UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2020
Casein-magnesium composite as an
intumescent fire retardant coating for wood
Uddin, Mezbah
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
© Elsevier Ltd.
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.firesaf.2019.102943
https://erepo.uef.fi/handle/123456789/7983
Downloaded from University of Eastern Finland's eRepository
Journal Pre-proof
Casein-magnesium composite as an intumescent fire retardant coating for wood Mezbah Uddin, Kalle Kiviranta, Sari Suvanto, Leila Alvila, Jari Leskinen, Reijo Lappalainen, Antti Haapala
PII: S0379-7112(19)30135-3
DOI: https://doi.org/10.1016/j.firesaf.2019.102943 Reference: FISJ 102943
To appear in: Fire Safety Journal Received Date: 17 March 2019 Revised Date: 20 December 2019 Accepted Date: 27 December 2019
Please cite this article as: M. Uddin, K. Kiviranta, S. Suvanto, L. Alvila, J. Leskinen, R. Lappalainen, A.
Haapala, Casein-magnesium composite as an intumescent fire retardant coating for wood, Fire Safety Journal (2020), doi: https://doi.org/10.1016/j.firesaf.2019.102943.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier Ltd.
1 1
Casein-magnesium composite as an intumescent fire retardant coating for wood 2
Mezbah Uddina, Kalle Kivirantab, Sari Suvantoc, Leila Alvilac, Jari Leskinend, Reijo 3
Lappalainend,e, Antti Haapalaa*
4
a School of Forest Sciences, University of Eastern Finland, FIN-80100, Joensuu, Finland 5
b Savonia University of Applied Sciences, FIN-70101, Kuopio, Finland 6
c Department of Chemistry, University of Eastern Finland, FIN-80100, Joensuu, Finland 7
d SIB Labs, University of Eastern Finland, Kuopio Campus, FIN-70211, Kuopio, Finland 8
e Department of Applied Physics, University of Eastern Finland, FIN-70211, Kuopio, Finland 9
⁎ Corresponding author: antti.haapala@uef.fi 10
11
ABSTRACT 12
Fire hazard associated to wood and wood-based products is a clear shortcoming of natural 13
building materials and hence the use of fire retardants is mandated by national building codes.
14
In this study, we have synthesized a simple and green fire retardant composite using 15
magnesium hydroxide (Mg(OH)2) with casein protein and tested its performance as self- 16
adhesive coating paste. The thermogravimetric analyses indicated that casein-magnesium 17
composites are able to produce thermally stable insulating magnesium oxide (MgO) that 18
suppresses the fire spread as a result of the synergistic combination of physical and chemical 19
actions—the intumescent action of casein, formation of char and thermally stable MgO, and 20
the simultaneous release of water vapor retained by the casein. The most effective 21
performance was obtained with a sample (coating weight of 1080 g/m2) for which the ignition 22
2 time of pinewood substrate was increased by 147%, while the smoke production rate and the 23
peak heat release rate were decreased by 53% and 30%, respectively.
24 25
Keywords 26
Casein; Flame retardancy; Magnesium hydroxide; Thermal analysis; Wood 27
Declaration of interest: none 28
29
1. Introduction 30
Wood is naturally a fire sensitive material because of its thermally active carbohydrate 31
polymeric constituents. Commercial fire retardants are abundantly available for wood products 32
as paints and surface coatings for e.g. in places where rigid gypsum boards and other 33
structural protections are not feasible but issues on performance, environmental sustainability 34
and cost remain to be improved for many contemporary solutions. For instance, boron-based 35
products have reportedly suffer from leaching problems [1], and the halogenated fire 36
retardants produce halogen gases known to potentially cause respiratory issues [1-5].
37
Similarly, polyurethane-based fire retardants release hydrogen cyanide and carbon monoxide, 38
which are toxic [6].
39
Several studies have been on modified wood materials [7,8], wood-plastic composite 40
materials [9,10] and halogen-free fire retardant compounds as coatings and nano-coating 41
agents. These include clay nanopapers [11] composites of waterborne polyurethane (WPU), 42
aluminum trihydroxide (ATH) and mica [12], TiO2/ZnO coatings [13] and exfoliated vermiculite- 43
sodium silicate composites [14]. Other approaches are also used to insert fire retardant 44
chemicals into wood. For example, the polymerization of aniline in wood veneers [15] and 45
3 impregnation using a mixture of ammonium polyphosphate, magnesium hydroxide and 46
aluminum hydroxide [16].Recently, Merk et al. [17,18] investigated the mineralization of wood 47
by calcium carbonate using different approaches for improving fire retardancy and Xiao et al.
48
[19] reported reduction of both heat release and smoke production using modified silica.
49
Similarly the uses of graphene [24], nanoclay [25] and synthetic sulfonamides [20] have been 50
reported. To reduce the flammability behavior of wood with non-toxic additives is a potentially 51
valuable research topic for the wood-converting industry [21] while similar development is 52
seen on thermoset resins [22,23] and other materials [24-28].
