mechanism of disperse phase formation
in this type of resins has to do with precipitation of linear, urea-terminated
oligomers.
Grunwald [7] observed a clear supernatant after centrifugation of UF samples at
high velocities for long times (16 h at 40000 rpm). The characterization of the
continuous and disperse phases by GPC/SEC and 1H-NMR indicated that the main
difference between the phases is the fraction of urea and methylolureas.
Colloidal particle formation, followed by clustering, has been shown to be the
normal way of ageing of aminoplastic resins. Melamine-urea-formaldehyde (MUF)
resins, in particular, have been the object of an extensive study done by Pizzi and
co-workers [3, 8-10]. Despres and Pizzi [10] used optical microscopy to identify the
colloidal structures, reporting considerable differences in shape for UF, MF and
MUF resins. Agglomerates were detected in SEC chromatograms of the UF resins.
They have also observed that, at early stages, colloidal aggregate formation
showed a star-like light interference pattern as in MF resins [11].
The main objective of the present work is to obtain some additional insights on the
nature of the colloidal particles present in UF resins and the role that these may
play on the wood-bonding process. The structural analysis of the colloids was
carried out by optical microscopy, transmission electronic microscopy (TEM) and
scanning electron microscopy (SEM). Particle Size Distribution (PSD) analysis give
information on the effects of condensation degree and ageing on the formation of
colloidal structures. Molecular weight distribution analysis based on GPC/SEC
measurements is influenced by the heterogeneous structure of the resins, since the
disperse phase cannot be completely solubilized. Still, the chromatograms provide
information on the fraction of insoluble material present and this can be related to
the resins synthesis conditions. Reactivity and cure enthalpies are evaluated by DSC
Chapter 2
38
for a resin and the individual phases separated by centrifugation. Finally, the shear
strength of glued joints is evaluated, allowing for some insights on the role of each
phase on the wood bonding process.
2.2 Materials and Methods
2.2.1 Materials
Several commercial UF resins were characterized. They were produced by
EuroResinas – Indústrias Químicas S.A., Portugal, with different synthesis
formulations under different operating conditions.
Error! Reference source not found. shows the technical data of four UF resins,
which were analysed in this work. All of them are UF resins in water solution. The
main differences in the synthesis of these resins are the duration of the
condensation step and the final amount of urea added. For example, resin D has
the longer duration of the condensation step and a larger amount of final urea,
while resin A has the shortest duration of the condensation step and the lowest
amount of final urea. These resins were available for analysis at different times
along the work, within the context of an ongoing project. For this reason, the
analysis techniques discussed below comprise different sets of samples.
Nevertheless, when relevant, the results obtained are related to the nature of the
resins under study.
To limit the ageing of the resins for posterior analysis, they were stored at 5 ºC
unless otherwise specified.
The colloidal nature of UF resins and its relation with adhesive performance
39
Table 2.1. Technical data for UF-resins analysed in this work
Resin Molar
ratio F/U Solids
content1 (%)
Gel time2
(s) pH range
(25 °C) Viscosity
3
(mPa.s)
A 1.30 63±1 45±10 7.5-8.5 150-350 B 1.20 64±1 80±30 8.0-9.5 150-350 C 1.08 64±1 80±30 7.0-9.0 150-300 D 1.00 64±1 70±30 8.0-9.5 150-300
1105 °C, 3 h
2Gel time at 100 °C with 3 % wt of NH4Cl (20 % wt solution)
3Brookfield viscometer at 25 °C
2.2.2 Methods
Morphology analysis
Direct visualization of UF resins was carried out with a Leica DM LB2 microscope
(Leica Microsystems) with 1000× magnification and equipped with a filter cube
with the following characteristics: excitation filter 340–380 nm, dichromatic mirror
400 nm and suppression filter LP 425, without oil immersion. The sample was
prepared diluting approximately 1-2 % of UF resin in water.
For visualization by TEM (50 kV; Zeiss), samples were adsorbed to glow-discharged
carbon-coated collodion film on 400-mesh copper grids during 1 minute. The
sample was prepared diluting approximately 1-2 % of UF resin in water.
SEM/EDS images were taken with a SEM-JEOL-JSM6301-F equipped with an Oxford
INCA/ENERGY-350 microanalysis system. The deposition of resin droplets on the
surface of a metallic sheet and on wood was also investigated. A water solution
with 20 wt % of ammonium chloride was used as hardener and mixed with the
resin before spraying on a metallic sheet. The hardener concentration was 3 wt %
(based on resin solids content) in the final mixture. The boards were then cured in
an oven at 110 °C for 8 min [12]. The samples were coated with gold by
vaporization before the analysis.
For structural analysis by CryoSEM, the sample was placed on a copper CryoSEM
Chapter 2
40
sample holder in a circular hole. The set was dipped into liquid nitrogen (slush
nitrogen) at -170 °C; the freezing occurred under vacuum. After freezing, the
sample was transferred to a Gatan Alto 2500 cryo-chamber of the microscope
(JEOL, JSM-6700F). The sample was fractured in the cryo-stage with a knife at -130
°C, after the aqueous phase sublimation was performed at -98 °C under vacuum
during 1 min. Finally, the sample was gold coated by vaporization before the
analysis. The sample was observed at a temperature of -130 °C and a voltage of 10
kV.
Particle size analysis
A Beckman Coulter LS230 (LS230) with a wet module was used in this study. The
dynamic range of the LS230 goes from 0.04 µm to 2000 µm. The LS230 instrument
uses Polarization Intensity Differential Scattering (PIDS) method. The PIDS system
uses three light wavelengths (450 nm, 600 nm and 900 nm, obtained by filtering
light from an incandescent bulb) at two polarizations. Measurements were carried
out at several scattering angles of the polarized vertical light to the scattering plane
and the polarized horizontal light to the scattering plane. The samples were
prepared diluting 5 drops of UF resin in 10 mL of distilled water, these were
ultrasonicated for 1 min and analysed in the Coulter-Counter.
Centrifugation
Phase separation of UF resin was carried out by centrifugation. The samples were
fractionated at 18000 rpm during 3 h using a JA25.50 rotor in Beckmann Coulter
Avanti J25 centrifuge at 12 °C in 50 mL polyallomer bottles. The supernatant was
withdrawn with a micropipette, while the sediment was withdrawn after removing
the intermediate phase. The supernatant, sediment and a sample of resin were
stored in closed vials at 5 °C for future analysis.
Molecular weight distribution
A Gilson HPLC system equipped with a Gilson Differential RI detector and Viscotek's
The colloidal nature of UF resins and its relation with adhesive performance
41
Dual Detector (differential viscometer and a light scattering detector RALLS). A
Rheodyne 7125 injector with a 20 µL loop was used for sample introduction. The
column used was a Waters Styragel HR1 5 µm. Dimethylformamide (DMF) was used
as the mobile phase. The column was conditioned at 60 °C using an external oven
and the flow rate was 1 mL/min. The Universal Calibration was done using
polyethylene glycol standards from PL with molecular weight between 12140 –
106. The RALLS detector was not used for the tri-SEC molecular weight calculations
because of the low response for lower molecular weights. However, the RALLS
signals were qualitatively taken into account.
The samples for SEC analysis were prepared dissolving 100 mg sample of resin in 3
mL of dimethylsulfoxide (DMSO), then stirring vigorously and filtering through a
0.45 µm Nylon syringe filter.
Curing and bonding strength characterization
About 15 mg of each sample was mixed with 3 wt % (based on resin solids content)
of hardener (20 wt % solution of ammonium chloride). A Setaram DSC 131 was
used with a temperature ramp of 5 °C/min, from 25 to 300 °C. The samples were
held in stainless steel crucibles with a crimpable seal and able to withstand 80 bars
up to 550 °C. The use of this crucible enables prevention of water evaporation in
the scanning temperature range. The analysis was performed under nitrogen
atmosphere.
The same mixtures used in the DSC were subjected to a tensile shear test in a
Servosis universal testing machine. Approximately 150 g/m2 adhesive were applied
to the bonding surfaces of lap joints test pieces.
The test was carried out according to the procedure described by EN 205. The
loading speed was 10 mm/min. The loading was continued until a break or
separation occurred on the surface of the test samples. The tensile shear strength
was then calculated in N·mm-2.
Chapter 2
42
2.3 Results and discussion
2.3.1 Disperse phase morphology
The resins are white opaque liquids, indicating the presence of an insoluble
disperse phase in the aqueous medium. The colloidal nature of UF resins has been
reported previously [4, 10, 13-14] and is associated to the presence of insoluble
molecular aggregates2. Furthermore, above about 100 % dilution with water,
flocculation takes place, originating a white precipitate and a translucent
supernatant liquid. The mechanism that originates this flocculation is not clearly
identified and is certainly dependent on the particular nature of the resins
considered.
Results from different microscopy techniques are presented below, providing a
consolidated view of the structure of the disperse phase and serving as a basis for
the formulation of some hypothesis.
Optical microscopy is the easiest method to observe the morphology of the resins.
Due to the high solids concentration, significant dilution is necessary for the
microscopic visualization to be possible. Therefore, flocculation is inevitable and
particle agglomerates are formed, which were not present in the original resin. As
an example, Figure 2.1a displays a sample of resin B diluted in water (1.6 % resin in
water, resin had been stored at 5 °C for about 90 days). A large agglomerate of
smaller particles (primary particles) is visible.
Interestingly, when dilution is performed with a urea solution, instead of pure
water, distinctly smaller floccules are formed. This difference is visible to the naked
eye and is confirmed in Figure 2.1b (1.3 % resin in urea solution at 10 %
concentration). The presence of excess urea may cause the solvation of the primary
2 In this paper we adopt the terminology suggested by Nichols et al. [15], using the term aggregate to describe “an
association of molecules into supramolecular structures, which in turn may develop into agglomerates.” Along this
paper, the terms aggregate and primary particle are used indistinctly.
The colloidal nature of UF resins and its relation with adhesive performance
43
particles and minimizes inter-particle interaction, namely hydrogen bonding,
therefore decreasing the tendency towards agglomeration. This behaviour is similar
to the one observed in the well known chemical denaturation of proteins by urea
solutions (e.g. Caballero-Herrera et al. [16]). The effect of urea on minimizing
colloidal aggregation on MUF resins has previously been discussed by Zanetti and
Pizzi [3].
Figure 2.1. Optical microscopy image of resin B: a) diluted in water (1.6 wt % resin in water) and b) diluted in 10 wt % urea solution (1.3 wt % resin in urea solution).
At this point, it is useful to formulate a descriptive model of the heterogeneous
nature of these resins, which will be helpful in the interpretation of the results
discussed throughout de rest of the paper. Figure 2.2 presents a schematic view of
the resins heterophasic structure.
Figure 2.2a shows particles formed by insoluble molecular aggregates, swollen by
Chapter 2
44
water and soluble oligomers. These particles are not necessarily individualized,
some agglomeration being possible. In addition to this dispersed phase, the
continuous phase contains unreacted urea and soluble polymer, ranging from small
oligomers to medium range molecular weight polymer.
Figure 2.2. Simplified representation of a UF resin heterogeneous nature: a) original resin, and b) resin after flocculation by water dilution.
Unreacted urea may form a solvation layer surrounding the particles, contributing
to its stabilization towards agglomeration. Upon dilution-induced flocculation,
particle agglomeration occurs, originating optically visible large floccules. This
instability towards dilution can be associated to the depletion of the solvation layer
due to diminishing urea concentration in the continuous phase. This explains why
flocculation with urea solution is more difficult and originates much smaller
floccules.
An additional colloidal stabilization mechanism may be taken into account, based
on electrostatic repulsion associated to ionic species retained at the particle
surface. A fraction of the polymeric species may have an intermediate hydrophilic
character which causes it to be distributed between the dispersed and continuous
phases. These species should have higher contents of very hydrophilic terminal
hydroxymethylene and primary urea groups than the insoluble fraction. This makes
possible the capture of ionic species, i.e. sodium cations, at the particle surface,
The colloidal nature of UF resins and its relation with adhesive performance
45
thus originating electrostatic stabilization. Upon dilution, the polymer at the
surface is partially displaced towards the aqueous medium, depleting the surface
from ionic species. This leads to agglomeration and flocculation.
TEM images for resin B fresh and aged (stored for approximately 2 months at 5 °C)
are shown in Figure 2.3. Since TEM analysis implies significant dilution, we are
again looking at the flocculated samples. These agglomerated structures are
evident in both samples. The primary particles have dimensions of tenths of
micron. Interestingly, the aged sample showed larger agglomerates, while in the
fresh sample these were smaller and less numerous. This relation between
floccules dimensions and ageing will be further discussed below.
Figure 2.3. TEM images of resin B diluted in water. a) fresh resin, and b) aged resin.
Chapter 2
46
In the case of CryoSEM analysis, resin dilution is not necessary. Additionally, no
particle coalescence occurs since the sample is kept frozen. Figure 2.4 shows an
agglomerate with approximately 3 µm. The relatively large size is probably
associated with the ageing process. The primary particles have dimensions in the
tenths of micron, in agreement with the TEM observations. Interestingly, the
particle geometry does not seem to be only spherical, but also lamellar. These two
types of geometries seem to be also present in the results reported by Despres and
Pizzi [10].
Figure 2.4. Cryo-SEM image aged resin B.
An example of the SEM images collected for the resin sprayed and cured on a metal
plate is shown in Figure 2.5. Two different phases are visible, as expected: a
continuous phase, corresponding to cured soluble polymer, and a dispersed phase
made of individualized primary particles and small agglomerates. The particles have
sizes in the range of a few tenths of micron. In the agglomerates, some particle
coalescence due to heating is visible. This image somehow illustrates how the resin
is deposited on the wood particle surface during the glue blending process. The
continuous phase should penetrate to some degree within the wood structure, but
the particles are likely to remain at the surface.
The colloidal nature of UF resins and its relation with adhesive performance
47
Figure 2.5. SEM image of UF resin B sprayed and cured on a metal plate.
Figure 2.6 presents the elementary analysis carried out at different points in the
cured resin surface. The particle surface shows a significantly higher content in
sodium, chlorine and potassium, due to the adsorption of ions present in water.
The ability for the surface to capture sodium cations may play an important role on
the colloidal stabilization process, as discussed previously. The surface also shows
higher nitrogen content relatively to carbon and oxygen. This could be related with
the presence of urea in the disperse phase, as also suggested before.
Chapter 2
48
Figure 2.6. Elementary analysis for UF resin B cured on a metal plate. a) particle surface, and b) in continuous phase.
Summarizing some of the conclusions derived from the previous observations:
- The colloidal structures are agglomerates of elementary particles. These
are a few tenths of micron in diameter and exhibit spherical and lamellar
geometries.
- Urea can play an important role on the stabilization of the colloidal
particles, through the formation of a solvation layer. Adsorption of ionic
species at the particles surface may also contribute to electrostatic
stabilization.
The colloidal nature of UF resins and its relation with adhesive performance
49
- Water dilution causes the weakening of both types of stabilization
mechanisms, due to surface depletion, and originates
agglomeration/flocculation.
- When a resin is sprayed over a porous substrate, such as wood, the
particles will remain at the surface while the soluble fractions should be
absorbed together with water. This will have an effect on the bonding
mechanism, as discussed later.
2.3.2 Particle size distribution
The purpose of these PSD measurements was to analyze the effect of the degree of
condensation and ageing. It must be noted that the Coulter-Counter system implies
dilution of the samples above the flocculation limit. Therefore, one is looking at
flocculated samples and not at the original PSD. Figure 2.7a and Figure 2.7c shows
the volume PSD obtained for resins A, B and C, relatively fresh (stored for 30 days
at 5 °C) and aged (stored for 170 days at 5 °C). For the fresh samples, it is clear that
resin A originates a different type of flocculate than resins B and C. The volume
distribution for resin A lies mostly between 0.2 to 12 μm, while the others are
displaced towards the right, in the region of hundreds of micron, but still showing
a small peak below 10 μm. On the other hand, after ageing all resins show similar
PSDs.
The much larger dimensions of the flocculates in the less aged resins B and C, in
relation to resin A, evidenced in the volume distributions, may be related to the
lower degree of condensation of resin A. The colloidal phase present in fresh resin
A will be composed of less numerous and not very agglomerated particles that do
not yield such large floccules. With ageing, more molecular aggregates are formed
and more agglomeration takes place, therefore approaching the morphology of
resins B and C and originating floccules with a similar PSD.
A subtle additional difference exists between the fresh resins and is more visible
Chapter 2
50
when looking at the number PSDs, shown in Figure 2.7b and Figure 2.1d. The PSD
for fresh resin A does not show particles below 0.1 μm, in contrast to resins B and
C. After ageing, however, the three resins again show similar number PSDs, all
starting below that size. The number distributions put into evidence the lower
particle sizes present, certainly displaying primary particles that have remained
unagglomerated. Resins B and C display a large number of particles with sizes
around 0.1 μm, which is consistent with the dimensions of the individual particles
identified in the previous microscopy sections. Resin A, on the other hand, being a
less condensed resin, contains a lower concentration of agglomerates, which is
confirmed by the lower amount of sediment observed after flocculation. Therefore,
a lower amount of the original primary particles remain after flocculation.
The reproducibility of these results, both for fresh and aged conditions, has been
confirmed with other batches of the same resins, not shown here.
In summary, even though these PSD analysis combines information on the original
primary molecular aggregates and on the agglomerates formed upon dilution,
some qualitative conclusions can be withdrawn. More condensed resins display
larger, more agglomerated, colloidal structures. Upon dilution, also larger floccules
are formed. A less condensed resin will approach this morphology along time, as
more particles (molecular aggregates) and particle agglomerates are formed.
The colloidal nature of UF resins and its relation with adhesive performance
51
Figure 2.7. Particle size distribution for three UF resins stored at 5 °C. a) PSD in volume for 30 days, b) PSD in number for 30 days – note that the curves for samples B and C are nearly superimposed, c) PSD in volume for 170 days, and d) PSD in number for 170 days.
In order to clarify the nature of the two phases present in these UF resins, some
samples were centrifuged and the sediment and supernatant phases separated.
Coulter Counter analysis was performed on the supernatant and compared to the
results from the original resins. The sediment consisted of a gel-like mass which
could not be redispersed for this analysis. These samples were also analysed in the
chromatographic and calorimetric studies described later in the paper.
A completely clear supernatant has never been obtained, indicating that a residual
colloidal phase was still dispersed in the continuous medium. On the other hand,
Chapter 2
52
the sediment is expected to still contain, in addition to water, soluble polymer and
oligomers, entrapped in inter and intraparticular voids, or retained due to surface
adsorption on the dispersed phase [6].
Resin A was studied for two different ages: 20 and 50 days at 25 °C, while resin D
was analysed only for 20 days, since it was gelified after 50 days at 25 °C. Table 2.2
displays the results obtained for the solids content (after evaporation at 105 °C for
3 hours) in the original resins and in the separated phases.
Table 2.2. Solids content of original resins, sediment and supernatant phases for UF resins A-2 and D
Resin Resin Age
(days at 25 °C) Resin solids
(wt %) Sediment
solids (wt %) Supernatant solids (wt %)
A-2 20 62.2 66.2 61.0
50 63.0 67.8 60.2
D 20 65.0 72.2 63.7
The PSD results obtained are shown in Figure 2.8 and Figure 2.9. The results for the
resins before centrifugation are qualitatively similar to the ones described before,
even though the ageing times are not comparable due to the different storage
temperatures. Resin A with 20 days shows a narrower PSD than the more
condensed resin D. As before, resin A with this lower age (Figure 2.8a) does not
exhibit particles in the neighbourhood of 0.1 µm, contrary to resin D (Figure 2.9).
As resin A ages (Figure 2.8b), however, the PSD of the flocculated material widens
and the mean value is significantly displaced towards the right. Particles in the
neighbourhood of 0.1 µm are now displayed in the analysis.
The PSD of the flocculated supernatant, on the other hand, is similar for all resins.
Interestingly, as can be seen in Figure 2.8a, this distribution is almost the same as
for the fresh uncentrifuged resin A. This was somehow expected, considering that
the supernatant should contain only small unagglomerated molecular aggregates,
similar to the ones present in the fresh low-condensed resin. Agglomeration caused
The colloidal nature of UF resins and its relation with adhesive performance
53
by ageing leads to a broader PSD of the flocculate, as seen in Figure 2.8b and Figure
2.9.
Figure 2.8. Particle size distribution in volume for UF resin A-2 and supernatant. a) resin stored for 20 days at 25 °C, and b) resin stored for 50 days at 25 °C.
Figure 2.9. Particle size distribution in volume for UF resin D and supernatant stored for 20 days.
