Abstract
Castor oil polyol (COP) having a hydroxyl number of 400 mg KOH/g was prepared through the transesterification reaction of castor oil with glycerol. The effect of reaction temperature on the composition, hydroxyl number and viscosity of the COP products was studied. A series of rigid polyurethane foams were synthesised using the mixtures comprising COP and a petroleum-based polyol with various proportions as polyol component. It was found that the foaming rate, compressive strength and dimensional stability and morphology of resulting foams were dominated by the foam formulation, in a more accurate way, COP content in the polyol mixtures. The combination of expandable graphite and dimethyl methyl phosphonate exhibited stronger flame retardant function which was ascribed to the synergistic effect associated with the simultaneous presence of the two additives. An improvement in thermal stability was observed due to the inclusion of the flame retardants.
Introduction
Rigid polyurethane foam (RPUF) constitutes a distinctive group of products in the PU family and accounts for about 23% of all PU production. 1 RPUF is used on a large scale mainly as thermal insulating material owing to its low thermal conductivity, high compression strength, low density, high strength-to-weight ratio and low moisture permeability. 2 – 3 This cellular material has been produced from polyol and isocyanate which are obtained mostly from petrochemicals. 4 With increasing concern in relation to the environment and the shortage of fossil resource, preparation of RPUF using vegetable oil derived polyol is especially advocated. 1 , 5 Castor oil (CO) is a naturally occurring triglyceride of ricinoleic acid and has been mainly considered for the synthesis of rigid foams. However, CO has some intrinsic defects including low hydroxyl number (OH number) and slow rate of curing thanks to the presence of secondary hydroxyl group. 6 One way to overcome these disadvantages is the transesterification of CO with polyhydroxy compounds such as glycerol,trimethylolpropane or pentaerythritol in the presence of basic catalyst, thus introducing reactive hydroxyl groups. The acquired castor oil polyol (COP) could fully or partially replace petroleum-based polyol for the foam preparation. 7 – 8
Like other polymers, the major weakness of RPUF lies in its high flammability which largely restricts its usable range. 9 – 10 Consequently, RPUF has to be endowed with flame retardancy to actualise the high fire-protection performance desirable in many practical applications. The most widely used approach for the aim is to add flame retardant additives into the foam and multifarious additives have been attempted. Of them, expandable graphite (EG) has attracted considerable attention because it possesses higher efficiency in fire performance and releases less toxic smoke during burning than other flame retardants.11-14 EG is a graphite intercalation compound in which sulphuric acid is inserted between the carbon layers of graphite. When exposed to a heat source, EG expands and generates a voluminous insulative layer, thus providing fire performance of interest to the polymeric matrix. 15 – 16 Besides EG, phosphorus containing compounds such as dimethyl methyl phosphonate (DMMP) are also able to markedly depress the flammability of RPUF and often exhibit fairly good fire-retarding performance. 17 Moreover, combination of two flame retardants is reported and a significant improvement in the fire performance of polymeric foams is observed. 18 – 19 So far, a large number of papers concerning flame retardancy and thermal stability of petroleum-derived RPUF have appeared in literatures, but such attention has seldom been given to the CO-derived RPUF products.
In this work, a COP having OH number of 400 mg KOH/g was synthesised through the transesterification of CO with glycerol. Gel permeation chromatography (GPC) and Fourier transform infrared spectroscopy (FTIR) were adopted to track this reaction. The resulting COP in combination with a commercially available petrochemical polyether-polyol (co-polyol) was employed to fabricate RPUF by virtue of reaction with polymeric diphenylmethane diisocyanante (PMDI). Effects of COP content on foaming rate, morphology and mechanical properties of RPUF were examined. In addition, foam flame retardant and thermal properties were investigated after inclusion of EG/DMMP mixture.
Material and methods
CO having OH number of 163 mg KOH/g and functionality of 2.7 was supplied by Wenshui Huanghe Oil & Fat Co., Ltd, China. A polyether-polyol, GR-4110 G with OH number of 400–460 mg KOH/g and functionality of 4.2 originated from the polypropylene oxide and sucrose/glycerol base, was obtained from Jiangsu Zhangshan Chemical Co. Ltd, China. Glycerol of analytical grade was purchased from Sinopharm Chemical Reagent, China. PMDI (functionality 2.7, NCO content 31.3 wt-%) was provided by Yantai Wanhua Polyurethane Co. Ltd, China. Distilled water as chemical blowing agent was generated in our laboratory. HCFC-141b, a co-blowing agent, was the commercial product of Solvay Flour GmbH, Germany. Surfactant (B8418) was obtained from Goldschmidt AG, Germany. The catalysts including dibutyltin dilaurate (T-12) and triethylene diamine dissolved in dipropylene glycol to 33 wt-% (Dabco 33LV) were supplied from Air Products and Chemicals Inc. EG was purchased from Baoding Aikesen Graphite Co. Baoding, China. DMMP was supplied by Sinopharm Chemical Reagent, China. All these chemicals were used as received.