53
The most ecofriendly nitrogen- and phosphorous-containing fire retardants (FRs) are the 54
casein [29] and whey proteins [30], deoxyribonucleic acid (DNA) [32] and DNA with chitosan 55
[33]. Caseins have traditional applications outside the food sector as e.g. binding materials as 56
crosslinking agents for tissue engineering [29]. Casein belongs to a family of phosphoproteins 57
containing 0.7-0.9% of phosphorous, which exist with protein through a serine ester linkage 58
called phosphoserine. The solubility of these calcium phosphor-caseinate clusters depends on 59
pH and casein micelles typically form stable colloidal dispersions with very few disulfide bonds 60
[31]. As caseins consist of phosphorous they can produce phosphoric acid to dehydrate 61
polymers. This is applied in the fire retardancy of textiles [29,32], where casein is shown to 62
promote char formation through dehydration of cellulose, however, to obtain significant 63
performance, rather high concentrations in relation to the mass of protected substrate are 64
required. They also generate nitrogen-phosphorus synergy in the combustion process [33,34].
65
Magnesium hydroxide is used as a fire retardant, which is attributed to the endothermic 66
decomposition into magnesium oxide and water that causes ignition to delay because energy 67
from heat source is absorbed by the reaction. Concurrently the release of water vapor dilutes 68
4 the flammable gases that are generated and the produced magnesium oxide can function as a 69
thermal barrier [35].
70
The study focuses on to develop approaches to synthesize a novel and cost-effective fire 71
retardant (FR) composite by preparing casein protein suspension in the presence of sodium 72
hydroxide (NaOH) followed by reacting with magnesium hydroxide (Mg(OH)2). The composite 73
provides multiple modes of actions to protect fire spread with relatively low content of casein.
74
The Scots pine sapwood was coated using prepared composite paste from 700 to 1080 g/m2. 75
The penetration depth of the coating components onto wood and formation of chemical bonds 76
between coating composite constituents was also investigated. The thermal stability of the 77
composite was investigated using thermogravimetry (TG) and differential thermogravimetry 78
(DTG). Cone calorimetry and TG were used to assess the self-extinguishing property and 79
thermal stability of the coated wood samples.
80
2. Experiments 81
2.1. Materials 82
Scots pine (Pinus sylvestris L.) sapwood samples (100 × 100 × 20 mm) were conditioned at 83
(22°C and 50% relative humidity RH) for a minimum of two weeks before cutting and coating.
84
Sodium hydroxide (NaOH) (≥ 97% pellet; Merck, Burlington, USA), casein from bovine milk 85
(technical grade), and Mg(OH)2 (reagent grade, 95%), both from Merck KGaA (Darmstadt, 86
Germany), were used directly without further purification. Deionized water was used during the 87
experiment when needed.
88 89
2.2. Sample preparation 90
First, 50 grams of water in a 250 ml beaker was heated to 60°C and 10 g of casein 91
introduced, while stirring continuously. Then 0.4 g of NaOH was added while stirring for 10 min 92
5 to facilitate even mixing and formation of sodium caseinate salt to complete the base mix.
93
Different composite formulations were prepared by adding 4 to 16 g of Mg(OH)2 (see Table 1) 94
into the suspension following continuous mixing for 10 min at 70°C to form a composite paste.
95
In preparing the coatings for 700, 775, 990 and 1080 g/m2 coating basis weights doses of 4, 8, 96
12, and 16 grams of Mg(OH)2 were used. The reaction mixture was stirred for 5 min to cool, 97
and then the viscous composites were cast onto two wood pieces, each measuring 98
100×100×20 mm and climatized at 22°C and 50% relative humidity (RH). The composites 99
were allowed to cover the specimen surfaces and the coating layer was made uniform with a 100
help of an adjustable coating blade (Fig. 1). Coated samples were dried at room temperature 101
for 48 h, and then kept at 22°C and 50% RH in a fixed humidity chamber for 7 days.
102
103
Fig 1. Reference (left) and a sample of coated Scots pine sapwood.
104 105
The weight of coating applied onto the samples (g/m2) was calculated by subtracting the 106
weight of uncoated sample from the weight of the coated and dried sample. The composition 107
and applied weight of coatings onto wood (g/m2) are summarized in Table 1.
108 109
Table 1. Composition of the samples coating and their weight onto each sample 110
Sample # Composition of coating
composite Coating applied (g) Coating basis weight (g/m2)
6
Reference - - -
Sample 1 4 g Mg(OH)2, 0.4 g NaOH, 10 g
casein 7 700
Sample 2 8 g Mg(OH)2, 0.4 g NaOH, 10 g
casein 7.7 770
Sample 3 12 g Mg(OH)2, 0.4 g NaOH, 10
g casein 9.9 990
Sample 4 16 g Mg(OH)2, 0.4 g NaOH, 10
g casein 10.8 1080
111 112
2.3. Surface analysis 113
The surface of the prepared composites was investigated with a FTIR spectrometer Vertex 114
70 (Bruker, Leipzig, Germany) equipped with an ATR platinum diamond, a sensitive 2 × 2 mm 115
diamond crystal surface, and a sample detector RT-DLaTGS. The same ATR-FTIR parameters 116
recorded within the range 4000–650 cm-1 were used in the OPUS 6.5 software (Bruker, 117
Leipzig, Germany) throughout the measurements.