Chapter 2
54
2.3.3 Molecular weight distribution
The work below attempts to relate the information obtained from conventional
GPC/SEC analysis with the known heterogeneous character of these resins. This
may provide additional information for the characterization of the resins molecular
and colloidal structures. Chromatograms for resins A-1 and D and the
corresponding supernatant and sediment phases, obtained after centrifugation, are
shown in Figure 2.10 and Figure 2.11. In all cases, at least two samples were
prepared and analysed in order to verify the reproducibility of the results. Figure
2.10a shows the normalized refraction index (RI) chromatogram for resin A-1. The
general features of the chromatogram are similar to those found in the literature
for UF resins [1, 7, 10, 17-19]. Three zones, based on the apparent limits of
detectable peaks, can be defined in the chromatograms:
Zone I - elution volume between 8 and 9 ml. Corresponds to the lower
molecular weight species, i.e. unreacted urea, methylolureas and
oligomeric species, such as methylenediurea,
monomethylolmethylenediurea, dimethylolmethylenediurea and others.
The methylenediurea was synthesized with Kadowaki’s method [20] in
laboratory and analysed in SEC. Molecular weight should be below about
600.
Zone II - elution volume between 5.8 and 8 ml. Corresponds to
intermediate molecular weight species, with molecular weights ranging
from a few tens of thousand to about 600.
Zone III - elution volume below 5.8 ml. This would correspond to polymer
with quite high molecular weights, eluting before the exclusion limit of the
SEC column. It has been suggested by some authors that this portion of the
chromatograms actually refers to molecular aggregates and not
individualized polymer molecules [10, 18]. These aggregates would be
The colloidal nature of UF resins and its relation with adhesive performance
55
insoluble in the original aqueous medium, probably forming larger colloidal
structures that partially disaggregated in the DMSO solvent. In this paper
we present some evidence in favour of this hypothesis.
Figure 2.10. Chromatogram for UF resin A-2 with 20 and 50 days, sediment and supernatant diluted 3 % in DMSO. a) normalized response of RI sensor for 20 days, b) response of RALLS sensor for 20 days, c) normalized response of RI sensor for 50 days, and d) response of RALLS sensor for 50 days.
Chapter 2
56
Figure 2.11. Chromatogram for UF resin D with 20 days, sediment and supernatant diluted 3 % in DMSO. a) normalized response of RI sensor, and b) response of RALLS sensor.
Since the chromatograms may reflect the presence of insoluble molecular
aggregates, a straightforward computation of average molecular weights would be
misleading. Two different approaches were therefore taken in order to
quantitatively represent the chromatographic data. On one hand, assuming that
zone III corresponds essentially to insoluble material, molecular weights were
computed neglecting this portion of the chromatograms. On the other hand the
two following parameters were introduced in order to complement the description
of the particular features of these chromatograms:
amchromatogr of areatotal
I Zone of areaf
1= (2.1)
III Zone II Zone of areas
III Zone of areaf2 += (2.2)
f1 should reflect the amount of low molecular weight species in the sample, while f2
is expected to be related to the relative importance of what would be high
molecular weight species in the polymerized material. Table 2.3 show the
numerical values obtained for these parameters in all the chromatograms.
The colloidal nature of UF resins and its relation with adhesive performance
57
Qualitatively, the RI chromatograms (Figure 2.11 and Figure 2.11) display similar
features for the original resins and the supernatant and sediment phases. Two
discrete peaks in the lower molecular weight region are common to all samples,
with retention volumes of 8.5 and 8.2 mL. The first peak represents urea and
methylolureas, as was confirmed by direct injection of urea and dimethylolurea.
The fraction which eluted at 8.2 mL is probability a polymer formed by reaction of
urea and methylolureas. The chromatogram of the sediment phase displays a peak
for a retention volume of 7.5-8.0 mL (Mw ~ 600-1000), which does not seem to be
present in the supernatant. A similar behaviour was observed by Motter [6] for the
SEC analysis of the sediment and supernatant phases. This author associates this
peak to insoluble linear urea-terminated oligomers, 4 to 8 urea units in length.
These oligomers would correspond to a critical molecular size for solubility in
water, above which the formation of primary colloidal particles takes place.
Table 2.3. Values of Mn, Mw, polydispersity, and parameters f1 and f2, obtained by SEC for resin, sediment and supernatant of the UF resin A-2 with 20 and 50 days and UF resin D with 20 days storage at 25 °C
Sample Mn Mw Mw/Mn f1 f2
Resin A-2 with 20 days storage
Resin A-2 4.69×102 3.40×10
3 7.2 0.296 0.168
Supernatant 4.17×102 2.84×10
3 6.8 0.313 0.167
Sediment 5.63×102 3.50×10
3 6.2 0.248 0.180
Resin A-2 with 50 days storage
Resin A-2 5.10×102 2.74×10
3 5.4 0.225 0.185
Supernatant 4.91×102 2.53×10
3 5.2 0.246 0.185
Sediment 6.39×102 3.64×10
3 5.7 0.190 0.192
Resin D with 20 days storage
Resin D 3.91×102
2.87×103 7.3 0.342 0.182
Supernatant 3.62×102 2.53×10
3 7.0 0.355 0.184
Sediment 5.19×102 3.06×10
3 5.9 0.269 0.200
Considering the data in Table 2.3, for resin A-2 with 20 days at 25 °C, the sediment
shows the highest molecular weights and the supernatant the lowest. The original
resin shows intermediate values. On the other hand, the fraction of low molecular
Chapter 2
58
weight species, f1, is highest for the supernatant and lowest for the sediment. Once
again, the original resin shows an intermediate value. Similar results are seen in
Table 2.3 for the other samples. It was expected that the supernatant would exhibit
the highest fraction of low molecular weight material, which was dissolved in the
aqueous medium of the original resin. The sediment also contains some of this
material, at a lower proportion, because it was sorbed within the solid phase or
was retained in the inter-particular space. Therefore, one concludes that
centrifugation does not lead to complete separation of the disperse and continuous
phases.
It is somewhat surprising that, in all samples, a not very substantial difference is
obtained for the computed molecular weights of sediment and supernatant.
Significantly higher molecular weights would, in principle, be expected for the
sedimented phase. Since this is not the case, we are led to believe that the colloidal
phase is actually composed by aggregated polymer with molecular weights not too
different from the polymer in solution. The more relevant difference between the
soluble and insoluble fractions may lie on chemical composition, namely hydroxyl
group content.
Even though the purpose of this paper is not to study the effect of ageing, when
looking at resin A-2 with 20 and 50 days (Table 2.3), it is noticeable that the
molecular weights increase and f1 decreases with ageing. This indicates that
condensation progresses with time, consuming urea and oligomers. Furthermore, f2
has increased, indicating the presence of more insoluble aggregates.
Resin D (Table 2.3) is the most condensed of the resins studied. This fact is not
evident from the computed molecular weights, since it is masked by the larger
amount of final urea added, as confirmed by the higher value of f1. However, this
resin shows the highest values for parameter f2, which may indeed be related to
the higher degree of condensation.
The colloidal nature of UF resins and its relation with adhesive performance
59
It is also interesting to look at the RALLS responses obtained in the different cases.
For resin A-2 after 20 days, Figure 2.10b shows several peaks roughly in the region
corresponding to zone III, attributable to the molecular aggregates present in the
samples. The peak magnitude is higher for the sediment, as this sample is expected
to be more concentrated in insoluble material. This difference becomes much more
significant when the resin is aged to 50 days, as seen in Figure 2.10d, where the
RALLS signal is actually saturated for the sediment, while it maintains
approximately the same magnitude for the supernatant. As expected, the original
resin displays an intermediate chromatogram. This indicates that the ageing
process has originated a significantly higher concentration of insoluble molecular
aggregates, in agreement with the PSD results discussed previously. These
aggregates would actually be agglomerated into larger structures in the original
resin, sedimenting during centrifugation. Contact with DMSO solvent caused
desagglomeration into particles visible to the RALLS detector.
On the other hand, for resin D, Figure 2.11b, the RALLS chromatogram for the
sediment is less intense than for the supernatant. This more condensed resin may
form insoluble particles that cannot be easily disaggregated in DMSO and are
therefore retained in the pre-column 0.45 µm filter.
Some key conclusions can be extracted from these SEC results:
- The lower retention volume portion of the RI chromatograms reflects the
presence of insoluble molecular aggregates. This fraction of the
chromatogram can be quantified and related to the resins degree of
condensation and ageing condition.
- Based on the analysis of centrifuged samples, the polymer aggregates
present in the colloidal phase do not seem to have higher molecular weight
than the water-soluble polymer. The cause for insolubility may have to do
with differences in chemical composition.
Chapter 2
60
- A larger degree of condensation may not translate into a significantly
higher molecular weight, but into a higher fraction of insoluble aggregates
formed.
- Resin ageing leads to the formation of a higher fraction of insoluble
molecular aggregates.
- The RALLS chromatograms confirm the presence of insoluble particles in
the samples and its relation with ageing.
2.3.4 Curing behaviour and bonding strength
DSC is an often used tool to evaluate the reactivity of UF resins. However, the
analysis conditions are different from those observed in the bonding process, so it
cannot be directly related to the process of mechanical strength development.
The same separate analysis for resin, supernatant and sediment was performed for
resin A. Figure 2.12 shows the DSC curves obtained and Table 2.4 summarizes the
results. The supernatant is more reactive, having a slightly lower cure temperature
than the resin and sediment, which show similar results. The cure enthalpies based
on the samples solids content (from Table 2.2), are significantly different. The
values increase in the following order: sediment < resin < supernatant. This may be
attributed to the higher molecular mobility and lower condensation (more reactive
groups) of the soluble material present in the supernatant.
The colloidal nature of UF resins and its relation with adhesive performance
61
Figure 2.12. DSC curves for UF resin A-2, supernatant and sediment. Resin had been stored for 50 days.
The endothermic peaks that appear between 160 °C and 190 °C can be attributed
to the release of entrapped water from the cured polymer [21]. It is also possible to
observe, above 200 °C, endothermic peaks associated to thermal degradation of
the methylene-ether [22-23].
Table 2.4. Cure temperatures and enthalpies for UF resin A-2, supernatant and sediment
Sample Peak
Temperature (°C) Enthalpy
(J/g)
Resin A-2 83.25 77.57 Supernatant 81.58 85.38 Sediment 83.04 72.18
For evaluating the bonding behaviour of resin A-2, supernatant and sediment, a
tensile shear test of lap joints was performed. The tensile shear strengths
measured are shown in Table 2.5, together with the percentages of cohesive failure
in substrate, obtained by visual inspection of the joint after fracture. Qualitatively
Chapter 2
62
similar results were obtained for resin D.
Table 2.5. Values of tensile shear strength and % of cohesive failure within wood for UF resin A-2, supernatant and sediment
Sample Tensile shear strength
(N·mm-2
) % cohesive failure
in wood
Resin A-2 13.4 90 Supernatant 11.3 60 Sediment 14.6 90
More relevant than looking at the actual tensile strengths measured, may be
comparing the percentages of cohesive failure within wood. These are much higher
for the original resin and the sediment (about 90%). This indicates that in these
samples the failure started within the wood material and not at the adhesive joint.
The measured strength values may therefore be limited by the tensile strength of
the wood substrate. On the other hand, the joint glued with supernatant appears
more fragile, showing a lower percentage of cohesive failure in wood and lower
tensile shear strength, which indicated that the failure was initiated at the adhesive
joint. The poorer performance of the supernatant may be associated to its ability to
penetrate to some degree into the substrate, depleting the interface from glue.
Recalling Figure 2.5, the colloidal structures, which are significantly present in the
sediment, are expected to remain in the wood surface. The presence of these
structures in the joint interface may play an important role in the bonding process,
acting as a reactive reinforcing filler.
There is a generalized concept that a well performing UF resin should incorporate
low molecular weight species, which are important for substrate wetting and
penetration, and higher molecular weight species, which have an important
contribution to cohesive strength [1, 17]. Our results emphasize that the “high
molecular weight” component of the resin may actually take the form of insoluble
molecular aggregates.
The colloidal nature of UF resins and its relation with adhesive performance
63
2.4 Conclusions
Different microscopy techniques allowed for the visualization of colloidal structures
in the UF resins analysed. Primary particles have dimensions in the order of tenths
of micron, showing spherical or lamellar shapes. These particles are found to be
agglomerated to some degree in the original resin. Upon water dilution,
flocculation occurs, originating much larger agglomerates. Two colloidal
stabilization mechanisms were suggested, based upon solvation by unreacted urea,
hindering inter-particle hydrogen bonding, and electrostatic repulsion due to
cations adsorbed at the particle surface. Both mechanisms explain the dilution-
flocculation process in terms of depletion of the stabilizing entities from the
particle surface.
PSD analyses of different resins showed distinct PSDs, which can be related to the
way how the degree of condensation affects the formation of the dispersed solid
phase. These differences tend to vanish as the resins age, indicating that molecular
aggregation and particle agglomeration proceed along time on a less condensed
resin.
The size exclusion chromatography analysis encompasses information on the
molecular weight distribution of the soluble polymer and on molecular aggregates
from the original dispersed phase that have not been completely solubilised in the
solvent. The quantification of the fraction of the RI chromatogram corresponding
to molecular aggregates is relatable to the resins degree of condensation and the
ageing status. The information from the RALLS sensor confirms qualitatively the
presence of insoluble material. Analysis of partially separated phases, obtained by
centrifugation, show that the insoluble portion of the resins does not seem to have
a significantly different molecular weight, in relation to the soluble material. The
formation of colloidal structures may have to do with aggregation caused by inter-
molecular interactions and not with precipitation above a critical molecular weight.
Chapter 2
64
Tensile shear strength experiments performed with the original resin and the
centrifugation products (sediment and supernatant) indicated that the presence of
the dispersed phase may be relevant for the strengthening of the adhesive bond,
acting as a reactive reinforcing filler.
Future work will be focused on the analysis of the resin ageing processes on the
resins characteristics, particularly on its heterogeneous morphology. In addition, it
will be important to study in more detail the bonding mechanism of this bi-phase
structure in the wood composite production.
2.5 References
[1] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998). [2] M. Dunky, Analysis of Formaldehyde Condensation Resins for the Wood Based
Panels Industry: Status and new challenges. In 1st European Panels Products
Symposium, Hague, Loxton, Bolton and Mott, Eds. Llandudno, North Wales, UK, 1997; p 217.
[3] M. Zanetti, A. Pizzi and P. Faucher, Journal of Applied Polymer Science 92, 672 (2004).
[4] T. J. Pratt, W. E. Johns, R. M. Rammon and W. L. Plagemann, Journal of Adhesion 17, 275 (1985).
[5] A. K. Dunker, W. E. John, R. Rammon, B. Farmer and S. J. Johns, Journal of
Adhesion 19, 153 (1986). [6] W. K. Motter, 1990. The formation of the colloidal phase in low mole ratio urea-
formaldehyde resins, Ph.D. Thesis, Washington State University, Washington, USA.
[7] D. Grunwald, Kombinierte analytische Untersuchungen von Klebstoffen für Holzwerkstoffe, Mensch & Buch Verlag, Berlin, Germany (2002).
[8] M. Zanetti, A. Pizzi, M. Beaujean, H. Pasch, K. Rode and P. Dalet, Journal of
Applied Polymer Science 86, 1855 (2002). [9] A. Pizzi, B. George, M. Zanetti and P. J. Meausoone, Journal of Applied Polymer
Science 96, 655 (2005). [10] A. Despres and A. Pizzi, Journal of Applied Polymer Science 100, 1406 (2006). [11] J. Mijatovic, W. H. Binder, F. Kubel and W. Kantner, Macromolecular Symposia
181, 373 (2002). [12] L. M. H. Carvalho, 1999. Estudo da Operação de Prensagem do Aglomerado de
Fibras de Média Densidade (MDF): Prensa Descontínua de Pratos Quentes, Ph.D. Thesis, University of Porto, Porto.
[13] J. Stuligross and J. A. Koutsky, Journal of Adhesion 18, 281 (1985).
The colloidal nature of UF resins and its relation with adhesive performance
65
[14] M. G. Kim, B. Y. No, S. M. Lee and W. L. Nieh, Journal of Applied Polymer
Science 89, 1896 (2003). [15] G. Nichols, S. Byard, M. J. Bloxham, J. Botterill, N. J. Dawson, A. Dennis, V.
Diart, N. C. North and J. D. Sherwood, Journal of Pharmaceutical Sciences 91, 2103 (2002).
[16] A. Caballero-Herrera, K. Nordstrand, K. D. Berndt and L. Nilsson, Biophysical
Journal 89, 842 (2005). [17] R. N. Kumar, T. L. Han, H. D. Rozman, W. R. W. Daud and M. S. Ibrahim, Journal
of Applied Polymer Science 103, 2709 (2007). [18] T. Hlaing, A. Gilbert and C. Booth, British Polymer Journal 18, 345 (1986). [19] G. Zeppenfeld and D. Grunwald, Klebstoffe in der Holz- und Möbelindustrie,
Drw Verlag Weinbrenner (2005). [20] H. Kadowaki, Bulletin of the Chemical Society of Japan 11, 248 (1936). [21] C. Xing, J. Deng, S. Y. Zhang, B. Riedl and A. Cloutier, Journal of Applied Polymer
Science 98, 2027 (2005). [22] M. Szesztay, Z. Laszlohedvig, E. Kovacsovics and F. Tudos, Holz als Roh- und
Werkstoff 51, 297 (1993). [23] T. Zorba, E. Papadopoulou, A. Hatjiissaak, K. M. Paraskevopoulos and K.
Chrissafis, Journal of Thermal Analysis and Calorimetry 92, 29 (2008).
CHAPTER 3
Effect of ageing on UF resins
69
3 Effect of ageing on UF resins1
Abstract
During the last 40 years, several analytical techniques have been
developed/adapted to characterize urea-formaldehyde (UF) resins. However, a
great part of the research about this kind of wood adhesives has been performed
by industrial producers and thus the main part of the existing knowledge is
retained within those companies.
This work describes a methodology for determining the molecular weight
distribution (MWD) of UF resins using Gel Permeation Chromatography (GPC)/Size
Exclusion Chromatography (SEC) with 2 detectors (differential refractive index (RI)
and differential viscometer). This method permitted to characterize/distinguish
commercial UF resins produced with different F/U molar ratios and to monitor the
molecular weight and molecular weight distribution with ageing.
An HPLC method was additionally used to evaluate the fraction of unreacted urea,
monomethylolurea and dimethylolurea present in commercial UF resins and
measure the evolution of these three compounds with ageing.
3.1 Introduction
The complex physics and chemistry of UF resins has been the subject of several
studies. These works have yielded further knowledge regarding these systems, but
still many issues remain concerning their structure as well as the kinetics and
mechanisms of their formation. The variety of randomly linked structural elements
1 J. M. Ferra, J. Martins, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho, Journal of Adhesion
Science and technology 24, 1535 (2010).
Chapter 3
70
such as methylene bridges, ether bridges, methylol and amide groups, and possible
cyclic derivatives makes their analysis a tough challenge. Moreover, these highly
reactive chemical systems have tendency to change during the preparation for the
analysis, or during analysis itself. The difficulty to find suitable solvents for these
resins is an additional problem [1]. Fortunately, the availability of modern
spectroscopic and chromatographic methods has led to considerable
improvements in the characterization of these resins: more specifically, 13C NMR
[2] and FTIR [3] for the structure and GPC/SEC [4-13] and even more recently
MALDI-TOF-MS [14] for the determination of the detailed chemical constitution,
although the true molecular weight distribution (MWD) remains elusive. More
recently, Minopoulou et al. [15] explored the capabilities of FT-NIR spectroscopy
[15-16] for on-line monitoring of the amino resin synthesis. The cured system has
been investigated by solid state 13C CP MAS NMR [17], FTIR [18] and Raman
spectroscopy [19]. The reaction kinetics has been studied experimentally [20] and
theoretically [19, 21-22].
In this work, the characterization of UF resins is focused on the determination of
the molecular weight distribution (MWD). The mechanical and bonding properties
of an adhesive are strongly dependent on its MWD [23-24]. This can be done by
GPC/SEC, but the low solubility of the colloidal fraction of these resins introduces
unique features in the chromatograms that must be taken into account [25].
GPC/SEC is a controlled separation technique in which molecules are separated on
the basis of their hydrodynamic molecular volume or size [26]. With proper column
calibration or using molecular weight-sensitive detectors, such as light scattering,
viscosimetry, or mass spectrometry, the MWD and the statistical molecular weight
averages can be readily obtained. In their review, Barth et al. [26] mentioned that
the GPC/SEC is the premier technique to evaluate these properties for both
synthetic polymers and biopolymers.