118 119
2.4. Microanalysis 120
The penetration capacity of the coating components and the morphology of the coated 121
sample were investigated using scanning electron microscopy-energy-dispersive X-ray 122
spectroscopy (SEM-EDS). A cross-section, including the borderline of the wood and coating, 123
was observed with a Zeiss SigmaHD|VP Field Emission SEM (Carl Zeiss NTS, Cambridge, 124
UK). Also, the Mg concentration was identified on both sides of the interface between the 125
coating and the wood substrate with EDS (Thermo NSS; Thermo Fisher Scientific, Madison, 126
WI, USA) line scan of 20 evenly distributed points along a 1.2 mm line. A small chip of samples 127
7 was attached to a standard 12 mm aluminum stub for SEM specimens using a piece of 128
double-sided copper adhesive tape and imaged without any sputter coating in low vacuum, N2 129
atmosphere at 30 Pa pressure using 10 kV acceleration voltage, a backscatter electron 130
detector, and a working distance of 8.3 mm.
131 132
2.5. Thermal gravimetric analysis 133
The thermal properties of the samples were studied using a thermogravimetric analyzer 134
(TGA/STDA 851e/LF/1100; Mettler Toledo, Greifensee, Switzerland). Approximately 10 mg of 135
dry coating were taken from the coated wood samples, along with a thin layer of wood 136
material, using a sharp blade to be used for TGA measurement. Thermogravimetric 137
measurement was run under a nitrogen flow rate of 50 ml/min, a heating rate of 10°C/min, and 138
the temperature range of 25 to 800°C. The sample (approx. 9–10 mg) in solid form was kept in 139
an open alumina pan (70 µl).
140 141
2.6. Cone calorimetric analysis 142
Flammability behavior of the samples was analyzed with a cone calorimeter (ConeTool 1;
143
SGS Govmark Ltd., Farmingdale, IL, USA) at a radiant flux of 50 kW/m2 based on the ISO 144
5660-2 Standard Test Method. The heat release rate was measured based on the oxygen 145
consumption and the flow rate in the combustion product stream. The samples were set up 146
horizontally in the sample holder, and only the surface of the samples was exposed.
147
148
3. Results and discussion 149
3.1. Surface properties of the composites 150
8 The following FTIR-ATR spectra (Fig. 2) belong to the surface of four different casein- 151
magnesium hydroxide composites. The peak at 3695 cm-1 corresponds to the stretching of OH 152
in the Mg(OH)2 [36].The broad bands at 3279 cm-1 are associated with the stretching of OH in 153
the protein structure. The band at 2920 cm-1 corresponds to the stretching of CH2 in the protein 154
skeleton. The large peak at 1632 cm-1 is attributed to the stretching of the C=O of the amide 155
bond, the so-called amide I group (alpha helix and beta sheet structure of proteins), while the 156
peak at 1531 cm-1 refers to the combination of the stretching of C-N and deformation bending 157
of N-H, the so-called amide II groups [37]. The signal at 1082 cm-1 corresponds to the 158
stretching of the C-O in the COO- groups. The band at 1238 cm-1 is attributed to the stretching 159
of the P=O in phosphodiesters based on the information from ref. [38] but it could be also 160
related to C-N bond.
161
162
Fig 2. FTIR-ATR absorbance spectra of casein and four prepared composites.
163 164
9 The intensity of all the assigned peaks decreased with the increasing concentration of 165
Mg(OH)2, the sole exception being the stretching of OH at 3695 cm-1. The reduction of 166
intensities indicates that magnesium may interact with the different active moieties of casein, 167
but only weakly as no chemical bond formation is observed. For instance, Mg may interact 168
with casein protein through the strong long-range interaction of PO2 in the phosphate of the 169
casein. In addition, Mg may interact with the carbonyl C=O and C-N bonds of casein. Along 170
with these, Mg may also interact with the basic region (amine) of the casein protein [39].
171
Another reason for the decreasing intensity of the peaks is that Mg(OH)2 covers the functional 172
moieties of the casein – however, the precise interfacial phenomena between the principal 173
components is considered beyond the scope of this study. Further study will focus on in the 174
direction of the precise interfacial phenomena between the principal components, as 175
mentioned here the focus was to illustrate the performance of the achieved composite.
176 177
3.2. Coating component penetration into wood 178
The following images show the morphology and elemental depth profiles of the coated wood 179
samples. In Fig. 3a, the arrangement of mineral coating particles on top of the wood substrate 180
is shown. The coating seems rather uniform, but the layer thickness has some variation, as it 181
was not optimized; rather, the content of solids in the coating chemical was controlled. The Mg 182
profile measured with EDS is highlighted in white. The Mg concentration of 15.9 ± 3.1% (w/w) 183
was measured within the coating (1080 g/m2 coating). The measured Mg concentration profile 184
revealed that the studied fire retardant had penetrated about 560 µm into the wood.
185
The Mg concentration in the wood substrate varied locally from 2 to 12%. However, the local 186
minimum concentrations were measured at the hollow points on the surface. Therefore, it is 187
obvious that those values were greatly underestimated (Fig. 3a). The EDS analysis pathway 188
10 from wood to coating over their interface is shown with a yellow arrow. In Fig. 3b, the 189
composition of identified elements in the wood and coating layers is shown. The proof of 190
MgOH penetration into the wood surface was obtained with EDS as the Mg concentration 191
dropped to 0.7% or less below 500 µm in the wood, which, in practice, can be considered as 192
retardant free wood.
193
194
Fig 3. a) SEM image from the coating interface shows the concentration gradient of Mg with 195
some penetration into the wood surface. The yellow arrow shows the pathway measured by 196
EDS analysis, along which b) the elemental distribution of C, O, and Mg were measured.
197 198
3.3. Thermal stability analysis 199
The thermal stability of the samples was investigated using thermogravimetry (TG) and 200
differential thermogravimetry (DTG), as shown in Fig. 4a and 4b. In the case of the untreated 201
sample (only wood), weight loss occurred in three steps as observed from the both TGA and 202
DTGA curves. Bound water and volatile organic compounds are lost in the first step;
203
hemicellulose and cellulose are decomposed into CO2, CO, CH4, CH3OH, and CH3COOH in 204
11 the second step, called the charring step; and, finally, lignin is decomposed and char residue is 205
oxidized in the third step, called the calcining step [40,41].
206
In the case of the treated samples, weight loss occurred in four steps. First, water and 207
organic volatiles are lost as in the untreated sample. In the second step, hemicellulose and 208
cellulose are decomposed; weight loss is lower than for the untreated reference, which could 209
be due to the formation of synergy, which may modify the pyrolysis process by interrupting the 210
formation of the main degradation product levoglucosan from cellulose that facilitates thermally 211
stable char formation. Casein begins to decompose at 270°C and may start to form phosphoric 212
acid that may promote char formation [29].
213
In the third step, Mg(OH)2 starts to decompose at just under 350°C and forms thermally 214
stable MgO that protects the wood surface from burning. From the DTGA, it is obvious that 215
samples containing more Mg(OH)2 lose more weight at 415°C; this indicates that the more 216
substantial MgO layer formation effectively blocks the heat transfer. In the fourth step, lignin 217
starts to decompose at 425°C; however, the mass loss of all the samples coated with 218
composites is lower than that of the untreated reference because of the char and MgO layer 219
formation.
220
221
12 Fig 4. a) TG and (b) DTG curves of untreated wood and samples treated with coating 222
composites.
223 224
3.4. Flame retardant performance 225
To assess the flame retardancy of the samples, we monitored time to ignition (TTI) in 226
seconds, peak heat release rate (HRR) in kW/m2, total heat released (THR) in MJ/m2, average 227
smoke production (ASP) in m2/s, and total mass loss (TML) in g. The cone calorimetric test 228
was performed three times for each sample and the average value was used to assess the fire 229
retardancy. The results are shown in the following Table 2.
230 231
Table 2. Cone calorimetry results of the reference and coated wood (TTI: Time to ignition, 232
Peak HHR: Peak heat release rate, THR: Total heat released, TML: Total mass loss, ASP:
233
Average smoke production) 234
Sample TTI (s) Peak HHR (kW/m²) THR (MJ/m²) TML (g) ASP (m2/s)
Ref.* 12.1 216.0 79.5 46.6 0.0047
700 g/m2 11.5 147.7 64.3 42.4 0.0035
770 g/m2 17.5 140.0 61.9 41.5 0.0031
990 g/m2 26.3 127.3 56.1 42.5 0.0035
1080 g/m2 30.4 119.0 53.3 40.8 0.0022
* The reference used was uncoated pinewood 235
236
3.4.1. Time to ignition 237
TTI of all the samples was higher than the reference except for sample containing 700 g/m2 238
coating in Table 2, probably to inherent variation in wood composition. Sample containing the 239
highest amount of composite (1080 g/m2) delayed fire spreading significantly, which could be 240
due to the release of more water vapor during combustion which cools the surface of the 241
13 substrate. During the combustion process Mg(OH)2 also forms thermally stable magnesium 242
oxide (MgO) on the coating layer. Coatings have an intumescent functionality due to the 243
presence of casein, which prevents wood from burning by retaining inflammable gasses 244
between the wood and coating, like in [42], where the components produced a synergistic 245
effect in extending the ignition time. The synergistic effect may be caused through the complex 246
formation between casein and magnesium hydroxide. Divalent magnesium ion (Mg2+) forms 247
octahedral complex during reaction with casein protein, in which four bonds form with four 248
ligands of casein such as either aspartic acid (E) or glutamic acid (D). Another bond forms with 249
carboxylate oxygen atom of either glutamic acid (E) or aspartic acid (D), which functions as 250
monodentate ligand. One further bond forms with water molecule shown in the following figure 251
[45-47].