Effect of ageing on UF resins
71
The main problem using GPC/SEC is the choice of the proper solvent and mobile
phase to ensure complete resin solubility. It is necessary to use dimethylformamide
(DMF) or even dimethylsulfoxide (DMSO) to dissolve the higher molecular mass
fractions. Salts such as lithium chloride (LiCl) or lithium bromide (LiBr) are often
used to increase the solvent polarity and consequent polymer solubility, thus
minimizing the formation of aggregates. Another problem is related to the complex
nature of the polymer present in UF resins, where linear and branched fractions
coexist. The calibration standards can not represent this accurately, leading to
inaccuracies in the measured distribution of molecular masses. In contrast to
methods such as the nowadays seldom used analytical ultracentrifugation, no exact
molecular mass distribution can be extracted from chromatographic traces without
several assumptions. But even for linear fractions of UF polymers, as there are no
commercial standards for molecular mass calibration (UF compounds with a single
molar mass and molecular structure), oligomers would have to be synthesized in
the laboratory. GPC/SEC together with light scattering detection, which should
avoid the need for external calibration, is not a viable solution, as only the high
molar masses are detected by the light scattering sensor [1] and for this reason, the
calculation of averages such as weight and number molar masses is still much
affected by the absence of a reliable calibration.
Some partial success has been claimed for the use of GPC/SEC in the analysis of UF
polymers in previous research works. For instance, Dankelman et al. [5] have
reported that GPC/SEC can estimate the ratio of low to high molecular mass
components as well as the amounts of some oligomers. Billiani et al. [4] used this
technique to characterize UF resins synthesized with different degrees of
condensation. They found that measured average molecular mass increased with
the duration of the condensation steps, from a few thousand up to more than 100
kDa.
In the present work, preliminary studies have pointed out that the Right Angle
Chapter 3
72
Laser Light Scattering (RALLS) signal is too weak in the low to moderate molecular
weight fractions of the chromatograms obtained with UF resins. Additionally,
preliminary tests with polystyrene standards indicated that this RALLS system was
not able to detect molecular weights below about 7000 g/mol. This implies the
need for using a traditional, universal calibration technique, with two detectors:
differential refractive index and differential viscometer. The information provided
by the RALLS detector was used only in a qualitative way.
The GPC/SEC technique is useful for the characterization of MWD, but does not
give complete information about the composition of the low molecular weight
species present. For this purpose, HPLC can be effectively used for identifying low
molecular weight components in UF resins [27-28], and may contribute to a deeper
understanding of UF chemistry.
Grunwald [29] mentioned the combination of HPLC with GPC techniques as a
relevant area for future R&D on UF resins. In our present work HPLC was used
successfully for determining urea, monomethylolurea and dimethylolurea in
different resins.
3.2 Materials and Methods
3.2.1 Resins preparation
All resins characterized in this work were produced according to the alkaline-acid
process, which consists basically of three steps: methylolation under alkaline
conditions, condensation under acidic conditions, and neutralization and addition
of the so-called final urea or last urea.
Samples of UF-R5 and UF-R2 were supplied by EuroResinas-Indústrias Químicas
S.A., Portugal, while sample of UF-Exp17 was prepared in our laboratory according
to procedure described elsewhere [30]. Table 3.1 shows the technical data
collected for these three resins. All of them are UF resins in water solution with low
Effect of ageing on UF resins
73
amounts of melamine and hexamine. The main differences in the synthesis are the
duration of the condensation step – leading to different kinds of polymers formed -
and the final amount of urea added - leading to different final F/U ratios. Resin UF-
R2 has the longest condensation step and the largest amount of final urea, while
resin UF-R5 has the shortest condensation step and the lowest amount of final
urea. Resin UF-Exp17 has a sequential addition of urea during the condensation
step and the preparation procedure is completely described by Ferra et al. [30].
Table 3.1. Technical data of UF-resins
Resin Molar
ratio F/U Solids
content1 (%)
Gel time2
(s) pH range
(25 °C) Viscosity
3
(mPa·s)
UF-R5 1.30 63±1 35-55 7.5-8.5 150-350 UF-R2 1.00 64±1 40-100 8.0-9.5 150-300 UF-Exp17 1.12 64±1 40-100 8.0-9.5 150-300
1105 °C, 3 h
2Gel time at 100 °C with 3 wt % of NH4Cl (20 wt % solution)
3Brookfield viscometer at 25 °C
Some commercial resins from several major European producers were also studied.
The principal characteristics are presented in Table 3.2.
Table 3.2. Technical data on UF-resins used from different producers
Resin Molar
ratio F/U Solids
content1 (%)
Gel time2
(s) pH value (25 °C)
Viscosity3
(mPa·s)
UF-A 1.03 67.4 56 8.25 220 UF-B 1.12 63.6 64 8.78 210 UF-C 1.11 69.0 54 8.33 310 UF-D 1.15 68.1 43 8.40 400 UF-E n.a. 64.0 57 8.30 258
1105 °C, 3 h
2Gel time at 100 °C with 3 wt % of NH4Cl (20 wt % solution)
3Brookfield viscometer at 25 °C
3.2.2 GPC/SEC analysis
The main instrument used was a Gilson HPLC system equipped with a Gilson
Differential RI detector and Viscotek Dual Detector (differential viscometer and a
Chapter 3
74
light scattering detector RALLS). A Rheodyne 7125 injector with a 20 µL loop was
used for injection. The column used was a Waters Styragel HR1 5µm column. DMF
was used as the mobile phase. The column was conditioned at 60 °C using an
external oven and the flow rate was 1 mL·min-1. The universal calibration was done
using poly (ethylene glycol) standards from Polymer Laboratories Ldd, UK, with
molecular weight between 106 – 12140. The RALLS detector was not used for the
determination of molecular weight, because of the absence of response of the
detector to lower molecular weights. However, the RALLS signals were analysed
qualitatively.
The samples for GPC/SEC analysis were prepared by dissolving the resin in DMSO,
then vigorously stirring and filtering through a 0.45 µm Nylon syringe filter. The
addition of LiCl and the use of an ultrasonic bath in the preparation of the samples
were also tested, but the differences found in chromatograms were slight or
nonexistent.
3.2.3 HPLC analysis
A JASCO HPLC system equipped with a JASCO Differential RI detector and a
Rheodyne 7725i injector with a 100 µL loop was used. The column used was a
Waters Spherisorb silica column. A mixture of acetonitrile and water (90/10) was
used as the mobile phase. The column was conditioned at 30 ºC using an external
oven and the flow rate was 1.5 mL·min-1.
The samples for HPLC analysis were prepared by dissolving the resin in 1 mL of
DMF. After vigorous agitation for 1 min, it was diluted in 2 mL of mobile phase.
When the mobile phase was added, flocculation occurred and the sample was
allowed to rest for 10 min. The supernatant was finally withdrawn with a
micropipette.
Effect of ageing on UF resins
75
3.3 Results and discussion
3.3.1 Characterization of UF resins
Determination of molecular weight distribution
In Figure 3.1 one can see the GPC/SEC chromatograms of resins UF-R5 and UF-R2.
In both cases, at least two samples were prepared and analysed in order to verify
the reproducibility of the results.
Figure 3.1a shows the normalized weight fraction (Wt Fr) for resins UF-R5 and UF-
R2 after 5 days. The general features of the chromatograms are similar to those
found in the literature on UF resins. Three zones can be identified in the
chromatograms, as discusses in a previous work [25]. Zone I (elution volume
between 8 and 9 ml) corresponds to the low molecular weight species. Zone II
(elution volume between 5.8 and 8 ml) corresponds to intermediate molecular
weight species, with molecular weights ranging from about tens of thousands Da to
about 600 Da. Zone III (elution volume below 5.8 ml) would correspond to species
with quite high molecular weights, eluting before the exclusion limit of the
GPC/SEC column. It has been suggested that these are molecular aggregates and
not individual polymer molecules [9, 31]. These aggregates would be insoluble in
the original aqueous medium, probably forming larger colloidal structures which
become partially disaggregated in the DMSO solvent.
Since the chromatograms may reflect the presence of molecular aggregates, a
straightforward computation of the average molecular weights would be
misleading. Two different approaches were therefore followed to quantitatively
represent the chromatographic data. On the one hand, assuming that zone III
corresponds essentially to insoluble material, molecular weights were computed by
neglecting this portion of the chromatograms. On the other hand, the two
following parameters were introduced in order to complement the description of
the particular features of these chromatograms:
Chapter 3
76
amchromatogr of areatotal
I Zone of areaf1 = (3.1)
III Zone II Zone of areas
III Zone of areaf2 += (3.2)
f1 reflects the amount of low molecular weight species in the sample, while f2
indicates the fraction of high molecular weight species, probably in the form of
molecular aggregates in the polymerized material.
The main difference between the chromatograms for resins UF-R5 and UF-R2
(Figure 3.1a) is observed in the zone of low molecular weights, which probably
originated from the larger amount of the urea added in the last step of the reaction
for resin UF-R2. In addition, resin UF-R2 shows a more pronounced tail for low
elution volumes, corresponding to zone III. This probably reflects the higher
condensation state of this resin, which induces higher molecular aggregation [32].
Other than this, the two chromatograms are generically very similar. However, as
will be discussed below, ageing will introduce more pronounced differences.
The RALLS response (Figure 3.1b) gives qualitative information on the insoluble
particles present in solution [32]. One can see that the RALLS chromatograms are
similar for the two resins, but the trace for resin UF-R2 is more intense at lower
elution volumes, once again indicating a more significant aggregation.
Effect of ageing on UF resins
77
Vretention/mL
3 4 5 6 7 8 9
Norm
alized W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
1.2
log Mw
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
UF-R5
UF-R2
Vretention/mL
3 4 5 6 7 8 9
RALLS response/mV
110
120
130
140
150
UF-R5
UF-R2
(b)(a)
Figure 3.1. Chromatograms for UF-R5 and UF-R2 diluted 3 % in DMSO and stored for 5 days at 25 °C). a) normalized weight fraction (Wt Fr), and b) RALLS response.
The quantitative data shown in Table 3.3 confirm the previous analysis: resin UF-R2
presents a higher fraction of lower molecular weight species (f1) as well a higher
fraction of insoluble aggregates (f2). Resin UF-R2 is the most condensed of the
resins studied in this work, but this feature is not evidenced from the molecular
weight results, because of the larger addition of urea in the last step of the
reaction. Nevertheless, this resin shows the highest value of parameter f2, which is
indeed relatable to the highest degree of condensation.
Table 3.3. Values of Mn, Mw, polydispersity (Mw/Mn), and parameters f1 and f2, obtained by SEC for UF-R5 and UF-R2 stored for 5 days at 25 °C
Resin Mn Mw Mw/Mn f1 f2
UF-R5 3.77×102 3.59×10
3 9.5 0.356 0.171
UF-R2 2.90×102 3.44×10
3 11.9 0.410 0.193
Determination of the fractions of urea and methylolureas
Figure 3.2 presents a typical chromatogram obtained for a UF resin in HPLC. The
three first peaks correspond to urea, monomethylolurea and dimethylolurea,
respectively. This was confirmed by injection of the isolated compounds. The other
Chapter 3
78
peaks in the chromatogram correspond to oligomeric species.
time/min
4 6 8 10 12 14
RI response/mV
0
100
200
300
400
500
600
UF-R5
Figure 3.2. Chromatogram obtained for resin UF-R5.
Figure 3.3 shows the distribution of the urea, monomethylolurea and
dimethylolurea present in solution for the two resins stored at 25 °C for 5 days
after synthesis. UF-R2 has a much larger fraction of unreacted urea than resin UF-
R5. The last urea was added in order to react with the free formaldehyde present,
but as the added amount was large, most of the urea remained unreacted in the
final resin. However, this unreacted urea may play another role, since it may form a
solvation layer surrounding the colloidal aggregates surface, contributing to its
stabilization against agglomeration [25].
Effect of ageing on UF resins
79
Peaks
U MMU DMU
Peak area/total area
0
10
20
30
40
50
60
UF-R5
UF-R2
Figure 3.3. Peak areas normalized by total chromatogram area for UF-R5 and UF-R2 stored at 25 °C.
Analysis of commercial UF resins
Five commercial UF resins from different European producers were analysed.
Figure 3.4a shows the normalized Wt Fr obtained by GPC/SEC. One can see that UF-
D presents a distinctly larger fraction of insoluble aggregates (Zone III) and a higher
fraction of oligomers in the elution volume range 7.8-8.2 mL. On the other hand,
resin UF-B has the lowest fraction of insoluble aggregates and a larger fraction of
polymer with moderate molecular weight. Resins UF-A, UF-C and UF-E present
similar chromatograms, but some differences in the three zones of the
chromatograms were found, namely the large amount of polymer with low
molecular weight present in UF-A and a large amount of insoluble aggregates
existing in UF-E.
The chromatographic trace from the RALLS detector shown in Figure 3.4b agrees
qualitatively with the previous analysis concerning the presence of insoluble
aggregates in the different resins.
Chapter 3
80
Vretention/mL
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
1.2
UF-A
UF-B
UF-C
UF-D
UF-E
Vretention/mL
3 4 5 6 7 8 9
RALLS response/mV
500
1000
1500
2000
2500
UF-A
UF-B
UF-C
UF-D
UF-E
(b)(a)
Figure 3.4. Chromatograms for five UF resins from different manufactures in Europe. a) normalized weight fraction, and b) RALLS response.
The distributions of urea and methylolureas present in the five commercial resins
are shown in Figure 3.5. The main difference among the resins is the fraction of
unreacted urea. In particular, the fraction of urea in resin UF-D is approximately
half of the value for the other resins. The final percentages of urea and
methylolureas as shown above are related to the amount of last urea added and
the free formaldehyde present in the final condensation step.
The results obtained by GPC/SEC and HPLC indicate that each producer had likely
used different processes for the production of UF resins.
Effect of ageing on UF resins
81
Peaks
U MMU DMU
Peak area/total area
0
10
20
30
40
50
UF-A
UF-B
UF-C
UF-D
UF-E
Figure 3.5. Ratios of peak areas / total area of urea (U), monomethylolurea (MMU) and dimethylolurea (DMU) for five UF resins from different producers.
3.3.2 Monitoring the ageing of UF resins
Two UF resins were analysed after ageing, using GPC/SEC and HPLC techniques. The
pH and viscosity of resins were also monitored.
Figure 3.6 depicts the evolution of viscosity for resins UF-R2 and UF-R5 during
storage at 25 °C. The initial slight decrease in viscosity is related to the migration of
hydroxymethyl groups (methylolureas) from the polymeric UF resin components to
the last urea as reported by Kim [33]. Resin UF-R2 gels faster than UF-R5, but both
are stable up to 30 days, which is the normal specification for UF resins. The
storage time limit for UF-R2 is about 40 days, while for UF-R5 it is about 60 days.
The higher degree of condensation of the resin UF-R2 can explain this behaviour.
Chapter 3
82
storage time/days
0 20 40 60 80
Viscosity/cP
0
200
400
600
800
1000 UF-R5
UF-R2
Figure 3.6. Change of viscosity of UF-R5 and UF-R2 with storage time at 25 °C.
UF-R5 and UF-R2 resins were monitored at six different ageing periods (5, 12, 19,
32, 50 and 53 days). Figure 3.7 and Figure 3.8 show the chromatograms for some
selected storage times. Note that no significant changes in the molecular weight
distributions were observed for the first two storage times (5 and 12 days). This is
consistent with the stable viscosity measurements for the first 20 days. Kim et al.
[34] reported similar results by monitoring the storage of UF resins using 13C NMR.
They showed that the degree of polymerization remained stable for about 15 days
but then increased rapidly until 30 days and remained constant afterwards until
gelling.
On analysing the chromatograms for 4 selected storage times, we find that urea
and methylolureas (peak at 8.5 mL) decrease with ageing whereas the peaks at 8.2
mL and 7.8 mL increase. This suggests that urea and methylolureas react during
storage to produce a polymer with a narrow range of molecular mass / size (peak at
8.2 mL), which, in turn, reacts to produce polymer eluting in the vicinity of 7.8 mL,
visible as a broad peak. Globally, there is a decrease in Zone I of the chromatogram,
Effect of ageing on UF resins
83
yielding an increase in moderate to high molecular weight polymer (Zone II) and in
molecular aggregates (Zone III). Interestingly, a well-defined separation between
Zones II and III becomes apparent (at an elution volume of about 6 mL), which was
also visible in the commercial resins (Figure 3.4a). This might be associated with the
process of agglomeration of smaller aggregates into larger particles, shifting
towards the left portion of the chromatogram.
RALLS responses show a sharp increase in insoluble aggregates between 19 and 32
days for resin UF-R2 and between 32 and 53 days for resin UF-R5. This is related to
the viscosity evolution previously measured for both resins (Figure 3.1): the pre-
gelling increase in viscosity for each resin is associated with a significant formation
of insoluble molecular aggregates or, as mentioned above, to agglomeration of
existent aggregates into larger particles. Zanetti and Pizzi [35] and Despres and Pizzi
[31] reported that the continuing formation of colloidal structures followed by the
formation of “superaggregates” (globular masses) were the normal steps for
physical gelation of MUF and UF resins.
Vretention/mL
3 4 5 6 7 8 9
Norm
alized W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
1.2
5 days
19 days
32 days
53 days
Vretention/mL
3 4 5 6 7 8 9
RALLS response/m
V
500
1000
1500
2000
2500
5 days
19 days
32 days
53 days
(b)(a)
Figure 3.7. Chromatograms for UF-R2 diluted 3 % in DMSO, for different storage periods at 25 °C. a) normalized weight fraction, and b) RALLS response.
Chapter 3
84
Vretention/mL
3 4 5 6 7 8 9
Norm
alized W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
5 days
19 days
32 days
53 days
Vretention/mL
3 4 5 6 7 8 9
RALLS response/m
V
500
1000
1500
2000
2500
5 days
19 days
32 days
53 days
(b)(a)
Figure 3.8. Chromatograms for UF-R5 diluted 3 % in DMSO, for different storage periods at 25 °C. a) normalized weight fraction, and b) RALLS response.
From the data in Table 3.4 for resins UF-R2 and UF-R5 at different ages, one can see
that the value of polydispersity decreases with time, due to the condensation of
the low molecular weight species to form polymers with moderate and high
molecular weights. This solubilized polymer might then form insoluble molecular
aggregates but the portion of the chromatogram corresponding to the insoluble
molecular aggregates is not included in the molecular weight calculation, as
mentioned above. The decrease of the low molecular weight fraction (f1) and the
increase of insoluble aggregates fraction (f2) with ageing corroborates the idea that
condensation progresses with storage time, consuming urea and oligomers. The
decrease in f2 observed for resin UF-R5 after 50 days is related to the large increase
in fraction of the polymer with moderate molecular weight (zone II).
It is also interesting to look at the RALLS responses obtained for resins UF-R2 and
UF-R5 with ageing. Figure 3.7b and Figure 3.8b show a peak located roughly in the
region corresponding to zone III, assigned to molecular aggregates present in the
samples. When resins were “fresh” the peak magnitude was very low, but with
ageing it increased sharply for both resins. This seems to indicate that the ageing
Effect of ageing on UF resins
85
process produces a significantly higher concentration of aggregated material. These
aggregates might actually be agglomerated into larger particles in the resin.
Table 3.4. Values of Mn, Mw, polydispersity (Mw/Mn), and parameters f1 and f2, obtained by GPC/SEC for UF-R2 and UF-R5 stored for different days at 25 °C
Storage period Mn Mw Mw/Mn f1 f2
UF-R2
5 2.90×102 3.44×10
3 11.9 0.410 0.193
32 3.79×102 2.50×10
3 6.6 0.315 0.201
50 4.14×102 2.03×10
3 4.9 0.303 0.217
UF-R5
5 3.77×102 3.59×10
3 9.5 0.356 0.171
32 4.55×102 2.30×10
3 6.6 0.268 0.190
50 5.10×102 2.75×10
3 5.4 0.225 0.185
Figure 3.9 and Figure 3.10 show the evolution with ageing of the fractions of urea,
monomethylolurea, dimethylolurea and three more oligomeric species present in
the HPLC chromatograms, for resins UF-R2 and UF-R5, respectively. According to
the results described by Ludlam et al. [28], who used the same analysis conditions
(silica columns with NH2 groups, mobile phase and sample preparation) for the
identification of oligomeric species present in UF resins, the three unknown peaks
can be identified as monomethylolmethylenediurea,
monomethyloloxymethylenediurea, and dimethylolmethylenediurea, respectively.
In both cases, the fraction of urea in the solution decreases significantly up to 30
days. It ends up becoming stable as the formaldehyde present in the solution is
consumed. It is interesting to note that the evolution of dimethylolurea is different
for the two resins. It remains almost constant for resin UF-R2 but it goes through a
maximum at about 30 days for resin UF-R5. This behaviour is related to the
existence of free formaldehyde in the solution for resin UF-R5, which reacts with
urea and mostly with monomethylolurea (present in excess at the solution)
forming dimethylolurea. These results indicate that the polymerization reactions
between free formaldehyde, urea and methylolureas continue during the storage
Chapter 3
86
of the resin. Similar observations were reported by Kim et al. [33], which reported
that the amount of monomethylolurea during storage could increase or decrease
depending on the amount of free formaldehyde present in the solution.
Peaks
U MMU DMU Peak 4 Peak 5 Peak 6
Peak area/total area
0
10
20
30
40
50
60
5 days
12 days
32 days
49 days
Figure 3.9. Evolution of the ratios of peak areas / total area of the urea (U), monomethylolurea (MMU), dimethylolurea (DMU) and three other oligomeric species for UF-R2 stored for various periods at 25 °C.
Peaks
U MMU DMU Peak 4 Peak 5 Peak 6
Peak area/total area
0
10
20
30
40
50
5 days
12 days
32 days
49 days
73 days
Figure 3.10. Evolution of the ratios of peak areas / total area of the urea (U), monomethylolurea (MMU), dimethylolurea (DMU) and three other oligomeric species for UF-R5 stored for various periods at 25 °C.
Effect of ageing on UF resins
87
3.3.3 Determination of water tolerance
UF resins are colloidal suspensions that tend to flocculate as they are diluted in
water [25]. Figure 3.11 shows the GPC/SEC chromatograms obtained for the
original resin UF-R5 after 5 days storage at 25 °C and for the resin flocculated with
a large excess of water. In this case the supernatant was collected after
sedimentation of the precipitate and analysed. The corresponding chromatogram
shows similar concentrations of moderate and low weight molecules (elution
volumes between 7.0 to 9.0 mL), but a higher concentration of the zone that
corresponds to high/moderate molecular weight (elution volumes between 5.8 to
7.0 mL) and a lower concentration of insoluble molecular aggregates (Zone III).
Apparently, the transfer of molecules from the high to the intermediate molecular
weight zone is related to disaggregation of the larger particles due to the dilution.
It can also be seen from Figure 3.11 that insoluble aggregates originally
corresponding to zone III form aggregates of larger dimensions after flocculation,
thus eluting earlier. These aggregates only are detected by the RALLS sensor due to
their low concentration. However, it is necessary to use some caution in analysing
the results from RALLS detector because it saturates in the zone between 3.5 to 4.5
mL of the chromatogram. A similar behaviour was observed for other UF resin,
namely UF-R2.
These results demonstrate that the GPC/SEC is an interesting technique to evaluate
the water tolerance. The evaluation of water tolerance using the common method,
which consists in the addition of small amounts of water until the resin flocculation
occurs, is very difficult and inaccurate. A high value of water tolerance confers good
wash down properties to the product and allows easy cleaning of the apparatus
used for production and storage of UF resins.
Chapter 3
88
Vretention/mL3 4 5 6 7 8 9
Norm
alized W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
RALLS response/mV
500
1000
1500
2000
2500
Resin
flocculated in water
Figure 3.11. Chromatograms for UF-R5 aged for 5 days, diluted in DMSO and very diluted (flocculated) in water.
3.4 Conclusions
Different UF resins were characterized by GPC/SEC and HPLC techniques.
The GPC/SEC analysis encompasses information on the MWD of the soluble
polymer as well as on insoluble molecular aggregates that constituted the original
dispersed phase and have not been completely dissolved in the DMSO solvent. The
information obtained by GPC/SEC is useful for characterization of the resins and
allows to distinguish resins obtained from different production processes. The
information from the RALLS detector complements qualitatively the information on
the insoluble material.
GPC/SEC and HPLC methods permitted to verify that UF resins produced by the
European companies had distinct characteristics (MWD and relative amounts of U,
MMU and DMU). These results suggest that each producer uses a particular
process for the production of UF resin.
The GPS/SEC analyses of the resins at different ageing periods indicated that both
Effect of ageing on UF resins
89
polymer condensation and aggregation/agglomeration proceed during storage.
This technique permits to monitor accurately the ageing of UF resins.
3.5 References
[1] M. Dunky, Analysis of Formaldehyde Condensation Resins for the Wood Based Panels Industry: Status and new challenges. In 1st
European Panels Products
Symposium, Hague, Loxton, Bolton and Mott, Eds. Llandudno, North Wales, UK, 1997; p 217.
[2] R. M. Rammon, W. E. Johns, J. Magnuson and A. K. Dunker, Journal of Adhesion 19, 115 (1986).
[3] S. S. Jada, Journal of Macromolecular Science-Chemistry A27, 361 (1990). [4] J. Billiani, K. Lederer and M. Dunky, Angewandte Makromolekulare Chemie 180,
199 (1990). [5] W. Dankelman, J. M. H. Daemen, A. J. J. D. Breet, J. L. Mulder, W. G. B.
Huysmans and J. D. Wit, Angewandte Makromolekulare Chemie 54, 187 (1976).
[6] M. Dunky and K. Lederer, Angewandte Makromolekulare Chemie 102, 199 (1982).
[7] M. Dunky, K. Lederer and E. Zimmer, Holzforschung Und Holzverwertung 33, 61 (1981).
[8] D. Grunwald, Kombinierte analytische Untersuchungen von Klebstoffen für Holzwerkstoffe, Mensch & Buch Verlag, Berlin, Germany (2002).
[9] T. Hlaing, A. Gilbert and C. Booth, British Polymer Journal 18, 345 (1986). [10] C. Y. Hse, Z. H. Xia and B. Tomita, Holzforschung 48, 527 (1994). [11] S. Katuscak, M. Tomas and O. Schiessl, Journal of Applied Polymer Science 26,
381 (1981). [12] P. R. Ludlam and J. G. King, Journal of Applied Polymer Science 29, 3863 (1984). [13] G. Zeppenfeld and D. Grunwald, Klebstoffe in der Holz- und Möbelindustrie,
Drw Verlag Weinbrenner (2005). [14] H. Mandal and A. S. Hay, Polymer 38, 6267 (1997). [15] E. Minopoulou, E. Dessipri, G. D. Chryssikos, V. Gionis, A. Paipetis and C.
Panayiotou, International Journal of Adhesion and Adhesives 23, 473 (2003).
[16] E. Dessipri, E. Minopoulou, G. D. Chryssikos, V. Gionis, A. Paipetis and C. Panayiotou, European Polymer Journal 39, 1533 (2003).
[17] S. Tohmura, C. Y. Hse and M. Higuchi, Journal of Wood Science 46, 303 (2000). [18] G. E. Myers, Journal of Applied Polymer Science 26, 747 (1981). [19] L. M. H. Carvalho, M. R. P. F. N. Costa and C. A. V. Costa, Journal of Applied
Polymer Science 102, 5977 (2006). [20] A. F. Price, A. R. Cooper and A. S. Meskin, Journal of Applied Polymer Science
Chapter 3
90
25, 2597 (1980). [21] M. R. P. F. N. Costa and R. Bachmann, Polycondensation. In Handbook of
Polymer Reaction Engineering, T. Meyer and J. Keurentjes, Eds. Wiley-VCH: 2005; Vol. 1.
[22] A. Kumar and A. Sood, Journal of Applied Polymer Science 40, 1473 (1990). [23] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998). [24] A. Pizzi and K. L. Mittal, Handbook of Adhesive Technology, Marcel Dekker,
New York (2003). [25] J. M. M. Ferra, A. M. Mendes, M. R. N. Costa, L. H. Carvalho and F. D.
Magalhães, Journal of Applied Polymer Science 118, 1956 (2010). [26] H. G. Barth, B. E. Boyes and C. Jackson, Analytical Chemistry 70, 251 (1998). [27] K. Kumlin and R. Simonson, Angewandte Makromolekulare Chemie 68, 175
(1978). [28] P. R. Ludlam, J. G. King and R. M. Anderson, Analyst 111, 1265 (1986). [29] D. Grunwald, Wood Adhesives. In COST Action E13: Wood Adhesion and Glued
Products, State of the art Report, M. Dunky, Ed. 2001. [30] J. M. Ferra, J. Pereira, B. Pinto, F. Silva, J. Martins, A. Mendes, M. R. Costa, L.
Carvalho and F. D. Magalhães, Optimization of the Synthesis of Urea-Formaldehyde resins by Response Surface Methodology. In International
Panel Products Symposium, M. Spear, Ed. Espoo, Finland, 2008; pp 97. [31] A. Despres and A. Pizzi, Journal of Applied Polymer Science 100, 1406 (2006). [32] J. M. Ferra, J. Martins, A. Mendes, M. R. Costa, L. Carvalho and F. D.
Magalhães, Characterization of the Two Different Phases in Urea-Formaldehyde Resins. In ECOWOOD, J. Caldeira, Ed. Porto, Portugal, 2008; pp 17.
[33] M. G. Kim, Journal of Applied Polymer Science 80, 2800 (2001). [34] M. G. Kim, H. Wan, B. Y. No and W. L. Nieh, Journal of Applied Polymer Science
82, 1155 (2001). [35] M. Zanetti and A. Pizzi, Journal of Applied Polymer Science 91, 2690 (2004).
CHAPTER 4
Optimization of the synthesis of UF resins using response surface methodology
93
4 Optimization of the synthesis of UF resins using response
surface methodology1
Abstract
In the near future, companies will face the need to produce low formaldehyde
emission resins, i.e., not above the emission level of natural wood. However, for
producing this new generation of urea-formaldehyde resins, it is necessary to
optimize the synthesis process.
This work describes an optimization procedure for UF resin synthesis, following an
alkaline-acid process, focusing on the conditions of the condensation step. A design
of experiments methodology was employed to optimize the 3 selected factors
(number of urea additions, time span between urea additions, and pH of
condensation reaction), in order to produce particleboards with maximum internal
bond strength and minimum formaldehyde release.
The condensation pH played a significant role in increasing the Internal Bond (IB)
strength and reducing the Formaldehyde Emission (FE). The sequential addition of
urea also has a noticeable influence on resin performance. Optimum conditions for
production of urea-formaldehyde resins have been proposed and tested by the
response surface methodology using the desirability function.
1 J. M. Ferra, P. C. Mena, J. Martins, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho, Journal of
Adhesion Science and Technology 24, 1455 (2010).
Chapter 4
94
4.1 Introduction
Nowadays, the main goal of urea-formaldehyde (UF) resin industry is to comply
with the standards requiring even lower formaldehyde release after curing. This
chemical has now been reclassified as carcinogen by different organizations [1] (see
Table 4.1). In California, the state Health and Safety Code mandates the California
Air Resources Board (CARB) to develop Airborne Toxic Control Measures (ATCM) to
protect public health from airborne carcinogens such as formaldehyde. The ATCM
proposed the reduction of formaldehyde emission from particleboard specifically,
in a first phase, until September 2009, from 0.30 ppm (specified in ASTM E 1333-
96) to 0.18 ppm, and in a second phase, until 2011-2012, to 0.09 ppm [2].
Table 4.1. Current classification of formaldehyde by some organizations
Organization Current classification
European Chemicals Bureau Category 3-R40 “limited evidence of a carcinogenic effect”
U.S. Environmental Protection Agency
Probable human carcinogen
International Agency for Research on Cancer
Group 1 - “there is sufficient evidence in humans & animals for the carcinogeneticity of formaldehyde”
Free formaldehyde is present in UF resins and hydrolytic degradation of UF resins
(reversibility of aminomethylene links) under moist and acidic conditions is known
to be responsible for formaldehyde emission from wood-based panels. So,
formaldehyde emission not only depends on synthesis conditions, but also on the
type of bonds in the cured resin [3].
Several strategies have been explored for reducing formaldehyde release. These
include the addition of formaldehyde scavengers directly to the resin or wood
particles, the treatment of final wood panels with scavengers or impermeable
coatings, and the development of improved resin formulations [4].
The main change in the synthesis processes has been the decrease of F/U molar
Optimization of the synthesis of UF resins using response surface methodology
95
ratio. However, this worsens the mechanical properties of the particleboards
formed and, moreover, increases the hardening time using the traditional
hardeners.
Different synthesis procedures for UF resins have been found in the literature [5-
11] which can be divided into two distinct processes:
i) The alkaline-acid process [3, 5-6, 8] is the most common. It involves
three steps, usually an initial alkaline methylolation followed by an acid
condensation and finally neutralization and addition of the last urea
fraction.
ii) The strongly-acid process was first described in the patents by
Williams[4, 11]. It consists of: (1) carrying out the condensation of urea
and formaldehyde under a highly acid environment and large excess of
formaldehyde; (2) continuing the reaction in alkaline medium after
additional urea is added to attain a predetermined U/F molar ratio; (3)
carrying out the reaction under a low pH of about 5 to allow further
condensation until a desired viscosity is reached; and (4) neutralization of
the product and addition of the final urea amount.
Rammon [12] studied the effect of several variables on the synthesis of UF resins
by the alkaline-acid process. The results obtained indicated that increasing the pH
of the alkaline phase or using a longer acid condensation step leads to a decrease
in the amount of ether linkages formed. These linkages can undergo
rearrangement and lead to formaldehyde release.
The Williams process [4, 11] entails minimal energy consumption and involves
Chapter 4
96
relatively short (3-4 hours) reaction times. The reduced formaldehyde emission and
increased hydrolytic stability have been attributed to the predominance of the
more stable methylene linkages in the cured resin, unlike the alkaline-acid process,
leading to a large amount of methylene ether linkages in the cured resin. The main
problem of this process lies in the control of the highly acid condensation step, due
to its exothermic character. According to Hatjiissaak and Papadopoulou [7], this
process needs careful control, which is difficult to achieve at an industrial scale, to
prevent resin gelling in the reactor.
Williams [11] demonstrated that the modification of the resins with the addition of
cross-linking agents (e.g., trimethoxymethylmelamine) improves the values of
internal bond (IB) strength by 28 % for the uncatalyzed resin and 47 % for the
catalyzed formulation.
Pizzi and coworkers [13-15] reported that multiple additions of urea during resin
preparation increases the bond quality, especially for resins with low F/U molar
ratios. This helps the build-up of the fraction of polymeric material in the final
resin, at the expense of low oligomeric species, thus increasing the condensation
degree of the UF resin [14].
Kim and coworkers [16-20] reported on the migration of methylol groups to the
last urea added in alkaline medium, leading to the reaction with the free amide
groups which were formed in the final stage, producing monomeric methylolureas.
This decreases the final viscosity and increases the formaldehyde emission from
particleboards. Kim and coworkers [16-20] reported also that this migration
continued over a period of 1 month or longer at room temperature and became
relatively fast above 50 °C.
Recently, Kumar et al. [21] have optimized the second step of the strongly-acid
process (alkaline methylolation) for producing UF resins. The effect of the “number
of additions” and “time span between additions” of the second urea on IB strength
Optimization of the synthesis of UF resins using response surface methodology
97
and formaldehyde emission (FE) was analysed. The results showed that the
sequential addition of urea in the methylolation step plays a significant role in
reducing the formaldehyde emission and increasing the internal bond strength. It
was also reported that these observations were due to the conversion of urea and
the migration of the hydroxymethyl groups.
Pizzi and coworkers [14-15] and Kumar et al. [21] state that the best strategy for
obtaining low formaldehyde emission UF resins, with suitable bond strength to
wood, is to produce resins with excess monomeric non-methylolated and
methylolated urea species (which lowers formaldehyde emission and, up to a
certain extent, improves resin adhesion to the wood substrate). On the other hand,
the presence of a certain fraction of non-methylolated and particularly
methylolated polymeric species, ensures that cross-linking density and cohesion,
and hence yield an acceptable mechanical strength will be obtained.
The present work describes the optimization of an alkaline-acid process, mainly
focusing on the condensation step, considering 3 factors: the number of urea
additions, the time span between additions, and the pH. A response surface
methodology (RSM) with a central composite design (CCD) was used in order to
evaluate the effect of independent variables on the process performance and to
optimize the operating conditions (maximize the IB strength and minimize the FE).
Second-order models were fitted using the CCD results, which describe the effect
of the operating conditions on the process responses. These models were used for
interpolating predicted values for the experimental conditions and, therefore,
comparing them with the experimental ones. On the other hand, the models were
used to evaluate the effect of the critical operating conditions on the responses
and to obtain the operating conditions that maximized the objective function.
Chapter 4
98
4.2 Materials and Methods
4.2.1 Experimental design
Response surface methodology (RSM) is a combination of mathematical and
statistical tools which is effective for studying and modelling processes where
responses are dependent on several operating variables [22]. The model
parameters are estimated using the least squares method. In this work a CCD was
selected, which is the most used method for fitting second-order models.
Pizzi [15] reports that the most important synthesis parameters with an influence
on the properties of UF resins are: temperature, intermediate U/F molar ratios, and
pH of the condensation step.
In the present work, focussing on an alkaline-acid synthesis process, three factors
were analyzed while maintaining the global F/U molar ratio constant: a) total
number of urea additions during the condensation step, b) time span between
consecutive additions of urea during the condensation step, and c) pH of the
condensation step. The pH of the methylolation step was also initially considered,
but preliminary experiments, performed with a pH range between 8.6 and 9.4,
indicated only small modifications in the final product characteristics.
The levels of the three selected factors are given in Table 4.2. The pH values
selected were based on the range used in industrial production.
Table 4.2. Experimental levels of the three factors
Factors levels
−1 0 1
No. of U additions in the condensation step (A) 2 3 4 Time span between additions of U in the condensation step (B) 10 15 20 pH of the condensation step (C) 5.6 5.9 6.2
The properties selected for evaluating the resins produced were: internal bond
strength, formaldehyde content measured by the perforator method, and the
Optimization of the synthesis of UF resins using response surface methodology
99
fraction of insoluble aggregates (FIA) measured by Gel Permeation
Chromatography (GPC) / Size Exclusion Chromatography (SEC). However, for all the
resins produced, several other properties were measured, namely, gel time, final
pH, viscosity, solids content, and stability at 25 °C.
For generating the matrix of experiments a CCD was employed [22] with 2 central
points, which resulted in 16 experiments (Table 4.3). Even though the number of U
additions is a discrete variable, it was assumed as a continuous variable so that the
RSM approach could be used directly. JMP 5 software (SAS, Cary, NC, USA) was
used for generating the matrix of experiments and for analysing the results.
Table 4.3. Central composite design matrix of experiments generated by the DoE tool
Run Experimental values
A – No. add. U B – Time Span C – pH
1 4 10 5.6
2 2 10 6.2 3 3 15 5.9 4 4 20 5.6 5 3 10 5.9 6 4 15 5.9 7 3 15 6.2 8 3 15 5.9 9 2 20 6.2 10 2 15 5.9 11 4 20 6.2 12 3 15 5.6 13 2 10 5.6 14 4 10 6.2 15 3 20 5.9 16 2 20 5.6
Statistical analysis
The analysis of variance (ANOVA) was used to determine the adequacy of the
model to describe the observed data. The R2 statistic indicates the degree of
variability of the optimization parameters that are explained by the model.
Chapter 4
100
Three-dimensional response surface plots were generated for each quality
parameter. Calculation of optimal synthesis conditions for optimum IB strength and
FE was performed using a multiple response method designated as desirability [22-
24]. This optimization method incorporates desired values and priorities for each
variable.
The natural variables (A, B and C) are associated with coded variables (X1, X2 and X3)
which are dimensionless, according to [22]:
0.3
5.9CX;
5
15BX;
1
3AX
321
−=
−=
−= (4.1)
The quadratic equation for the variables is defined as follows:
∑∑++=≥i ij
jiijii0XXβXΣββY (4.2)
Y is the predicted response; β0 is a constant; βi are the linear coefficients; βii are the
squared coefficients; and βij with i≠j are the cross-product coefficients. The above
quadratic equation was used to build response surfaces for the dependent
variables.
4.2.2 Synthesis of UF resins
In the preparation of UF resins, industrial-grade raw materials were used, provided
by EuroResinas – Indústrias Químicas S.A., Portugal, namely, urea, 50 % formalin,
sodium hydroxide solution, and acetic acid solution. The synthesis was carried out
in a laboratory-scale 5 L glass reactor.