252
253
Fig 5. The interaction of magnesium with casein protein.
254 255
Trey et al. [15] found no improvement of TTI when applying polyaniline in veneers. Pan et al.
256
[16] similarly failed to show improvement in TTI by adding Mg, Al, or Si derivatives to nitrogen- 257
phosphorus fire retardants. Merk et al. [17] also reported no change in TTI for CaCO3 258
mineralized beech and spruce wood. Chang et al. [12] found that a WPU coating on wood 259
14 panels benefited from ATH and mica as fillers. The addition of inorganics resulted in an 260
extended TTI from 34 sec with WPU to 75 sec with ATH (WPU/ATH in 10/100 g), and the best 261
sample, with 15 g ATH and 85 g mica mixed in 10 g WPU, did not ignite at all. Their coating 262
system thickness was, in comparison approximately four times that of Sample 4 presented 263
here, making direct comparison difficult. However, a thicker coating generally provides better 264
flame retardant performance.
265
In addition, the study with a layer-by-layer coating using vermiculite and sodium silicate 266
(double layer) on wood substrate observed that the TTI was increased by 494% [14]. Film with 267
clay nanopaper increased the TTI of wood to 4.5 minutes; however, the heat flux in this study 268
differed from that of our study, being only 35 kW/m2 [11]. Although the performance is difficult 269
to compare directly due to differing test setups the displayed casein-magnesium composites 270
can clearly prolong the durability of wood in the presence of fire.
271 272
3.4.2. Heat release rate 273
While burning samples with a cone calorimeter, the untreated wood sample produces a 274
mixture of combustible gases, releases substantial heat, and forms a large exothermic peak 275
with a heat release rate (HRR) of 216 kW/m2. However, three exothermic peaks are formed for 276
all the treated samples: The first peak formed might be due to the intumescent behavior of the 277
coatings. When the treated samples are burned, Mg(OH)2 decomposes and produces water 278
vapor and MgO. This results in released heat and forms the second exothermic peak with a 279
low peak HRR. In the third combustion step, the char layer is decomposed, resulting in tiny 280
cracks that facilitate the evaporation of volatiles and release heat (the third exothermic peak).
281
The results indicate that the HRR of all the samples treated with composites containing 282
Mg(OH)2 is lower than the reference sample (Fig. 6) and decreases with the increasing 283
15 concentration of Mg(OH)2. The reason for the reduction of the HRR of the samples treated with 284
composites is the formation of thermally stable MgO and release of water from Mg(OH)2 285
decomposition inside the protective swelling coating layer of casein. Thermally stable MgO 286
may block the heat transfer through the material surface and results in char formation [12].
287
MgO may also block the heat transfer via suppressing the spread of oxygen gas and 288
combustible volatile compounds [16].
289
290
Fig 6. HRRs of the uncoated wood and coated wood samples.
291 292
3.4.3. Total heat released 293
The total heat released (THR) refers to the total amount of heat released from the materials 294
during combustion. Results indicate that the THR of all the treated samples is lower than that 295
of the untreated sample and that the THR is decreased with increasing the concentration of 296
Mg(OH)2. This is due to the fact that the samples containing more Mg(OH)2 produced a 297
considerable amount of thermally stable magnesium oxide which may absorb the heat.
298
Furthermore, water released from the decomposition of Mg(OH)2 is absorbed during 299
16 combustion, resulting in the total heat reduction [43,44]. The THR of coating containing 1080 300
g/m2 of composite was decreased by 41%.
301 302
3.4.4. Total mass loss 303
The total mass loss (TML) prior ignition is related to the moisture content of the coatings and 304
the water content of the wood itself [14]. The pine wood as a coating substrate contained ca.
305
9.5% of water (following the conditioning at RH 50%) and the moisture content in the coating 306
composites in these conditions was 4.5% of water (w/w). In terms of coated wood specimen, 307
the difference of moisture content between the lowest and the highest basis weight coatings (7 308
and 10.8 grams in absolute values) was hence less than 0.2 grams, which is not considered to 309
contribute significantly to their perceived mass loss. The TML of all the samples containing 310
Mg(OH)2 is lower than that of the untreated samples, and the TML of sample with the most 311
coating was the lowest (Table 2), which may be due to the higher char and thermally stable 312
magnesium oxide (MgO) formation (Fig. 4). Merk et al. [17] observed no significant change in 313
the TML for the CaCO3 mineralized beech and spruce wood. Several reasons can explain the 314
inconsistency of mass loss and smoke production trends observed between 770 and 990 g/m2 315
coatings. For instance, the anisotropic structure and possible differences in the wood substrate 316
composition, or small defects in some of the coating surfaces of 990 g/m2 specimen may have 317
contributed, but which we were regardless unable to identify from burnt samples.
318
Images of the surface of the samples after testing with cone calorimeter are shown in Fig. 7.
319
The char formation on the surface of the coated samples is clearly visible, but differences 320
among the coated samples were not significant, as the variation of mass loss between 321
samples was also limited.
322
17 323
Fig. 7. Untreated pinewood as a reference and the composite coated and charred pinewood 324
samples after a 600-second cone calorimetry test using a heat flux of 50 kW/m2. 325
326
3.4.5. Smoke production 327
The ASP decreases with increasing the concentration of Mg(OH)2. All the samples treated 328
with composite release less smoke than the untreated sample. During the combustion 329
process, casein-magnesium composite produced MgO layers having a high surface area, 330
which facilitates the absorption of smoke and smoke producer gasses [16]. Among all the 331
samples, sample containing 1080 g/m2 of composite releases the lowest amount of smoke, 332
which is decreased by about 53% compared with the untreated wood. The synergistic effect of 333
magnesium hydroxide and casein may suppress the smoke production similar to the study of 334
the wood sample impregnated with the mixture of 25% nitrogen–phosphorous fire retardant 335
and 10% magnesium and aluminum hydroxides [16].
336
4. Conclusions 337
Casein functions as an adhesive onto the wood surface and provides active sites for 338
interaction with magnesium while a pure casein suspension in reasonable concentration does 339
not improve the fire retardancy of wood and Mg(OH)2 alone lacks adhesive capacity to wood.
340
The deposition of magnesium in the pores of the pinewood at a 560 µm depth was confirmed 341
by SEM-EDS. Intumescent activity of the coating arises from casein, while the decomposition 342
of magnesium hydroxide at 415°C into thermally stable magnesium oxide and release of water 343
18 vapor provide additional protection. The cone calorimeter results showed that all the prepared 344
composites function as fire retardants and that the best performance was obtained with a 345
sample containing 1080 g/m2 of composite, extending TTI from 12 to 30 sec, while also 346
decreasing peak heat release from 216 to 119 kW/m2 and ASP from 0.005 to 0.002 m2/s, 347
respectively. Along with the thermally stable magnesium oxide and water vapor present, the 348
synergy obtained from casein and magnesium hydroxide composite coating can play a role for 349
improving the fire retardancy of wood products.
350 351
Acknowledgments 352
Teollisuusneuvos Heikki Väänänen’s fund and the North Karelia regional foundation of the 353
Finnish Cultural Foundation are acknowledged for their grants to support this work as is the 354
FORES graduate school.
355
Notes 356
The authors declare that there are no competing financial interests concerning this work.
357
References 358
[1] C. Thomsen, E. Lundanes, G. Becher. Brominated flame retardants in plasma samples from three different
359
occupational groups in Norway. J. Environ. Monit. 3 (4) (2001) 366–370.
360
[2] C. Thomsen, H. Leknes, E. Lundanes, G. Becher. A new method for determination of halogenated flame-
361
retardants in human milk using solid-phase extraction. J. Anal. Toxicol. 26 (3) (2002) 129–137.
362
[3] I. Watanabe, S. Sakai. Environmental release and behavior of brominated flame-retardants. Environ. Int. 29
363
(6) (2003) 665–682.
364
[4] M. Osako, Y. Kim, S. Sakai. Leaching of brominated flame-retardants in leachate from landfills in japan.
365
Chemosphere 57 (10) (2004) 1571–1579.
366
19
[5] T. Debenest, F. Gagné, A. Petit, C. André, M. Kohli, C. Blaise. Ecotoxicity of a brominated flame retardant
367
(tetrabromobisphenol A) and its derivatives to aquatic organisms. Comp. Biochem. Physiol. C Toxicol.
368
Pharmacol. 152 (4) (2010) 407–412.
369
[6] S.T. McKenna, T.R. Hull. The fire toxicity of polyurethane foams. Fire Sci. Rev. 5 (3) (2016) pp. 27.
370
[7] J. Jiang, J. Li, J. Hu, D. Fan. Effect of nitrogen phosphorus flame retardants on thermal degradation of wood.
371
Constr. Build. Mater. 24 (12) (2010) 2633–2637.
372
[8] Y. Xie, N. Liu, Q. Wang, Z. Xiao, F. Wang, Y. Zhang, H. Militz. Combustion behavior of oak wood (Quercus
373
mongolica L.) modified by 1,3-dimethylol-4,5-dihydroxyethyleneurea (DMDHEU). Holzforschung 68 (8) (2014)
374
881–887.
375
[9] L. Liu, M. Qian, P. Song, G. Huang, Y. Yu, S. Fu. Fabrication of green lignin-based flame retardants for
376
enhancing the thermal and fire retardancy properties of Polypropylene/Wood composites. ACS
377
Sustain. Chem. Eng. 4 (4) (2016) 2422–2431.