The reaction of urea and formaldehyde consists basically of a three-step process
(alkaline-acid process):
i) methylolation under alkaline conditions;
ii) condensation under acidic conditions;
Optimization of the synthesis of UF resins using response surface methodology
101
iii) neutralization and addition of the so-called final urea.
Methylolation step: after the required amount of urea was added to produce an
initial F/U molar ratio of 2.15, the reaction mixture was held at the final
temperature (T = 95 °C) for about 30 min under alkaline conditions (pH = 9.0). The
pH was afterwards adjusted to the pH of the condensation step using acetic acid.
Condensation step: the second urea was added sequentially to the previous
reaction mixture, varying the number of intermediate additions and the time
interval between additions. The F/U molar ratio after this step was 1.8. After the
desired viscosity was reached, the reaction was stopped by alkalinisation with
sodium hydroxide solution and cooling. The third (final) urea was then added at 60
°C, yielding a final F/U ratio of 1.12.
The UF resins were stored at 25 °C for three days and then were analysed and
incorporated into wood panels.
4.2.3 GPC/SEC analysis
A Gilson HPLC system equipped with a Gilson Differential RI detector and Viscotek's
Dual Detector (differential viscometer and a light scattering detector RALLS). A
Rheodyne 7125 injector with a 20 µL loop was used for sample introduction. The
column used was a Waters Styragel HR1 5 µm. Dimethylformamide (DMF) was used
as the mobile phase. The column was conditioned at 60 °C using an external oven
and the flow rate was 1 mL/min. The Universal Calibration was based on
poly(ethylene glycol) standards from Polymer Laboratories, Germany with
molecular weight between 106 - 12140. The RALLS detector was not used for the
tri-SEC molecular weight calculations owing to the very weak response at lower
molecular weights. However, the RALLS signals were qualitatively taken into
account.
The samples for GPC/SEC analysis were prepared by dissolving approximately 100
mg sample of resin in 3 mL of dimethylsulfoxide (DMSO), then stirring vigorously
Chapter 4
102
and filtering through a 0.45 µm Nylon syringe filter.
4.2.4 Preparation of laboratory-made particleboards
Wood particles were provided by a particleboard producer (Sonae Industria,
Oliveira do Hospital plant, Portugal), a standard mix of wood particles was used for
the face layer and core layer. The standard mix included about 30 % maritime pine
(Pinus Pinaster), 15 % eucalypt Eucalyptus Globulus, 25 % pine sawdust and 30 %
recycled wood. Wood particles with 4 % of moisture content were blended with
the resins and paraffin in a laboratory glue blender. The UF resins were applied at 8
% resin solids in both the face and core layers, based on the oven-dry weight of the
respective particles. The resin formation was more catalysed in the core layer (3 %
solids based on oven-dry weight of resin) than in the face layer (1 % solids based on
oven-dry weight of resin).
A three-layer particleboard was hand-formed in a metallic container with
dimensions 220 mm x 220 mm x 80 mm. The total percentages of board mass
were: 20 % for the upper face layer, 62 % for the core layer and 18 % for the
bottom face layer. Boards were then pressed in a laboratory scale hot-press,
controlled by a computer and equipped with a displacement sensor (LVDT), a load
cell, thermocouples and pressure transducers. The mat was pressed at 195 °C for
2.8 min to produce a board with a target density of 650 kg/m3 and with 17 mm
thickness.
After pressing, boards were conditioned for 3 days at normal conditions (20 °C, 65
% RH).
4.2.5 Particleboard Testing
Samples were then tested accordingly to the European standards and the following
physico-mechanical properties were evaluated: density (EN 323), moisture content
(EN 322) and IB strength (EN 319).
Optimization of the synthesis of UF resins using response surface methodology
103
The formaldehyde content of all samples was determined according to EN 120
(perforator method) and the formaldehyde emission was determined for some
samples, according to EN 717-2 (gas analysis method).
4.3 Results and Discussion
4.3.1 Characteristics and performance of the UF resins produced
In the resin synthesis, the condensation reaction step is stopped when the desired
viscosity is attained for the reaction mixture. Figure 4.1 shows the time necessary
for the condensation reaction step as a function of the pH in this step. There is a
strong influence of pH on the reaction rate. Interestingly, it was observed that the
other two factors, time span between urea additions and number of urea
additions, had negligence influence on the reaction time.
pH of condensation step
5.5 5.6 5.7 5.8 5.9 6.0 6.1 6.2 6.3
time /min
0
50
100
150
200
250
300
350
400
Figure 4.1. Condensation reaction time versus pH of the condensation step for the resins produced.
Chapter 4
104
Table 4.4 reports the characteristics measured for all the UF resins produced in this
work, namely, gel time, viscosity, solids content and final pH.
Table 4.4. Characteristics of UF resins produced
Run Gel time
1
(s) Viscosity
2
(mPa·s) Solids content
3
(%) pH
(25 °C)
1 67 190 63.92 9.05
2 77 170 63.86 8.66
3 94 170 64.36 8.73
4 115 200 64.04 8.69
5 66 240 63.59 9.10
6 69 260 64.22 8.18
7 60 130 64.73 9.40
8 93 160 64.99 8.63
9 74 130 64.30 8.02
10 80 150 63.86 9.26 11 80 170 65.35 9.54
12 68 130 63.37 9.18
13 71 130 63.32 8.87
14 58 130 65.02 9.16
15 62 190 64.69 8.90
16 80 210 63.52 9.30 1Gel time at 100 °C with 3 wt % of NH4Cl (20 wt % solution)
2Brookfield viscometer at 25 °
3120 °C, 3 h
All resins were analysed by GPC/SEC three days after synthesis, to evaluate the
fraction of insoluble molecular aggregates. Table 4.5 presents the resins properties
considered for the experimental design. Figure 4.2 shows a typical chromatogram
for a UF resin synthesised in this work. The peak with larger retention volume
corresponds to free urea, methylolureas and oligomers (MW < 400), the
intermediate zone of the chromatogram (5.8 – 8.2 mL) corresponds to polymer
with moderate molecular weight (400 < MW < 12140). The leftmost portion of the
chromatogram would correspond to polymer with unrealistic high molecular
weight (MW > 12140). This fraction may actually represent insoluble molecular
aggregates (colloids) smaller than 0.45 μm and not actually individualized
molecules, as suggested by Hlaing et al [25] and Despres and Pizzi [26]. The
Optimization of the synthesis of UF resins using response surface methodology
105
presence of these colloidal particles will influence the resin performance during the
bonding process [27]. The fraction of insoluble aggregates is computed here as the
ratio of the area of the chromatogram corresponding to elution volumes below 5.8
mL to the total area. This segmentation of UF resin chromatogram into three zones
has been previously suggested by different authors [25, 27-28].
Vretention/mL
3 4 5 6 7 8 9
Norm
alized RI response /mV
0.0
0.2
0.4
0.6
0.8
Figure 4.2. Normalized response of RI sensor for UF resin run 5 diluted 3 % in DMSO.
In Table 4.5, it is possible to observe quite distinct values of IB strength and FE,
indicating that the produced resins effectively have different properties, caused by
the changes in the operating conditions.
Note that for the particleboards produced with all resins, other physical-
mechanical properties were also measured, namely, density, moisture content, and
thickness swelling. All the results obtained comply with standard specifications for
particleboard (EN312).
Chapter 4
106
According to the standards specifications, the minimum acceptable value for IB
strength is 0.35 N·mm-2 (type P2 board, EN 312). Although for formaldehyde
content, the limit for class E1 is 8 mg/100 g oven-dry board, perforator value (EN
312; EN 13986), particleboard producers are compelled by customers to lower the
limits to 4 mg/100 g oven-dry board (EN 120). Some of the resins (runs 3, 7, 8 and
9) lead to excellent values of IB strength, while others lead to very low FE (runs 2
and 14), maintaining acceptable values of IB strength. In particular, one of the
resins (run 6) produced panels with very good properties, i.e., a high value of IB
strength (0.80 N·mm-2) and a low value of FE (4.5 mg/100 g oven-dry board).
Table 4.5. Experimental results for the three measured responses
Run IB strength
1
(N·mm-2
)
FE2
FIA3
Perforator method (mg/100 g dry board)
Gas analysis method
(mg/m2.h)
1 0.56 6.5 - 0.079 2 0.59 4.1 3.61 0.093 3 0.72 6.7 - 0.051 4 0.37 5.7 - 0.054 5 0.39 5.7 - 0.057 6 0.80 4.5 4.24 0.098 7 0.62 5.8 - 0.062 8 0.63 6.2 - 0.065 9 0.72 7.4 - 0.054 10 0.59 6.2 - 0.052 11 0.58 5.3 - 0.069 12 0.38 6.6 - 0.036 13 0.50 6.5 - 0.049 14 0.41 4.3 3.62 0.076 15 0.57 5.8 - 0.074 16 0.44 6.1 - 0.061
1IB strength – internal bond strength
2FE – formaldehyde emission
3FIA – fraction of insoluble aggregates
It is known that the release of formaldehyde from wood panels is caused by three
factors: a) residual formaldehyde trapped as gas in the board structure, b)
formaldehyde dissolved in the retained water, and c) hydrolysis of weakly bound
Optimization of the synthesis of UF resins using response surface methodology
107
formaldehyde, in the form of methylols, acetals and hemiacetals and, in more
severe cases, hydrolysis of methylene ether bridges (at high relative humidity) [3].
Figure 4.3 and Figure 4.4 show how the fraction of insoluble aggregates may be
related to IB strength and FE. Despite the scatter in the data, it seems acceptable to
establish a trend such that as FIA increases, IB strength increases and FE decreases.
The contribution of the dispersed insoluble phase to the strengthening of the
adhesive bond, by acting as a reactive reinforcing filler, has been previously
suggested [27, 29]. On the other hand, the association of this disperse phase to a
lower formaldehyde release, as suggested by Figure 4.4, can be interpreted in
terms of the presence of polymer with higher content in methylene linkages.
These results point a way to produce UF resins with improved performance, by
obtaining high fractions of insoluble aggregates combined with low molecular
weight polymer. Dunky [3] reports that higher molecular weight species improve
the cohesive strength, while the low molecular weight species contribute to the
wetting and penetration of the resin in the substrate.
IB strength (N.mm-2)
0.4 0.5 0.6 0.7 0.8
Fraction of insoluble aggregates
0.02
0.04
0.06
0.08
0.10
0.12
Figure 4.3. Relation between insoluble aggregates and internal bond strength. The dashed lines represent the 90 % confidence intervals.
Chapter 4
108
F content (mg/ 100 g dry board)
4 5 6 7
Fraction of insoluble aggregates
0.02
0.04
0.06
0.08
0.10
0.12
Figure 4.4. Relation between insoluble molecular aggregates and F content. The dashed lines represent the 90 % confidence intervals.
4.3.2 Model fitting
The application of RSM gave the regression equations shown below (Eqs. 4.3-4.5),
which are empirical relations between the three responses and the three test
variables. The analysis of variance (ANOVA) indicated that the second-order
polynomial model (Eq. 4.2) was adequate to represent the FE (p-value = 0.05 and R2
= 0.86). For IB strength (p-value = 0.25 and R2 = 0.73) the analytical method
employed for quantifying IB strength did not have enough resolution, within the
range considered for each factor, resulting in a smaller confidence level of the
fitted model. The parameters of the ANOVA indicated that the model was not
adequate to represent the FIA.
It must be remarked that in the cases where the error in Eqs. 4.3-4.5 was equal or
higher than the corresponding coefficient, the associated variable was not included
in the models, as usual [22].
Optimization of the synthesis of UF resins using response surface methodology
109
As can be seen in Figure 4.5, the values predicted by the second-order models for
response IB strength and FE agree reasonably well with the experimental data.
Figure 4.5. Experimental and calculated results of the responses considered. Y1 – internal bond strength and Y2 – formaldehyde emission.
After removing the negligible parameters from the original fitting polynomial
equations, one obtains:
2
3
2
2
2
1
323
0.06688)X0.093793(0.06688)X0.113793(0.06688)X0.1012069(
X0.038393)X0.06875(0.03434)X0.067(0.051411)0.6208621( strengthIB
±−±−±+
±+±+±= (4.3)
2
1
32213
21
0.333988)X0.546552(
X0.19173)X0.6875(X0.19173)X0.3375(0.171488)X0.45(
0.17488)X0.32(0.17488)X0.4(0.256739)6.0810345(FE
±−
±+±−±−
±+±−±=
(4.4)
2
31
31
0.01081)X0.011741(0.01081)X0.0142586(
0.00555)X0.0075(0.00555)X0.0067(0.008309)0.0598276(FIA
±−±+
±+±+±=2
(4.5)
where Xi are the coded variables (Eq. 4.1) for each factor: X1 is the number of urea
additions in the condensation step, X2 is the time span between successive urea
additions in the condensation step, and X3 is the pH of condensation step.
Considering just the first-order effects of each variable in Eq. 4.3, it is clear that the
main factor that affects IB strength is the pH of the condensation step. For the
Chapter 4
110
formaldehyde emission property (Eq. 4.4), the number of additions of urea and the
pH of condensation steps are the main factors, but the time span between urea
additions also plays a noticeable role.
4.3.3 Effect of all three factors on the measured responses
Figure 4.6 and Figure 4.7 depict the response surfaces, showing the effect of the
number of urea additions, the time span between consecutive urea additions, and
the pH of the condensation step on the IB strength and FE of the panels produced.
The polynomial expression in Eq. 4.4 was used to calculate the response surface
illustrated in Figure 4.6. It can be seen from Figure 4.6a that depending on the pH
of condensation, the time span between urea additions may have different effects
on the IB strength. For the lower condensation pH (5.6), increasing the time span
affects negatively the IB strength, whereas for the higher pH (6.2) the effect is
reversed. As the time span between urea additions increases, if the reaction rate is
sufficiently high, the urea might be depleted before the next addition takes place.
As we can see from Figure 4.6, for pH 5.6 the reaction is about 5 times faster than
for pH 6.2. Hence, there is a high probability that urea added too late during the
condensation step will not be totally consumed and will be carried on to the last
step. At pH = 5.9 there is a mixed effect on the IB strength value. By allowing a
higher conversion of urea in this reaction step, a higher fraction of polymer is
obtained in the final resin, in contrast to oligomeric species. This leads to an
improved bond quality, as previously discussed by Pizzi [15].
Figure 4.6b shows that the number of urea additions does not change the effect of
the other two factors on IB strength. Once again, it can be seen that each of the
two factors has a positive or negative effect on IB strength, depending on the value
of the other factor (cross-effects).
Optimization of the synthesis of
Figure 4.6. Response surface for internal bond strength as a function of: a) time span between urea additions and number of urea additions (for different pH values of condensation step), b) pH of condensation step and time span between urea additions (for different numbers of urea additions) and c) pH of condensation step and number of urea additions (for different time spans between urea additions).
The response surfaces illustrated in
on the formaldehyde emission (FE). It is particularly evident from
FE tends to decrease as pH is increased and the time span between additions is
decreased.
The lower emission at the higher pH might be
stable methylene linkages instead
condensation process. The former can rearrange to methylene bridges by splitting
Optimization of the synthesis of UF resins using response surface methodology
111
Response surface for internal bond strength as a function of: a) time span between urea additions and number of urea additions (for different pH values of
step), b) pH of condensation step and time span between urea additions (for numbers of urea additions) and c) pH of condensation step and number of urea
additions (for different time spans between urea additions).
ed in Figure 4.7 show the effects of the three factors
on the formaldehyde emission (FE). It is particularly evident from Figure 4.4b that
FE tends to decrease as pH is increased and the time span between additions is
The lower emission at the higher pH might be related to the formation of the more
stead of methylene ether linkages during the
condensation process. The former can rearrange to methylene bridges by splitting
Chapter 4
112
off formaldehyde [3].
The cross-effects discussed above justify
optimization of the synthesis of UF resin.
Figure 4.7. Response surface for formaldehyde emission as a function of: a) time span between urea additions and number of urea additions (for different pH values of condensation step), b) pH of condensation step andifferent numbers of urea additions) and c) pH of condensation step and number of urea additions (for different time spans between urea additions).
4.3.4 Optimization of Operating Conditions
The optimization consisted in the simultaneous maximization of the IB
value and minimization of the FE value. This was done based on the
justify the use of design of experiments tools for
optimization of the synthesis of UF resin.
Response surface for formaldehyde emission as a function of: a) time span between urea additions and number of urea additions (for different pH values of condensation step), b) pH of condensation step and time span between urea additions (for different numbers of urea additions) and c) pH of condensation step and number of urea additions (for different time spans between urea additions).
Optimization of Operating Conditions
the simultaneous maximization of the IB strength
value and minimization of the FE value. This was done based on the desirability
Optimization of the synthesis of UF resins using response surface methodology
113
function methodology [23], assigning preponderance factors of 0.5 and 1 to the IB
strength and FE responses, respectively, on a 0-1 scale. The optimum formulation
found is presented in Table 4.6.
Table 4.6. Operating conditions that produce the minimum formaldehyde emission
Factor Optimum
values
No. of U additions in the condensation step (A) 4
Time span (min) between U additions in the condensation step (B) 13
pH of the condensation step (B) 6.1
These synthesis conditions were reproduced in order to validate the predictions.
Table 4.7 shows the characteristics obtained for this resin.
Table 4.7. Characteristics of the UF resin optimized for minimizing the formaldehyde emission
Run Gel time1
(s) Viscosity
2
(mPa·s) Solids content
3
(%) pH
(25 °C)
Optimum 1 74 150 64.22 9.27 1Gel time at 100 °C with 3 wt % of NH4Cl (20 wt % solution)
2Brookfield viscometer at 25 °C
3120 °C, 3 h
The experimental values of the three responses for this optimized resin are
presented in Table 4.8, together with the values predicted by the empirical models
(Eqs. 4.3 to 4.5). The particleboard properties (IB strength and FE) are sufficiently
close to the predicted values, considering the inevitable variability induced by the
use of industrial grade reagents, complex control of synthesis conditions (namely
the pH history and monitoring of the viscosity in the condensation step) and the
natural heterogeneity of the wood mix used for particleboards.
The measured value of insoluble molecular aggregates is slightly lower than the
predicted value (the fitted model has a small p-value). Also, this result can be
explained by some variation in the final viscosity at condensation reaction. The
Chapter 4
114
value of insoluble molecular aggregates is significantly affected by the duration of
the condensation step.
The optimized UF resin has a good quality for a class E1 resin (the value measured
by the perforator method should be below 8 mg/100 g oven-dry board). The IB
strength is well above the minimum acceptable value of 0.35 N·mm-2 for a P2 type
board (EN312). However, resin reactivity (gel time) has to be improved.
Table 4.8. Predicted and experimental values for the three responses
Response Optimum 1
predicted experimental
IB strength1 (N·mm
-2) 0.65 ± 0.11 0.51
FE2 (mg/100 g dry board) 4.5 ± 0.5 4.2
FIA3
0.84 ± 0.018 0.076 1IB strength – internal bond strength
2FE – formaldehyde emission
3FIA – fraction of insoluble aggregates
4.4 Conclusions
An alkaline-acid synthesis process was studied using a design of experiments
methodology, in order to optimize both the adhesion performance as well as
formaldehyde emission of UF resins. It was concluded that the pH and the time
span between consecutive urea additions in the condensation step have a strong
influence on the analysed properties.
The fraction of insoluble molecular aggregates in the produced resin, detectable by
GPC/SEC analysis, is related to the particleboard performance. A larger amount of
dispersed phase seems to lead to higher internal bond strength and lower
formaldehyde emission.
In order to optimize the resins performance, in terms of internal bond strength and
formaldehyde emission, the desirability method was used. An optimized resin was
identified and produced. The properties obtained were within or close to the
values predicted by the empirical models employed.
Optimization of the synthesis of UF resins using response surface methodology
115
4.5 References
[1] E. Athanassiadou, S. Tsiantzi and C. Markessini, Towards composites with formaldehyde emission at natural wood levels. In COST Action E49
"Measurement and Control of VOC Emissions from Wood-Based Panels", Braunschweig, Germany, 2007.
[2] CARB, Airborne Toxic Control Measure to Reduce Formaldehyde Emissions from Composite Wood Products, California Environmental Protection Agency, 2007.
[3] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998). [4] J. H. Williams, US Patent 4482699 (1984). [5] P. Christjanson, T. Pehk and K. Siimer, Journal of Applied Polymer Science 100,
1673 (2006). [6] L. Graves and J. Mueller, US Patent 5362842 (1994). [7] A. Hatjiissaak and E. Papadopoulou, WO Patent 138364 (2007). [8] H. Kong, US Patent 4603191 (1986). [9] H. Spurlock, US Patent 4381368 (1983). [10] S. Vargiu, S. Giovanni, G. Mazzolen and U. Nistri, US Patent 3842039 (1974). [11] J. H. Williams, US Patent 4410685 (1983). [12] R. M. Rammon, 1984. The influence of synthesis parameters on the structure
of urea-formaldehyde resins, Ph.D. Thesis, Washington State University, Washington, USA.
[13] A. Pizzi, Wood adhesives : Chemistry and Technology, Marcel Dekker, New York (1983).
[14] A. Pizzi, L. Lipschitz and J. Valenzuela, Holzforschung 48, 254 (1994). [15] A. Pizzi and K. L. Mittal, Handbook of Adhesive Technology, Marcel Dekker,
New York (2003). [16] M. G. Kim, Journal of Polymer Science Part a-Polymer Chemistry 37, 995 (1999). [17] M. G. Kim, Journal of Applied Polymer Science 75, 1243 (2000). [18] M. G. Kim, Journal of Applied Polymer Science 80, 2800 (2001). [19] M. G. Kim, B. Y. No, S. M. Lee and W. L. Nieh, Journal of Applied Polymer
Science 89, 1896 (2003). [20] M. G. Kim, H. Wan, B. Y. No and W. L. Nieh, Journal of Applied Polymer Science
82, 1155 (2001). [21] R. N. Kumar, T. L. Han, H. D. Rozman, W. R. W. Daud and M. S. Ibrahim, Journal
of Applied Polymer Science 103, 2709 (2007). [22] D. C. Montgomery, G. C. Runger and N. F. Hubele, Engineering statistics, John
Wiley and Sons Inc., Hoboken, NJ (2001). [23] G. C. Derringer, Quality Progress 27, 51 (1994). [24] G. E. P. Box and K. B. Wilson, Journal Royal Statistical Society Series B 13, 1
(1951). [25] T. Hlaing, A. Gilbert and C. Booth, British Polymer Journal 18, 345 (1986). [26] A. Despres and A. Pizzi, Journal of Applied Polymer Science 100, 1406 (2006).
Chapter 4
116
[27] J. M. M. Ferra, A. M. Mendes, M. R. N. Costa, L. H. Carvalho and F. D. Magalhães, Journal of Applied Polymer Science 118, 1956 (2010).
[28] G. Zeppenfeld and D. Grunwald, Klebstoffe in der Holz- und Möbelindustrie, Drw Verlag Weinbrenner (2005).
[29] J. M. Ferra, J. Martins, A. Mendes, M. R. Costa, L. Carvalho and F. D. Magalhães, Characterization of the Two Different Phases in Urea-Formaldehyde Resins. In ECOWOOD, J. Caldeira, Ed. Porto, Portugal, 2008; pp 17.
CHAPTER 5
Comparison of UF synthesis by alkaline-acid and strongly acid processes
119
5 Comparison of UF synthesis by alkaline-acid and strongly
acid processes1
Abstract
This work discusses two processes for producing urea-formaldehyde (UF) resins.
One is the alkaline-acid process, which has three steps: usually an alkaline
methylolation followed by an acid condensation and finally the addition of a final
amount of urea. The other process, the strongly acid process, consists of four steps,
in which the first step involves a strongly acid condensation followed by an alkaline
methylolation, a second condensation under a moderately acid pH and finally,
methylolation and neutralization under a slight alkaline pH.
Two resins were produced using the two above described processes. The molecular
weight distribution (MWD) of the resins was monitored off-line by GPC/SEC and the
final resins were characterized by GPC/SEC and HPLC. These studies showed that
the two resins differ greatly in chemical structure, composition, viscosity and
reactivity. The monitoring of MWD indicated that the first condensation under a
strongly acid environment leads to the production of a polymer with a distinctly
different chemical structure, therefore increasing the flexibility of polymer
synthesis and opening the way to the improvement of end-use properties.
1J. M. Ferra, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho, Journal of Applied Polymer Science
(2010).Submitted
Chapter 5
120
5.1 Introduction
Owing to environmental concerns stemming from the formaldehyde emission in
wood-based panels, the formulation of urea-formaldehyde (UF) resins has evolved
towards a significant decrease in the mole ratio of formaldehyde to urea.
Unfortunately, when the aim is to achieve particularly low formaldehyde contents,
experience shows that compliance to this goal also brings about a drop in reactivity
and in resin stability as well as the degradation in the mechanical properties of the
finished boards. It is necessary to further optimize the synthesis of UF resins,
studying how the production process can be adjusted to obtain the desired
performances.
Several possibilities for the production of UF resins can be found in the literature
[1-8]. The most common is the alkaline-acid process [6-9], which has three steps -
usually an alkaline methylolation followed by an acid condensation and finally the
neutralization and addition of the last urea.
Maslosh et al. [10]reported that the alkaline-acid process produces resins with
lower content of free formaldehyde as compared with the acid-alkaline process.
According to Christjanson et al. [8] the main advantage of this process consists in
the first methylolation step, which allows a higher methylolurea content. The acid
condensation step before the methylolation stage promotes the tri-substitution in
urea through the formation of branched chains. These species decrease the
compatibility with water and the adhesion performance, namely the reactivity.
Graves et al. [6] have found that the addition of triethanolamine in the first
methylolation reaction improves not only the resistance to hydrolysis, but also the
rate of cure and decrease the formaldehyde rate of release.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
121
The strongly acid process was described for the first time by Williams [2, 4], who
has developed a method for producing low emission UF resins, consisting of:
(1) Carrying out the condensation of urea and formaldehyde under a highly acid
environment and large excess of formaldehyde;
(2) Pursuing the reaction in alkaline medium after urea addition, in order to
achieve a predetermined F/U molar ratio;
(3) Further carrying out the reaction under a pH of about 5 to allow a
supplementary condensation until the desired viscosity is reached;
(4) Neutralization and addition of a final amount of urea to obtain the intended
low F/U molar ratio.
The Williams process [2, 4] entails minimal energy consumption and involves
relatively short (3-4 hours) reaction times. The reduced formaldehyde emission and
increased hydrolytic stability have been attributed to the predominance of the
more stable methylene linkages in the cured resin, unlike the alkaline-acid process
which leads to a larger amount of methylene ether linkages in the cured resin.
The main problem of this process lies in the control of the strongly acid
condensation step, due to its exothermic character. According to Hatjiissaak and
Papadopoulou [5], this implies careful control, which may be difficult to achieve on
an industrial scale, to prevent resin gelling in the reactor.
In this work, two resins were produced using the alkaline-acid and strongly acid
processes. The synthesis of the resins was off-line monitored by GPC/SEC. The
monitoring shows that the polymerization path is very different for each process.
The strongly acid process forms, initially, polymer with high molecular weight,
which is insoluble due to the long size of chains with methylene bridges. The
determination of MWD (in fact, it could be better described as the molecular size
Chapter 5
122
distribution in terms of the molecular weight of the poly(ethylene glycol) with the
same hydrodynamic radius, we refer MWD for simplicity) of the final resins
indicated that the resin produced by the strongly acid process has a large fraction
of insoluble molecular aggregates in contrast with the alkaline-acid process.
5.2 Materials and Methods
5.2.1 Resin preparation
In the preparation of UF resins, industrial-grade raw materials were used, namely:
urea, 50 % formalin, melamine, hexamine, sodium hydroxide solution, acetic acid
and sulfuric acid. The synthesis of resins was carried out with a bench scale 5 L glass
reactor. They were produced according to two synthesis methods.
Resin UF-Exp7 was produced according to the alkaline-acid process and resin UF-
W6 according to the strongly acid process. The two procedures are described
below.
The two resins are water dispersions with a low amount of melamine (0.3 %) and
hexamine (0.1 %) which are typically added for improving hydrolysis resistance and
increase the buffering capacity, respectively.
Alkaline-acid process (synthesis of UF-Exp7)
The reaction of urea and formaldehyde consists basically in a three step process:
i. methylolation step at alkaline conditions;
ii. condensation step at acidic conditions (pH about 5.6) up to 350-450 mPa·s;
iii. neutralization (pH > 7.5) of the product and addition of a final amount of
urea to obtain a desired low formaldehyde/urea ratio.
The formulation of resin UF-Exp7 was based on a previously reported [11]
optimization study of the traditional alkaline-acid process. From the design of
Comparison of UF synthesis by alkaline-acid and strongly acid processes
123
experiments procedure employed, UF-Exp7 yielded the highest value of internal
bonding while maintaining an acceptably low formaldehyde emission value.
After the required amount of 55.5 % formaldehyde solution and urea were loaded
into a reactor (F/U molar ratio of 2.00 - 2.15), the reaction mixture was held at the
final temperature (95 °C) for about 30 min under alkaline conditions (pH = 8.5 -
9.5).
The pH was then adjusted to about 5.6 - 6.1 with acetic acid solution and the
second urea was added sequentially to the previous reaction mixture (four
additions of urea and thirteen minutes between additions). The F/U molar ratio
after this step is approximately 1.8.
After the desired viscosity being reached, the reaction was stopped by
alkalinisation with sodium hydroxide solution and cooled down. The third (final)
urea was then added at 60 °C, yielding a final F/U ratio of 1.12. Figure 5.1 shows a
diagram describing the temperature and pH histories during resin preparation.
Time / min
0 50 100 150 200
Temperature / º C
20
40
60
80
100pH
4
5
6
7
8
9
10
add. U2
add. U1 add. U3
S1 S2 S3 S4 S5
Figure 5.1. Reaction temperature (—) and pH (---) histories for resin UF-Exp7. These simplified history curves are based on the experimentally measured values. The urea addition (Ui) and sample collection (Si) times are also indicated in the graph.
Chapter 5
124
Strongly acid process (synthesis of UF-W6)
The reaction consists basically in a four step process:
i. condensation of urea and formaldehyde under a strongly acid environment
and a large excess of formaldehyde;
ii. methylolation step at alkaline medium;
iii. condensation at a low pH ( about 5) as soon as the desired viscosity was
reached;
iv. neutralization (pH > 7.5) and fast cooling; final addition of urea to obtain
the specified low F/U molar ratio.
Formaldehyde 55.5 % solution was charged into the reactor and the pH value was
adjusted to 1-2 using sulfuric acid. Urea (F/U ~3.00 - 3.25) was added in 15 equal
parts over a time span of 15 min. The reaction is very exothermic and the
temperature increased to 80 °C without external heat supply. The temperature was
kept at 80 °C for 10 min. After 10 min of reaction the pH was adjusted to 7.3-7.5
with a 50 % sodium hydroxide solution. A second amount of urea and melamine
were charged into the reactor and the temperature was increased to 95 °C and the
reaction followed during 15 min. The pH was again adjusted to 5.2 - 6.0 with 25 %
acetic acid solution and the polymer was condensed until the desired viscosity (450
- 550 mPa·s) was reached. The pH was then adjusted to 7.5 with 50 % sodium
hydroxide solution and the reacting media was cooled down. At 60 °C, the last
amount of urea and hexamine were added to the reactor to obtain the specified
F/U molar ratio (1.10), and again the pH was adjusted to obtain a pH 7.5 - 8.5. After
the whole amount of urea was dissolved the solution was cooled to room
temperature. The temperature and pH histories along resin preparation are shown
in Figure 5.2.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
125
Time / min
0 40 80 120 160 200 240
Temperature / º C
20
30
40
50
60
70
80
90
pH
0
2
4
6
8
seq. add. U1
add. U2
add. U3
S1 S2S3
S4 S5S6
Figure 5.2. Reaction temperature (—) and pH (---) histories for resin UF-W6. These simplified history curves are based on the experimentally measured values. The urea addition (Ui) and sample collection (Si) times are also indicated in the graph.
5.2.2 GPC/SEC analysis
A Gilson HPLC system equipped with a Gilson Differential RI detector and Viscotek
Dual Detector (differential viscometer and a light scattering detector RALLS).and a
Rheodyne 7125 injector with a 20 µL loop have been used. The selected column
was a Waters Styragel HR1 with 5 µm particle size. The RALLS detector was not
used for the GPC/SEC molecular weight calculations because of the weak response
for lower molecular weights. The analysis conditions and samples preparation have
been described in detail elsewhere [12].
5.2.3 HPLC analysis
A JASCO HPLC system equipped with a JASCO Differential RI detector, and a
Rheodyne 7725i injector with a 100 µL loop was used. The chosen separating
column was a Waters Spherisorb NH2 with 5 µm nominal particle size.
Acetonitrile/water (90/10) was used as the mobile phase. The column was
conditioned at 30 °C using an external oven and the flow rate was 1.5 mL·min-1.The
samples were prepared according to the procedure described in Ferra et al. [12].
Chapter 5
126
5.3 Results and discussion
5.3.1 Characteristics of the produced resins
Table 5.1 shows the technical data of the two UF resins characterized in this work.
The UF resins were stored at 25 °C for the chemical analysis in GPC/SEC and HPLC.
It should be noticed that resin UF-W6 had a high gel time (low reactivity), while
viscosity was comparable to the UF-Exp7.
Table 5.1. Technical data of UF-resins
1120 °C, 3 h
2Gel time at 100 °C with 3 wt % NH4Cl (20 wt % solution)
3Brookfield viscometer at 25 °C
5.3.2 Monitoring of UF synthesis
In order to improve the knowledge on the polymerization reactions occurring in
alkaline-acid and strongly acid processes, each synthesis was monitored off-line by
GPC/SEC.
Alkaline process
Samples were collected at some critical moments of the synthesis: start and end of
methylolation stage, start and end of condensation stage and before the addition
of the last urea. The complete description of the sampling procedure is presented
in Table 5.2. The instants when the samples were collected are also shown on
Figure 5.1.
Resin Molar ratio
F/U Solids
content1 (%)
Reactivity2
(s) pH value
(25 °C) Viscosity
3
(mPa.s)
UF-Exp7 1.12 64.1 71 8.64 150 UF-W6 1.10 63.3 112 8.16 270
Comparison of UF synthesis by alkaline-acid and strongly acid processes
127
Table 5.2. Identification of different stages during the synthesis (UF-Exp 7)
Stage Reaction
time (min) Reaction step T (°C) pH
Sample 1 27 At the middle of methylolation step
90-100 9.0-9.4
Sample 2 55 At the end of methylolation step
90-100 9.0-9.4
Sample 3 106 During condensation step, when visc. ~80 mPa·s at 25 °C
90-100 5.6-6.2
Sample 4 141 During condensation step, when visc. ~400 mPa·s at 25 °C
90-100 5.6-6.2
Sample 5 159 Before addition of last urea <60 >8.5
Figure 5.3 shows the GPC/SEC chromatograms obtained for the different samples.
It is interesting to follow the evolution of the low molecular weight species during
the reaction, taking into account the assignment of the peaks, discussed in a
previous work [12]: Peak 1 - urea and methylolureas; Peak 2 - methylolureas and
some oligomeric species; Peak 3 - other oligomeric species. In Figure 5.3a it is
possible to observe the formation of methylolureas and oligomeric species in the
methylolation stage (see Figure 5.4).
Vretention/ml
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
sample 1
sample 2
UF-Exp7
Vretention/ml
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
1.0
sample 3
sample 4
sample 5
UF-Exp7
(a) (b)
Peak 3
Peak 2Peak 1
Figure 5.3. Monitoring of UF-Exp 7 synthesis by GPC/SEC: a) samples collected during methylolation step, including the chromatogram of the final resin, and b) samples during condensation step, including the chromatogram of the final resin.
Chapter 5
128
Figure 5.4. Formation of methylolureas (mono-, di- and trimethylolurea) by the addition of formaldehyde to urea.
The chromatograms obtained for the samples collected during the condensation
reaction are presented in Figure 5.3b. During this process, the methylolureas are
condensated to form linear and/or branched polymers linked by methylene-ether
and methylene bridges (see Figure 5.5). The almost complete disappearance of the
peaks of urea and methylolureas indicates the progress of the condensation
reaction. At the same time, moderate molecular weight polymer is formed. The
chromatogram of sample 4 shows the presence of a small amount of insoluble
molecular aggregates in solution (elution volume below 5.8 mL) [12]. Samples 4
and 5 do not differ significantly indicating that the reaction was blocked by the
neutralization and cooling of the solution. Comparing the chromatograms for
sample 5 and the final resin, we can see that methylolureas were formed by
reaction of the last urea added with free formaldehyde in solution. Note that Peak
1 has been identified as a combination of urea and methylolureas and not urea
alone [12]. In addition, some condensation of higher polymer has occurred.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
129
Figure 5.5. Condensation of the methylolureas and urea to form methylene-ether bridges and methylene bridges.
The values of the molecular weight averages (based on polystyrene standards)
have been computed for all samples collected during the synthesis and are
presented in Table 5.3. According to Ferra et al. [12] the chromatogram fraction
corresponding to elution volumes below 5.8 mL corresponds essentially to
insoluble material, so the molecular weights were computed neglecting this portion
of the chromatograms. The obtained results show the increase of polydispersity of
the polymer with the progress of reaction. It is evident the growth of the
polydispersity from sample 2 (methylolation reaction) to sample 3 (condensation
reaction) and the stabilization of polydispersity along the condensation step. The
final increase, from sample 5 to the end product, is related to the last urea
addition.
Table 5.3. Values of Mn, Mw, polydispersity (Mw/Mn) obtained by GPC/SEC for samples collected during the synthesis of UF-Exp7
Sample Mn Mw Mw/Mn
Sample 1 5.70×103 6.18×10
3 1.1
Sample 2 2.98×103 4.66×10
3 1.6
Sample 3 3.51×103 1.20×10
4 3.4
Sample 4 8.70×103 3.18×10
4 3.7
Sample 5 7.39×103 3.04×10
4 4.1
UF-Exp7 5.50×102 2.78×10
3 5.1
Chapter 5
130
Strongly acid process
Table 5.4 summarizes the information about the samples collected during the
monitoring of the synthesis of UF-W6, at key points of the synthesis path: first
strongly acid condensation reaction, methylolation reaction, second condensation
reaction and after the addition of the last urea (final methylolation).
Table 5.4. Identification of different stages during the synthesis
Stage Reaction
time (min) Reaction step T (°C) pH
Sample 1 40 At the end of the 1st
condensation step
76 2.0
Sample 2 80 At the end of the 1st
methylolation step
74 7.4
Sample 3 112 At the 2nd
condensation step, viscosity ~200 mPa·s
89 5.5
Sample 4 130 At the 2nd
condensation step, viscosity ~300 mPa·s
90 5.5
Sample 5 161 At the 2nd
condensation step, viscosity ~500 mPa·s
88 5.5
Sample 6 220 After addition of the last urea 35 >7.5
As it can be seen from the GPC/SEC analysis in Figure 5.3 and Figure 5.6, the
polymerization reactions carried out by this process are very different from the
ones found with the alkaline-acid process.
Figure 5.6 shows the MWD of the products obtained. High molecular weight
material can be seen, which should be constituted by extensive methylene or
methylene-ether bridges chains and urons.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
131
Vretention/ml
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
sample 1
sample 2
Vretention/ml
4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
sample 6
UF-W6
(a) (b) Vretention/ml
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
sample 3
sample 4
sample 5
(c)
Figure 5.6. Monitoring of UF-W6 synthesis by GPC/SEC: a) 1ª condensation step and 1ª methylolation step, b) 2ª condensation step, and c) 2ª methylolation step and final resin.
In the first condensation reaction under strongly acid conditions the methylolureas
were formed, but condensate rapidly to form methyleneureas (e.g.
methylenediurea, dimethylenetriurea, trimethylenetetraurea, etc.) [13]. Kadowaki
[14] has reported the synthesis of methylenediurea by the reaction of
formaldehyde and urea at acid conditions. The progress of the reaction produces
linear products with methylene bridges. Motter [15] has suggested that this
polymer is insoluble above 4 to 8 urea units in the chain. Probably, uron groups are
also formed in this step. The formation of uron groups in strongly acid
Chapter 5
132
condensation of urea and formaldehyde was reported by Beachem et al. [16]. The
main reactions taking place in this first stage are presented in Figure 5.7.
Figure 5.7. Condensation reactions of urea and formaldehyde to form methylolureas that form methyleneureas and urons.
After the condensation reaction, the synthesis proceeds by the methylolation of
unreacted formaldehyde and urea (second amount of urea was added in these
stage) to form methylolureas (see Figure 5.4). This is shown in Figure 5.3a by
development of the rightmost peak in the chromatogram of sample 2.