378
[10] E.N. Kalali, L. Zhang, M.E. Shabestari, J. Croyal, D. Wang. Flame-retardant wood polymer composites
379
(WPCs) as potential fire safe bio-based materials for building products: Preparation, flammability and
380
mechanical properties. Fire Saf. J. (2017) 10.1016/j.firesaf.2017.11.001
381
[11] F. Carosio, F. Cuttica, L. Medina, L.A. Berglund. Clay nanopaper as multifunctional brick and mortar fire
382
protection coating-wood case study. Mater. Des. 93 (2016) 357–363.
383
[12] W. Chang, Y. Pan, C. Chuang, J. Guo, S. Chen, C. Wang, K. Hsieh. Fabrication and characterization of
384
waterborne polyurethane (WPU) with aluminum trihydroxide (ATH) and mica as flame retardants. J. Polym.
385
Res. 22 (12) (2015) 1–9.
386
[13] Q.F. Sun, Y. Lu, Y.Z. Xia, D.J. Yang, J. Li, Y.X. Liu. Flame retardancy of wood treated by TiO2/ZnO coating.
387
Surf. Eng. 28 (8) (2012) 555–559.
388
[14] S.P. Kumar, S. Takamori, H. Araki, S. Kuroda. Flame retardancy of clay-sodium silicate composite coatings
389
on wood for construction purposes. RSC Adv. 5 (43) (2015) 34109–34116. [15] S. Trey, S. Jafarzadeh, M.
390
Johansson. In situ polymerization of polyaniline in wood veneers. ACS Appl. Mater. Interfaces 4 (3) (2012)
391
1760–1769.
392
[16] J. Pan, J. Mu, Z. Wu, X. Zhang. Effect of nitrogen-phosphorus fire retardant blended with Mg(OH)2/Al(OH)3
393
and nano-SiO2 on fire retardant behavior and hygroscopicity of poplar. Fire Mater. 38 (5) (2014) 817–826.
394
20
[17] V. Merk, M. Chanana, S. Gaan, I. Burgert. Mineralization of wood by calcium carbonate insertion for
395
improved flame retardancy. Holzforschung 70 (9) (2016) 867–876.
396
[18] V. Merk, M. Chanana, T. Keplinger, S. Gaan, I. Burgert. Hybrid wood materials with improved fire retardance
397
by bio-inspired mineralisation on the nano- and submicron level. Green Chem. 17 (3) (2015) 1423–1428.
398
[19] Z. Xiao, J. Xu, C. Mai, H. Militz, Q. Wang, Y. Xie. Combustion behavior of Scots pine (Pinus sylvestris L.)
399
sapwood treated with a dispersion of aluminum oxychloride-modified silica. Holzforschung 70 (12) (2016)
400
1165–1173.
401
[20] G. Chen, Y. Hu, F. Peng, J. Bian, M. Li, C. Yao, R. Sun. Fabrication of strong nanocomposite films with
402
renewable forestry waste/montmorillonite/reduction of graphene oxide for fire retardant. Chem. Eng. J. 337
403
(2018) 436–445.
404
[21] R.-N. Amiri, T. Tirri, C.-E. Wilen. Flame retardant polyurethane nanocomposite: Study of clay dispersion and
405
its synergistic effect with dolomite. J. Appl. Polym. Sci. 129 (4) (2013) 1678–1685.
406
[22] T. Tirri, M. Aubert, W. Pawelec, A. Holappa, C.-E. Wilén. Structure-property studies on a new family of
407
halogen free flame retardants based on sulfenamide and related structures. Polymers 8 (10) (2016) 360–370.
408
[23] F. Carosio, J. Kochumalayil, F. Cuttica, G. Camino, L. Berglund. Oriented clay nanopaper from biobased
409
components - mechanisms for superior fire protection properties. ACS Appl. Mater. Interfaces 7 (10) (2015)
410
5847–5856.
411
[24] C. Ma, S. Qiu, B. Yu, J. Wang, C. Wang, W. Zeng, Y. Hu. Economical and environment-friendly synthesis of
412
a novel hyperbranched poly(aminomethylphosphine oxide-amine) as co-curing agent for simultaneous
413
improvement of fire safety, glass transition temperature and toughness of epoxy resins. Chem. Eng. J. 322
414
(2017) 618–631.
415
[25] Y. Shi, T. Fu, Y. Xu, D. Li, Wang X, Wang Y, Novel phosphorus-containing halogen-free ionic liquid toward fire
416
safety epoxy resin with well-balanced comprehensive performance. Chem. Eng. J. 354 (2018) 208–219.
417
[26] X. Shi, Y. Xu, J. Long, Q. Zhao, X. Ding, L. Chen, Y. Wang. Layer-by-layer assembled flame-retardant
418
architecture toward high-performance carbon fiber composite. Chem. Eng. J. 353 (2018) 550–558.
419
[27] Y. Kim, H.S. Kim, S.M. Jo, S.Y. Kim, B.J. Yang, J. Cho, S. Lee, J.E. Cha. Thermally insulating, fire-retardant,
420
smokeless and flexible polyvinylidene fluoride nanofibers filled with silica aerogels. Chem. Eng. J. 351 (2018)
421
473–481.