Then, the new condensation reaction takes place, forming chains of methylene and
methylene-ether bridges, the product of the reaction containing preferentially
methylene bridges according to Williams [2, 4]. Chromatograms in Figure 5.6b
show the condensation of methylolureas forming polymer with larger molecular
weight. The polymerization reaction proceeds until the desired viscosity is reached
forming a large amount of polymer with high molecular weight, which might be in
the form of insoluble molecular aggregates (see chromatograms of samples 3, 4
and 5 in Figure 5.6b).
Finally, Figure 5.6c shows the last reaction that takes place. The appearance of the
peak corresponding to urea and methylolureas (sample 6) indicates the presence of
unreacted urea and its reaction with the remaining formaldehyde.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
133
The values of the apparent molecular weight averages (see Table 5.5) show the
broad distributions of the polymer present in the resin UF-W6. This resin has a
large fraction of higher polymer, which is particularly important for the cohesive
strength of the resins [9].
Table 5.5. Values of Mn, Mw, polydispersity (Mw/Mn) obtained by GPC/SEC for samples collected during the synthesis of UF-W6
Sample Mn Mw Mw/Mn
Sample 1 7.13×102
1.64×104 22.9
Sample 2 3.25×102 3.34×10
3 10.3
Sample 3 7.30×102 5.96×10
3 8.2
Sample 4 6.48×102 4.75×10
3 7.3
Sample 5 7.49×102 6.66×10
3 8.9
Sample 6 2.06×102 1.69×10
3 8.2
UF-W6 2.75×102 7.23×10
3 26.3
5.3.3 Comparison of the two resins
Determination of molecular size distribution
In Figure 5.8 we can see the GPC/SEC chromatograms for two resins (UF-Exp7 and
UF-W6), in terms of the normalized weight fraction (Wt Fr), measured 5 days after
synthesis. In both cases, at least two samples were prepared and analysed in order
to verify the reproducibility of the results. Three zones, based on the apparent
limits of detectable peaks in the chromatograms, can be defined in the
chromatograms [12]. The Zone I (elution volume between 8 and 9 mL) corresponds
to the lower molecular weight species. Zone II (elution volume between 5.8 and 8
mL) corresponds to intermediate molecular weight species, with apparent
molecular weights between about a few tens of thousand Da and about 600 Da.
Zone III (elution volume below 5.8 mL) would correspond to polymer with quite
high molecular weights, eluting before the exclusion limit of the GPC/SEC column.
Ferra et al. [12] and, other authors [17-18] suggested that this portion of the
chromatograms actually consists of molecular aggregates and not of solubilized
polymer chains. These aggregates are insoluble in the original aqueous medium,
Chapter 5
134
forming larger colloidal structures that partially disaggregated in the DMSO solvent
used in GPC/SEC sample preparation [12].
Observing the two chromatograms it can be readily noticed that the UF-W6 resin
presents a large fraction of insoluble aggregates as compared with UF-Exp7 and
other resins produced according with this method and reported in Ferra et al. [12].
It is clear that resin UF-W6 has a smaller fraction of polymer with moderate
molecular weights, which reacted previously forming large linear chains of
methylene bridges and urons. The fraction of low molecular weight species is
superior for UF-Exp7 resin, therefore explaining the higher reactivity of this resin.
Vretention/ml
3 4 5 6 7 8 9
Norm
alize W
t Fr
0.0
0.2
0.4
0.6
0.8
UF-W6
UF-Exp7
Figure 5.8. Chromatogram for four resins (UF-W6 and UF-Exp7) with 5 days (stored at 25 °C).
As it has been above stated, the molecular weights were computed neglecting the
insoluble aggregates.
Comparison of UF synthesis by alkaline-acid and strongly acid processes
135
Ferra et al. [12] have defined two empirical parameters as an aid for quantifying
the features of the MWD in UF resins;
• f1 reflecting the amount of low molecular weight species in the sample;
amchromatogr of areatotal
I Zone of area=f
1 (5.1)
• f2 related to the relative importance of what would be the high molecular
weight species in the polymerized material;
III Zone +II Zone of area
III Zone of area=f
2 (5.2)
Table 5.6 shows the numerical values obtained of these parameters for the two
resins studied in this work. These results support the previous statements derived
from the direct inspection of the chromatograms.
Table 5.6. Values of Mn, Mw and parameters f1 and f2, obtained by GPC/SEC for resins (UF-W6 and UF-Exp7)
Sample Mn Mw f1 f2
UF-Exp7 5.50×102 2.78×10
3 0.40 0.045
UF-W6 2.75×102 7.23×10
3 0.35 0.13
Determination of the fraction of urea and methylolureas
Following the procedure described by Ferra et al. [19] the fraction of free urea and
methylolureas present in final resins was measured by HPLC, as seen in Figure 5.9.
The distribution of the unreacted urea, monomethylolurea and dimethylolurea
present in solution 5 days after the synthesis shows that the UF-W6 resin has a
much larger fraction of unreacted urea than resin UF-Exp7. As the amount of last
urea incorporated was similar for the two resins, other characteristics had an
impact on the distribution of unreacted urea in the final resin. The main difference
Chapter 5
136
in the last step of the process is a rapid cooling that takes place after addition of
the last urea for the strongly acid process. This action delays the consumption of
urea, which remains unreacted in the final resin. However, this unreacted urea may
play another role, since it may form a solvatation layer surrounding the colloidal
aggregates surface, contributing to its stabilization towards agglomeration [12].
Recently, Park et al. [20] reported that the unreacted urea can also be a good
scavenger of formaldehyde during the hot-pressing of the particleboards.
U MMU DMU
Peak area/total area
0
10
20
30
40
50
60
UF-W6
UF-Exp7
Figure 5.9. Ratios of peak areas / total area of the urea (U), monomethylolurea (MMU) and dimethylolurea (DMU) for UF-W6 and UF-Exp7 resins, with 5 days (stored at 25 °C).
Resin UF-Exp7 has a large fraction of monomethylolurea as compared with the
other resin, but the fraction of dimethylolurea is similar.
5.4 Conclusions
This work describes two pathways for producing UF resins, the alkaline-acid and
the strongly acid processes; the synthesised resins were followed by GPC/SEC. It
was found that the structures of the UF polymer formed by the two processes are
Comparison of UF synthesis by alkaline-acid and strongly acid processes
137
quite different. At the first stage of the strongly acid process, polymer with high
molar mass (containing essentially links of methylene groups) is formed, in contrast
with the alkaline-acid process, where this has to be avoided because of possible
resin gelling inside the reactor.
This study shows the capabilities of the GPC/SEC technique for monitoring and
controlling the synthesis of UF resins. This is more evident when comparing resins
produced by different processes. The technique permits to quickly identify the kind
of polymer obtained in different steps of reaction, with a high reproducibility.
The measured molecular weight distribution of UF resins shows that the resin
produced by the strongly acid process has a great fraction of insoluble molecular
aggregates, which is believed to contribute to a higher resin cohesive strength.
However, it has also a smaller fraction of low molecular weight species as
compared with the alkaline-acid process, which might contribute to improve resin
wettability on the wood substrate.
The results obtained by HPLC showed that the resins produced by the strongly acid
process have a large fraction of unreacted urea and a small fraction of
monomethylolurea in contrast with the alkaline-acid process.
5.5 References
[1] S. Vargiu, S. Giovanni, G. Mazzolen and U. Nistri, US Patent 3842039 (1974). [2] J. H. Williams, US Patent 4410685 (1983). [3] H. Spurlock, US Patent 4381368 (1983). [4] J. H. Williams, US Patent 4482699 (1984). [5] A. Hatjiissaak and E. Papadopoulou, WO Patent 138364 (2007). [6] L. Graves and J. Mueller, US Patent 5362842 (1994). [7] H. Kong, US Patent 4603191 (1986). [8] P. Christjanson, T. Pehk and K. Siimer, Journal of Applied Polymer Science 100,
1673 (2006). [9] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998). [10] V. Z. Maslosh, V. V. Kotova and O. V. Maslosh, Russian Journal of Applied
Chemistry 78, 685 (2005).
Chapter 5
138
[11] J. M. Ferra, P. C. Mena, J. Martins, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho, Journal of Adhesion Science and Technology 24, 1455 (2010).
[12] J. M. M. Ferra, A. M. Mendes, M. R. N. Costa, L. H. Carvalho and F. D. Magalhães, Journal of Applied Polymer Science 118, 1956 (2010).
[13] A. Pizzi and K. L. Mittal, Handbook of Adhesive Technology, Marcel Dekker, New York (2003).
[14] H. Kadowaki, Bulletin of the Chemical Society of Japan 11, 248 (1936). [15] W. K. Motter, 1990. The formation of the colloidal phase in low mole ratio
urea-formaldehyde resins, Ph.D. Thesis, Washington State University, Washington, USA.
[16] M. T. Beachem, Schicked.Pd, J. C. Oppelt, D. V. Maier and F. M. Cowen, Journal
of Organic Chemistry 28, 1876 (1963). [17] D. Grunwald, Kombinierte analytische Untersuchungen von Klebstoffen für
Holzwerkstoffe, Mensch & Buch Verlag, Berlin, Germany (2002). [18] T. Hlaing, A. Gilbert and C. Booth, British Polymer Journal 18, 345 (1986). [19] J. M. Ferra, J. Martins, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H.
Carvalho, Journal of Adhesion Science and technology 24, 1535 (2010). [20] B. D. Park, E. C. Kang and J. Y. Park, Journal of Applied Polymer Science 110,
1573 (2008).
CHAPTER 6
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
141
6 Evaluation of UF adhesives performance by IPATES and
ABES mechanical tests1
Abstract
In the open literature, two main strategies can be found for synthesizing urea-
formaldehyde (UF) resins. One is the alkaline-acid process, which takes place in
three steps, usually an alkaline methylolation followed by an acid condensation and
then the addition of a final amount of urea. The other process consists of four
steps, the main difference being an initial condensation in strongly acid
environment.
In this work, we evaluate the curing behaviour of four resins produced using the
aforementioned processes by the Integrated Pressing and Testing System (IPATES)
and the Automated Bonding Evaluation System (ABES).
The characterisation of the bond strength development during hot pressing by
ABES and IPATES shows that the four resins will have different performances in the
bonding process of wood-based composites. For each resin, the effect of pressing
parameters such as temperature, adhesive and hardener ratios on shear strength
(ABES) and internal bond (IPATES) during hot pressing is put into evidence.
1 J. M. Ferra, M. Ohlmeyer, A. Mendes, M. R. Costa, L. Carvalho and F. D. Magalhães, International Journal of
Adhesion and Adhesives (2010).Accepted
Chapter 6
142
6.1 Introduction
Despite the serious issue of formaldehyde emissions, urea-formaldehyde (UF)
resins are still the most used binders in the wood based industry. The main strategy
to reduce the formaldehyde emission of UF resins has been the change on its
formulation by decreasing the molar ratio of formaldehyde to urea. However, the
reduction weakens the mechanical properties of particleboard and moreover it
increases the time required for hardening under the action of current hardeners
(latent acids) [1]. Hence, it is necessary to optimize the synthesis of UF resins,
therefore changing the production process.
In literature, several possibilities for producing UF resins [2-11] have been
described by various researches. However, these can be summarized into two main
classes: the strongly-acid and the alkaline-acid processes. The details on the
reaction pathways occurring in each process have been described elsewhere by
Ferra et al. [12].
The understanding of adhesive cure behaviour and its dependence on temperature
and chemical conversion is an important starting point for the establishment of
processing windows for board production and prediction of the properties of cured
bondlines.
According to Schultz and Nardin [13], the main adhesion theories are: mechanical
interlocking, electrostatic, weak boundary layer, adsorption, diffusion and chemical
bonding. The adsorption hypothesis, which explains adhesion by intermolecular
and interatomic forces, such as van der Waals bonds, hydrogen bonds and
electrostatic interactions, is considered by many authors as the most applicable to
wood-polymer adhesion [14].
Marra [15] described adhesive bond formation in wood based panels as a dynamic
process, which depends on five actions: flow, transference, penetration, wetting
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
143
and solidification (cure).
Chemical curing (build-up of the three-dimensional network) can be monitored by
differential scanning calorimetry (DSC) [16-18], which allows the measurement of
the degree of chemical curing, as well as the heat of polymerization. Adhesive bond
strength development (sometimes called mechanical curing) can be monitored by
thermo-mechanical analysis (TMA) [19-21] and dynamic mechanical analysis (DMA)
[22]. Šernek and Kamke [23] reported that the dielectrical analysis (DEA) can be
used for in situ monitoring of the cure of thermosetting adhesives. The results thus
obtained suggest that the dielectric changes were related to the extent of cure of
the adhesive. The Automated Bonding Evaluation System (ABES) [24] technique
allows exploring the strength development of a diversity of wood adhesives [25].
Dunky [26] suggested that ABES jointly with DMA are the best methods for
monitoring bond strength development during the resin cure. More recently,
Integrated Pressing and Testing System (IPATES) has demonstrated to be a
powerful technique to characterize curing and bonding processes in resinated
fibres [25].
In this work, the evaluation of the adhesive bond strength development during the
hot-pressing by ABES showed that the shear strength of the four resins tested was
mainly affected by press temperature. IPATES results suggested that the effect of
press temperature and adhesive ratio on internal bond values was different for
each resin. The results also indicated a logarithmic correlation between IB and
press time.
6.2 Materials and Methods
6.2.1 Raw materials
Table 6.1 shows the technical data of four UF resins produced either with the
strongly-acid (UF-W6) or alkaline-acid (UF-Exp7, UF-R8 and UF-R2) processes. The
Chapter 6
144
formulation of resin UF-Exp7 is described on a previously reported optimization
study [27] for the traditional alkaline-acid process. Resins UF-R8 and UF-R2 were
produced using the same procedure of UF-exp7, but with a different
formaldehyde/urea (F/U) molar ratio. This allows evaluating the effect of molar
ratio F/U on the performance of the adhesives. In the preparation of UF-W6 resin, a
formaldehyde 55.5 % solution was charged into the reactor and the pH value was
adjusted to 1-2 using sulfuric acid. Urea (F/U ~3.00 - 3.25) was added in 15 equal
parts over a time span of 15 min. The reaction is very exothermic and the
temperature increased to 80 °C without external heat supply. The temperature was
kept at 80 °C for 10 min. After 10 min of reaction the pH was adjusted to 7.3 - 7.5
with a 50 % sodium hydroxide solution. A second amount of urea and melamine
were charged into the reactor and the temperature was increased to 95 °C and the
reaction followed during 15 min. The pH was again adjusted to 5.2 - 6.0 with 25 %
acetic acid solution and the polymer was condensed until the desired viscosity (450
- 550 mPa·s) was reached. The pH was then adjusted to 7.5 with 50 % sodium
hydroxide solution and the reacting media was cooled down. At 60 °C, the last
amount of urea and hexamine were added to the reactor to obtain the specified
F/U molar ratio (1.10), and again the pH was adjusted to obtain a pH 7.5 - 8.5.
Table 6.1. Technical data of UF-resins
Resin Molar
ratio F/U Solids
content1 (%)
Reactivity2
(s) pH value (25 °C)
Viscosity3
(mPa.s)
Alkaline-acid process
UF-Exp7 1.12 64.1 71 8.64 150 UF-R8 1.12 64.4 86 8.22 210 UF-R2 1.00 64.8 90 8.49 216
Strongly-acid process
UF-W6 1.10 63.3 112 8.16 270 1105 °C, 3 h
2Gel time at 100 °C with 3 % wt of NH4Cl (20 wt.% solution)
3Brookfield viscometer at 25 °C
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
6.2.2 Methods
Integrated Pressing and Testing System (IPATES)
IPATES (see Figure 6.1) is a recent
Heinemann [25] at the University of Hamburg (Germany) to characterize the cure
of adhesives by the evaluation of the internal bond in wood
during hot pressing. In this work, resinated wood particle mats are pressed under
strictly controlled conditions regarding
mats are then destructively tested in tension mode, without
the system (see Figure 6.2). A commercial polyurethane
bond the particle mat to the pressing platens.
In this test, wood particles usually chosen
particleboards were used. Wood particles with approximately 3 % o
content were blended with the resins and hardener (ammonium sulphate) in a
laboratory-scale glue blender. All boards
density of 650 kg/m3 and 6 mm thickness.
Figure 6.1. Overview and schematic of the IPATES machine.
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
145
Integrated Pressing and Testing System (IPATES)
) is a recent characterization method developed by
at the University of Hamburg (Germany) to characterize the cure
of adhesives by the evaluation of the internal bond in wood-based composites
during hot pressing. In this work, resinated wood particle mats are pressed under
egarding temperature, thickness and pressure. The
destructively tested in tension mode, without being removal from
commercial polyurethane-adhesive was applied to
the pressing platens.
chosen for the surface layer in the production of
particleboards were used. Wood particles with approximately 3 % of moisture
content were blended with the resins and hardener (ammonium sulphate) in a
glue blender. All boards have been produced with a nominal
thickness.
Overview and schematic of the IPATES machine.
Chapter 6
146
Figure 6.2. Particleboard mat during the pressing and subsequent testing.
This equipment allows testing the parameters that
development of wood particles boards
adhesive and adhesive ratio. The summary of the parameters tested in this work
present on Table 6.2.
Table 6.2. Parameters evaluated by IPATES
Test parameters
Press temperature
Pressing time
Adhesive ratio
Figure 6.3 shows the board core temperature
temperatures of 130 °C and 160 °C, measured with a thermocouple inserted at the
beginning of the pressing time. In the first
cure temperature of UF adhesives) in the core is reached only for pressing times
above 90 s at 130 °C, whereas, for a pressing temperature of 160
attains 90 °C after about 60 s. The change in slope
temperature is associated to a change in the heat transfer mechanism. The initial
heating rate is determined by the inward displacement of steam generated in the
outer layers of the mat. When the pressing temperature is hig
and displacement is faster, but the water available in the wood particles ends up
Particleboard mat during the pressing and subsequent testing.
testing the parameters that influence the bond strength
boards, namely: press temperature, press time,
adhesive and adhesive ratio. The summary of the parameters tested in this work is
s evaluated by IPATES
Press temperature 130 °C and 160 °C
30 s – 360 s
8 % and 10 %
core temperature history of the board for pressing
, measured with a thermocouple inserted at the
beginning of the pressing time. In the first case, a temperature of 90 °C (typical
temperature of UF adhesives) in the core is reached only for pressing times
C, whereas, for a pressing temperature of 160 °C, the board core
The change in slope observed for the latter pressing
temperature is associated to a change in the heat transfer mechanism. The initial
heating rate is determined by the inward displacement of steam generated in the
outer layers of the mat. When the pressing temperature is high, steam production
and displacement is faster, but the water available in the wood particles ends up
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
147
being completely vaporized. When this happens, heat conduction slows down and
the core heats up at a lower rate.
Pressing time / s
0 50 100 150 200 250 300
Temperature / ºC
20
40
60
80
100
120
140
130 ºC
160 ºC
Figure 6.3. Core layer temperature of the board at different pressing temperatures.
Automated Bonding Evaluation System (ABES)
The increase of mechanical bond strength with time was followed using ABES (see
Figure 6.4). Two beech veneer strips measuring 0.5 mm tick, 20 mm wide and 117
mm in length, stored at 25 °C and 65 % RH were glued together with an overlap of
5 mm. The veneers were prepared using a die cutter supplied by Adhesive
Evaluation Systems (Corvallis, Oregon). Adherent pairs were mounted in the system
with an overlapping area of 100 mm2 and pressed together at 1.2 N·mm-2. The
amount of adhesive system used for each test was 15 µL. The bond strength was
tested almost instantaneously in shear mode (the system is digitally controlled and
pneumatically driven). The parameters evaluated on ABES are summarized in Table
6.3. The higher temperature tested for resin UF-W6 was 130 °C. However, for the
other resins we observed that strength development was too fast at this
Chapter 6
148
temperature. Therefore, the higher temperatu
Exp7, UF-R8 and UF-R2 was 120 °C.
Figure 6.4. Schematic of the ABES test procedure.
Table 6.3. Parameters evaluated on ABES
Test parameters
Press temperature 90
Press time
Hardener ratio
6.3 Results and discussion
6.3.1 Integrated Pressing and Testing
Comparison of UF resins
The bond strength values of the resinated wood particle discs, pressed at 130
and 160 °C, were tested in tension mode after hot
resins are shown in Figure 6.5 and
pressing times, the values of the bond strength also increase. However, after a
certain press time, the values decrease, probably due to thermal degradation of
wood and cured resin.
higher temperature tested for the other resins UF-
Schematic of the ABES test procedure.