422
21
[28] X. Qiu, Z. Li, X. Li, Z. Zhang. Flame retardant coatings prepared using layer by layer assembly: A review.
423
Chem. Eng. J. 334 (2018) 108–122.
424
[29] J. Alongi, R.A. Carletto, F. Bosco, F. Carosio, A. Di Blasio, F. Cuttica, V. Antonucci, M. Giordano, G. Malucelli.
425
Caseins and hydrophobins as novel green flame retardants for cotton fabrics. Polym. Degrad. Stab. 99 (1)
426
(2014) 111–117.
427
[30] F. Bosco, R.A. Carletto, J. Alongi, L. Marmo, A. Di Blasio, G. Malucelli. Thermal stability and flame resistance
428
of cotton fabrics treated with whey proteins. Carbohyd. Polym. 94 (1) (2013) 372–377.
429
[31] A. Khandual. Green flame retardants for textiles. In: S.S. Muthu, M.A. Gardetti (Eds.), Green fashion.
430
Springer, Singapore, 2016, pp. 171–227.
431
[32] J. Alongi, R.A. Carletto, A. Di Blasio, F. Carosio, F. Bosco, G. Malucelli. DNA: A novel, green, natural flame
432
retardant and suppressant for cotton. J. Mat. Chem. A 1 (15) (2013) 4779–4785.
433
[33] Carosio, F., Di Blasio, A., Alongi, J., & Malucelli, G., (2013). Green DNA-based flame retardant coatings
434
assembled through layer by layer. Polymer (United Kingdom), 54(19), 5148-5153.
435
[34] X. Zhang, J. Mu, D. Chu, Y. Zhao. Synthesis of fire retardants based on N and P and poly(sodium silicate-
436
aluminum dihydrogen phosphate) (PSADP) and testing the flame-retardant properties of PSADP impregnated
437
poplar wood. Holzforschung 70 (4) (2016) 341–350.
438
[35] Z. Wu, N. Hu, Y. Wu, S. Wu, Z. Qin. The Effect of Ultrafine Magnesium hydroxide on the tensile properties
439
and flame retardancy of wood plastic composites. J. Nanomater. 2014, ID 945308, 8 p.
440
[36] K. Nakamoto. Infrared and Raman spectra of inorganic and coordination compounds: Part B: Applications in
441
coordination, organometallic, and bioinorganic chemistry. 6th ed., John Wiley & Sons, Inc., Hoboken, New
442
Jersey, USA, 2008.
443
[37] J. Kong, S. Yu. Fourier transform infrared spectroscopic analysis of protein secondary structures. Acta
444
Biochim. Biophys. Sin. 39 (8) (2007) 549–559.
445
[38] N.B. Colthup, L.H. Daly, S.E. Wiberley. Introduction to Infrared and Raman Spectroscopy, 2nd ed. Academic
446
Press, New York, 1975.
447
[39] K. Serec, S.D. Babić, R. Podgornik, S. Tomić. Effect of magnesium ions on the structure of DNA thin films: An
448
infrared spectroscopy study. Nucleic Acids Res. 44 (17) (2016) 8456–8464.
449
[40] F. Shafizadeh. Thermal uses and properties of carbohydrates and lignins. Academic Press, 1976, 1–17.
450
[41] F. Shafizadeh, Y. Sekiguchi. Development of aromaticity in cellulosic chars. Carbon 21 (5) (1983) 511–516.
451
22
[42] M.A. Hassan, R. Kozlowski, B. Obidzinski. New fire-protective intumescent coatings for wood. J. Appl. Polym.
452
Sci. 110 (1) (2008) 83–90.
453
[43] S. Fang, Y. Hu, L. Song, J. Zhan, Q. He. Mechanical properties, fire performance and thermal stability of
454
magnesium hydroxide sulfate hydrate whiskers flame retardant silicone rubber. J. Mater. Sci. 43 (3) (2008)
455
1057–1062.
456
[44] M.A. Bahattab, J. Mosnáček, A.A. Basfar, T.M. Shukri. Cross-linked poly(ethylene vinyl acetate) (EVA)/low
457
density polyethylene (LDPE)/metal hydroxides composites for wire and cable applications. Polym. Bull. 64 (6)
458
(2010) 569–580.
459
[45] Tang S., Yang J.J. (2013) Magnesium Binding Sites in Proteins. Kretsinger R.H., Uversky V.N., Permyakov
460
E.A. (eds) Encyclopedia of Metalloproteins. Springer, New York, NY.
461
[46] Oh, H. E., & Deeth, H. C. (2017). Magnesium in milk. International Dairy Journal, 71, 89-97.
462
[47] Cuomo, F., Ceglie, A., & Lopez, F. (2011). Temperature dependence of calcium and magnesium induced
463
caseinate precipitation in H2O and D2O. Food Chemistry, 126(1), 8–14.
464
Declaration of interests
☒The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