Parameters evaluated on ABES
90 °C, 100 °C, 110 °C and 120 °C or 130 °C
10 s – 360 s
1.5 % and 3.0 %
Integrated Pressing and Testing System (IPATES)
The bond strength values of the resinated wood particle discs, pressed at 130 °C
C, were tested in tension mode after hot-pressing. The results for four
and Figure 6.6. We can observe that increasing the
the values of the bond strength also increase. However, after a
certain press time, the values decrease, probably due to thermal degradation of
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
149
Effect of the production process
The results in Figure 6.5 show that the resin produced with alkaline-acid process
(UF-Exp7) present a higher performance and faster IB strength development
compared with the resin produced with strongly-acid process (UF-W6). The highest
IB strength (0.72 N·mm-2) was obtained for resin UF-Exp7 for 150 s of pressing time
at the highest temperature tested, 160 °C.
Analysis of the molecular weight distribution of the two resins showed that the one
produced by the strongly acid process (UF-W6) presents a higher fraction of
material in the high molecular weight region, and a smaller fraction of low
molecular weight species, as compared with the alkaline-acid process [12]. An UF
resin is an aqueous dispersion of low molecular weight (soluble) and high molecular
weight (insoluble) species. It is known that a well performing resin should show a
balance between the two fractions, in order to combine good flowability,
wettability and penetration with good cohesive strength and bondline coverage
[28-29]. We believe that after the optimization of the strongly-acid process, the
performance of the resin should increase. Another important issue related to this
process is the formation of uron structures, which are important for lowering
formaldehyde emission, due to their high stability [11, 20]. However, the reactivity
of the methylol group in these structures is much lower than for the methylol
group in methylolureas [1], obtained in the traditional alkaline-acid formulations.
Chapter 6
150
Pressing time / s
30 60 90 120 150 180 210 240
IB / N·mm-2
0.0
0.2
0.4
0.6
0.8
UF-W6
UF-Exp7
Pressing time / s
0 30 60 90 120 150
IB / N·mm-2
0.0
0.2
0.4
0.6
0.8
UF-W6
UF-Exp7
a) b)
Figure 6.5. Bond strength curves for particle mats pressed at 130 °C and 160 °C for UF-W6 and UF-Exp7 (8 % resin and 3 % hardener). a) 130 °C, and b) 160 °C.
Effect of F/U molar ration
Observing the results of resins UF-R8 and UF-R2 (see Figure 6.6), it is evident that
the resin with a higher formaldehyde/urea molar ratio (UF-R8) yields a higher
performance at 130 °C. However, for higher press temperatures, this difference
becomes very slight. These two resins effectively differ only in terms of free
formaldehyde present and monomeric species that have been formed in the
reaction with the last added urea with the high free formaldehyde content. As may
be expected, differences concerning essentially the low molecular weight species
present in the resins only affect the reactivity and, mainly, the formaldehyde
emission levels, which was not measured in this work.
Effect of temperature
Figure 6.5 and Figure 6.6 also depict the effect of press temperature on the boards’
bond strength of the boards as a function of pressing time, for the four resins
studied. It is evident that the IB development is significantly accelerated at high
temperatures. The resinated particle mats cured at 130 °C reach approximately the
same bond strength level as the mats cured at 160 °C, but with a significant delay.
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
151
This behaviour is related with the cure temperature of the adhesive system and is
caused by a lower mobility of the reactive groups at lower temperatures. As
mentioned before, a temperature of 90 °C of the UF adhesive systems) in the core
layer of the board is reached after 90 s at 130 °C and after 60 s at 160 °C. If we
draw a straight horizontal line at the reference value of 0.4 N/mm2, we can
determine the decrease of the press time for each resin with the increase of the
press temperature. UF-W6 reaches 0.4 N·mm-2 for 225 s, at 130 °C and for 125 s, at
160 °C. Resins UF-Exp7, UF-R8 and UF-R2 show a decrease of 60 s, 50 s and 75 s in
press time, respectively.
Pressing time / s
0 30 60 90 120 150
IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
UF-R8
UF-R2
Pressing time / s
30 60 90 120 150 180 210 240
IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
UF-R8
UF-R2
a) b)
Figure 6.6. Bond strength curves for particle mats pressed at 130 °C and 160 °C for UF-R8 and UF-R2 (8 % resin and 3 % hardener). a) 130 °C; b) 160 °C.
Effect of adhesive ratio
Figure 6.7 shows the effect of adhesive ratio on internal bond strength
development of resinated particle mats at 130 °C. Two adhesive contents were
analyzed for each resin (8 % and 10 %) at 130 °C and 160 °C.
For resin UF-W6, as expected, higher adhesive ratios result in an increase of bond
strength levels. However, this is not evident for the other resins. Similar results
Chapter 6
152
were obtained at 160 °C (not shown here). This behaviour can be associated with
the low wettability and penetration of the resin produced by strongly-acid process,
due to the lower fraction of low molecular weight species in solution, as previously
discussed.
Pressing time / s
50 100 150 200 250 300
IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
Pressing time / s
40 60 80 100 120 140 160 180IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
Pressing time / s
50 100 150 200 250 300
IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
Pressing time / s
50 100 150 200 250 300
IB / N·m
m-2
0.0
0.2
0.4
0.6
0.8
a) b)
c)d)
Figure 6.7. Bond strength curves for particle mats at 130 ºC with 8 % (•) and 10 % (○) of adhesive (3 % hardener). a) UF-W6 resin, b) UF-Exp7 resin, c) UF-R8 resin, and d) UF-R2 resin.
6.3.2 Automated Bonding Evaluation System (ABES)
Effect of the temperature
The effects of pressing time and testing temperature on the strength of several
cured joints evaluated using ABES for UF-W6 resin are shown in Figure 6.8. For all
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
153
temperatures, the bond strength was tested immediately after each pre-selected
pressing time in shear mode.
As reported by Heinemann [25], it is appropriate to sub-divide the curves into 3
zones. The initial delay in the onset of the bond strength development may be
caused by the loss of energy due to evaporation of water. The adhesive may still be
in the fully liquid state. The second and near-linear stage is likely a direct
consequence of the polymerization reaction in the adhesive systems by chain-
extension and cross-linking processes. In the third and final stage, the curves level
off until they reach maximum values. This behaviour is more evident at 90 °C, 100
°C and 110 °C.
Figure 6.9 shows the values of shear strength for the near-linear stage as a function
of the pressing times for two UF resins at different temperatures.
Pressing time / s
0 100 200 300 400
Shear Strength / N·m
m-2
0
1
2
3
4
5
6
7
80 ºC
90 ºC
100 ºC
110 ºC
130 ºC
Figure 6.8. Shear strength evolution with time for UF-W6 resin at pressing temperature of the 80 °C (•) 90 °C (○), 100 °C (▼), 110 °C (∆) and 130 °C (■).
Chapter 6
154
Pressing time / s
0 50 100 150 200 250
Shear Strength / N·m
m- 2
0
1
2
3
4
c) d)
a) b)
Pressing time / s
0 50 100 150 200 250
Shear Strength / N·m
m- 2
0
1
2
3
4
5
Pressing time / s
0 50 100 150 200 250
Shear Strength / N·m
m- 2
0
1
2
3
4
5
Pressing time / s
0 50 100 150 200 250
Shear Strength / N·m
m- 2
0
1
2
3
4
Figure 6.9. Shear strength evolution with time of four resins at pressing temperature 80 °C (•) 90 °C (○), 100 °C (▼), 110 °C (∆) and 130 °C (■). a) UF-W6 resin; b) UF-Exp7 resin; c) UF-R8 resin and d) UF-R2 resin.
From Figure 6.9, we can calculate the slopes of the linear regions for the four resins
and for all tested press temperatures. Plotting the natural logarithm of the
regressed isothermal bond strength development rate as a function of the
reciprocal of absolute temperature (see Figure 6.10), this relationship can be
represented by a straight line. The correlation coefficient (R2) for resins UF-W6, UF-
Exp7, UF-R8 and UF-R2 was 0.95, 0.91, 0.87 and 0.85 respectively.
Using the Arrhenius equation (Eq. 6.1) relating the rate coefficient and
temperature, we can determine a reactivity index, similar to an activation energy
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
155
(Ea). Eq. 6.2 shows the logarithmic form of Arrhenius equation;
−=
RT
EAexpk
a (6.1)
RT
ElnAlnk a−= (6.2)
Where k is the rate constant, A is the pre-exponential factor, Ea (kJ.mol-1) is the
activation energy, T (K) is the absolute temperature and R (kJ.mol-1·K-1) is the
universal gas constant.
1 / Temperature / K x 10-3
2.5 2.6 2.7 2.8
ln k / N·m
m-2·s-1
2
3
4
5
6 UF-W6
UF-R8
UF-R2
UF-Exp7
Figure 6.10. ABES-derived Arrhenius plots for four adhesive systems.
As Heinemann [25] has reported, the parameter Ea of the adhesive systems can be
obtained from the slope of each regressed line plot in Figure 6.10. The values of Ea
for all resins tested in this work are presented in Table 6.4. These are very similar
and, due to the associated error, it is not possible to distinguish the resins. This
seems to be a limitation of this analysis strategy. A more expedite test, based on
Chapter 6
156
the ABES setup, was implemented to obtain a more effective comparison between
the resins, as described further below.
Table 6.4. Values of the activation energy (Ea) for each adhesive system
Adhesive Ea (kJ·mol-1
)
UF- Exp7 76.7 UF-R8 76.6 UF-R2 71.6 UF-W6 74.8
Effect of hardener ratio
In order to analyse the effect of hardener ratio on bond strength curves, a lower
content of ammonium sulphate, 1.5 %, was tested and compared to the previous
results, with 3.0 %. However, equivalent results were obtained in all resins,
showing that an excess amount of hardener was being used. Figure 6.11 illustrates
the results obtained for resin UF-W6. One may note the good reproducibility shown
by the two data sets.
Pressing time / s
0 20 40 60 80 100 120 140
Shear Srength / N·m
m-2
0
1
2
3
4
5
6
1.5 % of hardener
3.0 % of hardener
Figure 6.11. Shear strength evolution with time UF-W6 resin with two hardener ratios, 1.5 % (•) and 3.0 % (○), at 110 °C.
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
157
A simplified comparison method for different resins
Using pressing conditions relatively close to those generally observed in industrial
hot-presses (100 °C in core layer and 80 s of pressing time), ten samples of each
resin were tested on the ABES setup. In Figure 6.12 the average shear strength
obtained for each resin is presented. It can be observed that UF-Exp7 presents the
best performance of all evaluated resins. UF-R2 shows the lowest value. In contrast
to the results obtained by IPATES, the performance of the resins UF-W6 and UF-R8
is very similar for the operation conditions tested and the error intervals are
overlapping. These results suggest that the reactivity of UF resins, as measured by
the gel time method, is not linked with the “ABES-performance” of UF resins. The
gel time value of the resins may be an unreliable predictor of bonding
performance.
As mentioned above, the main limitations of the resin produced by the strongly-
acid process is the lower fraction of monomeric species. However, in the case of
ABES tests, the impact of lower wettability of the adhesive system is smaller than
with the IPATES tests. Indeed, in the latter the wetting and penetration of the resin
on the wood particles takes place in a rotary mixer, where the resin is sprayed onto
the wood particles prior to forming and pressing the test specimen. On the other
hand, in the ABES test a measured amount of resin is placed on one veneer strip
and the second strip is rubbed over it in order to force a uniform distribution of the
resin throughout the entire contact area of the joint.
This expedite test method may be used to evaluate the performance of an adhesive
system, but further data must be collected in order to validate its results.
Chapter 6
158
UF-W6 UF-EXP 7 UF-R8 UF-R2
Shear strength / N·mm
- 2
0
1
2
3
Figure 6.12. Breaking load in shear mode for four resins by the ABES method at 100 ºC and 80 s (ten tests were performed for each resin to evaluate the error).
6.4 Conclusions
Four UF resins were produced using two different processes (alkaline-acid and
strongly-acid process). For evaluating the cure of these adhesive systems, two
recent characterization methods were used (IPATES and ABES). These allowed
describing the behaviour of each adhesive system under different conditions,
namely pressing temperature, pressing time, adhesive ratio (only for IPATES) and
hardener ratio (only for ABES).
Generally, the results obtained by IPATES showed that the UF-Exp7 resin produced
by alkaline-acid process always presents a better performance as compared with
the resin produced by strongly acid process.
F/U molar ratio of the resins produced by alkaline-acid process has an effect on
performance of the adhesive systems. The resin UF-R8 (F/U = 1.12) always presents
a better performance with UF-R2 (F/U = 1.00) in IPATES tests.
A quick test methodology implemented on ABES equipment, showed that the resin
Evaluation of UF adhesives performance by IPATES and ABES mechanical tests
159
produced by alkaline-acid process yields the best values of shear strength at 100 °C
and 80 s of press temperature and press time, respectively. Again, the results show
that the resin produced by alkaline-acid process give the best results, compared
with resin produced by the strongly-acid process.
6.5 References
[1] A. Pizzi and K. L. Mittal, Handbook of Adhesive Technology, Marcel Dekker, New York (2003).
[2] P. Christjanson, T. Pehk and K. Siimer, Journal of Applied Polymer Science 100, 1673 (2006).
[3] L. Graves and J. Mueller, US Patent 5362842 (1994). [4] A. Hatjiissaak and E. Papadopoulou, WO Patent 138364 (2007). [5] H. Kong, US Patent 4603191 (1986). [6] W. K. Motter and D. M. Harmon, WO Patent 2006-015331 (2006). [7] H. Spurlock, US Patent 4381368 (1983). [8] S. Vargiu, S. Giovanni, G. Mazzolen and U. Nistri, US Patent 3842039 (1974). [9] R. Whiteside, US Patent 4968773 (1990). [10] J. H. Williams, US Patent 4410685 (1983). [11] J. H. Williams, US Patent 4482699 (1984). [12] J. M. Ferra, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho,
Polymer International (2010). [13] J. Schultz and M. Nardin, Theories and Mechanisms of Adhesion. In Handbook
of Adhesive Technology, A. Pizzi and K. L. Mittal, Eds. Marcel Dekker: 2003. [14] A. Pizzi, Advanced wood adhesives technology, Marcel Dekker, New York
(1994). [15] G. G. Marra, Technology of wood bonding: principles in practice, Van
Nostrand, New York (1992). [16] S. Chow and P. R. Steiner, Holzforschung 29, 4 (1975). [17] G. E. Myers and J. A. Koutsky, Holzforschung 44, 117 (1990). [18] M. Szesztay, Z. Laszlohedvig, E. Kovacsovics and F. Tudos, Holz als Roh- und
Werkstoff 51, 297 (1993). [19] R. O. Ebewele, B. H. River and G. E. Myers, Journal of Applied Polymer Science
52, 689 (1994). [20] C. Soulard, C. Kamoun and A. Pizzi, Journal of Applied Polymer Science 72, 277
(1999). [21] S. Yin, 1994. Charactérisation par analyses thermiques de la polycondensation
d’adhésifs aminoplastes et du durcissement de composites modèles bois-adhésif, Ph.D. Thesis, Université de Nancy I, Nantes, France.
[22] K. Umemura, S. Kawai, Y. Mizuno and H. Sasaki, Mokuzai Gakkaishi 42, 489
Chapter 6
160
(1996). [23] M. Sernek and F. A. Kamke, International Journal of Adhesion and Adhesives
27, 562 (2007). [24] P. E. Humphrey, US Patent 5176028 (1993). [25] C. Heinemann, 2004. Characterization of the curing process of adhesives in
wood-particle matrices by evaluation of mechanical and chemical kinetics, Ph.D. Thesis, University of Hamburg, Hamburg.
[26] M. Dunky, Analysis of Formaldehyde Condensation Resins for the Wood Based Panels Industry: Status and new challenges. In 1st
European Panels Products
Symposium, Hague, Loxton, Bolton and Mott, Eds. Llandudno, North Wales, UK, 1997; p 217.
[27] J. M. Ferra, P. C. Mena, J. Martins, A. M. Mendes, M. R. N. Costa, F. D. Magalhães and L. H. Carvalho, Journal of Adhesion Science and Technology 24, 1455 (2010).
[28] G. He and N. Yan, Holzforschung 59, 635 (2005). [29] M. Dunky, International Journal of Adhesion and Adhesives 18, 95 (1998).
CHAPTER 7
General Conclusions and Future Work
163
7 General Conclusions and Future Work
7.1 General Conclusions
The objective of this doctoral work was to study and optimize the properties of
urea-formaldehyde (UF) resins used in the production of wood-based composites.
This work evolved in close cooperation with Portuguese company EuroResinas, a
producer of UF resins.
The first studies were focused on the analysis of the colloidal character of UF
resins, complementing the somehow unconsolidated information existing in the
open literature. Based on the results obtained, two hypothesis were formulated
about the stabilization of the colloidal suspension of UF resins: the first states that
unreacted urea may form a solvation layer surrounding the particles, retarding
agglomeration; the second one is based on stabilization by electrostatic repulsion
associated to ionic species retained at the particle surface. Additionally, tensile
strength tests performed on wood samples glued with the insoluble and soluble
phases indicated that the dispersed phase may be relevant for strengthening the
adhesive bond, acting as a reactive reinforcing filler.
In order to characterise the UF resins, two methods were implemented, using GPC/
SEC and HPLC, which permitted to obtain insight on the molecular weight
distribution of the polymer formed, and on the fractions of unreacted urea,
monomethylolurea and dimethylolurea. These techniques were applied on the
characterization of industrial resins, produced with different formulations.
A design of experiments tool was used to optimize the alkaline-acid synthesis
process, acting on the conditions of the condensation reaction. A response surface
methodology permitted to evaluate the effect of three selected factors (number of
urea additions, the time span between additions, and the pH) on resin performance
Chapter 7
164
and also to maximize the internal bond strength and minimize the formaldehyde
release on particleboard. It was concluded that the pH and the time span between
consecutive urea additions in the condensation step have a strong influence on the
analysed properties. An optimized resin was produced in laboratory and the
properties obtained were within or close to the values predicted by the empirical
model. This resin presented a high performance when compared to a standard
resin.
In the last part of the thesis, a new synthesis process was studied and tested. The
strongly acid process is a complex process due to the simultaneous formation of
methylolureas and condensation products. The determination of molecular weight
distributions by GPC/SEC indicated that the resin produced by this process has a
significant fraction of high molecular weight (or molecularly aggregated) polymer,
in direct contrast to the alkaline-acid process.
The mechanical performance of laboratory and commercial resins was evaluated at
vTi (Johann Heinrich von Thünen-Institut) institute in Hamburg, Germany, using
two recent methods (IPATES and ABES). The results allowed to assess the
mechanical behavior of each adhesive system under different conditions, namely in
terms of pressing temperature, pressing time, adhesive ratio (only for IPATES) and
hardener ratio (only for ABES). The resin selected in the synthesis optimization
work mentioned before presented the best results.
A major difficulty encountered during the work was the still reduced amount of
information available in the open literature on UF resins. This is a complex system,
both in terms of reaction mechanisms and performance evaluation. This doctoral
work has brought relevant practical knowledge regarding these issues, specially in
terms of UF resin characterization methodologies and on the influence of operation
conditions on the product properties. This may provide a solid support for new
projects on formaldehyde derived wood adhesives within the research group.
General Conclusions and Future Work
165
7.2 Future Work
Although the use of UF resins in wood-based panels industry has been questioned
in the last years, mostly due to environmental concerns related to formaldehyde
emission, these materials are not expected to be replaced by another kind of
adhesive in the immediate future. Their advantages (in terms of raw materials and
production costs, and performance) are still very relevant. Thus, it is necessary to
develop solutions for reducing formaldehyde emissions while maintaining the
mechanical performance of the wood composites. One strategy may be the
incorporation of other substances in the polymeric structure, in order to improve
moisture resistance and decrease formaldehyde emissions. Existing examples are
the use of soy protein, sugars and melamine.
Regarding the synthesis procedures, a deeper study of the strongly-acid process
should be performed. The advantages described in the literature for this approach
have not been confirmed by the results obtained in the preliminary tests carried
out in this thesis. The impact of the strongly-acid process in the polymeric
structures formed during synthesis and in the quality of the final produced resins
should be accessed.
Finally, it is important to implement a more advanced strategy for on-line control
of the polymerization reaction, in industrial context, allowing for a higher
reproducibility of the synthesis process. In particular, the application of near-
infrared (NIR) techniques can be used to monitor/control the addition and
condensation reactions in the synthesis of UF resins. One practical example of
industrial use of this technique has been reported by Chimar Hellas, a Greek
company, in 2007.