The OH number of polyol was determined according to ASTM D1957–86. The viscosity was measured on a Brookfield DV-II viscometer. Number average molecular weight (Mn) of polyol was measured by GPC at room temperature (Shimadzu LC-20AD). Dimensional stability was measured following ASTM D 2126. The apparent density of the foams was measured according to ASTM D 1622–03 with sample size of 30 × 30 × 30 mm (width × length × thickness), and an average of at least five measurements was taken to report. A compressive stress at 10% strain in parallel to foam rise direction was measured with an Instron Universal Testing Machine (Model LLOYD instrument) according to ASTM D 1621–00. The chemical structures of the foams were characterised using FTIR(Nicolet380). Thermal gravimetric analysis (TGA) was carried out on a thermogravimetric analyser TGA-50 with a heating rate 5°C/min from ambient temperature to 800°C in air. The morphology of the foams was observed under a scanning electron microscope (SEM, Hitachi S 3500N). The limiting oxygen index (LOI) test was carried out with a JF-3 oxygen index test instrument (Jiangning analytical instrument factory, China) according to ASTM D 2863–97.
Experimental
Transesterification of CO with glycerol
The reaction was conducted in a round bottomed flask equipped with a stirrer, thermometer, reflux condenser and nitrogen inlet. The flask was first charged with CO and litharge (0.1 wt-%) under stirring and heated up to desired reaction temperature. Glycerol was then added dropwise in terms of CO/glycerol molar ratio of 1:2. Reaction was carried out at 160–240°C for 4 h. The resultant COP products were dried at 80°C under vacuum.
Foam preparation
Typical formulations of neat RPUF
The polyol mixture, surfactant, blowing agents and catalysts were first mixed in a plastic beaker. The pre-weighed PMDI was then added and stirred acutely. Subsequently, the mixture was poured into an open mould and allowed to produce free-rise foam. All products were postcured at ambient temperature for one week before measurement and characterisation. These pure RPUFs were hereafter referred to as non flame retardant foams (NFFs). Flame retardant foams were made by adding various amounts of EG (i.e.10, 20, 30 and 40 wt-%) and constant amount of DMMP (10 wt-%) into a chosen NFF sample. They were called FF10/DMMP, FF20/DMMP, FF30/DMMP and FF40/DMMP in terms of respective EG content.
Results and discussion
Transesterification of CO with glycerol
Scheme 1 shows the schematic drawing of the modification reaction
The schematic drawing of the modification reaction.
GPC was used to monitor the progress of the reaction and results were depicted in Fig. 1. It was found that at 160°C, primary OH group containing mono-, di-glycerides created accompanied by the reduction of CO and glycerol. With further rise of reaction temperature to 200°C, more mono-, di-glycerides were gained. However, these desired products started to lessen beyond the temperature of 200°C.
GPC chromatograms of CO transesterification products. Molar ratio of CO/Glycerol: 1:2; reaction time: 4 h.
Effect of reaction temperature on COP synthesis
FTIR spectroscopy was used to analyse the functional groups existing in CO and COP, respectively. As shown in Fig. 2, characteristic bands observed at 3350 and 1744 cm−1 were ascribed to hydroxyl and ester carbonyl groups, respectively, while bands corresponding to 2870 and 2970 cm-1 were assigned to C–H stretching vibration in CH2 and CH3. Compared with CO, COP showed the enhanced band of hydroxyl group at 3350 cm−1. This well agrees with the increase in OH number after the CO transesterification. In addition, a new, intensive absorption peak at 1035 cm−1 in COP sample should be related to the generation of primary hydroxyl group in mono-, di-glycerides. All of above results validated that transesterification of CO with glycerol under appropriate conditions can provide a feasible way for upgrading CO.
FTIR spectra of CO and COP. Transesterification conditions: molar ratio of CO/Glycerol: 1:2; 200°C; 4 h.
Substitution of COP for GR-4110 G
A series of NFF samples (without flame retardant) were made by replacing up to 90% GR-4110 G with COP. It was noticed that the cream, tack-free and gel times shortened with the increase of COP mass fraction (data not shown). This enhancement in reactivity may be due to relatively higher primary hydroxyl content in COP than in GR-4110 G. COP addition also affected the mechanical properties of resultant foam. As shown in Fig. 3, the compression strength as a whole eased along COP content. By contrast, the dimensional stability (volume change, △V%) exhibited a reverse trend. Moreover, it was found that upon increasing COP content until 50%, the changes in mechanical properties were slow. However, the alternations became obvious by further increasing COP content from 50 to 90%.
Effect of COP mass fraction on foam properties.
It is known that the mechanical characteristics of RPUF are directly influenced by the functionality and structures of the polyol used. RPUF is a cross-linking polymer made through the reaction of high functionality polyol with PMDI. The compression strength and dimensional stability depend on the foam cross-linking density. With the percentage reduction of GR-4110 G having higher functionality (f = 4.2) in the foam formulation (namely, more COP addition), the fall in cross-linking density would lead to the important worsening of the mechanical properties. On the other hand, more side groups in COP than of GR-4110 G weaken hydrogen bond force between hard domains and intermolecular forces, deteriorating compression strength and dimensional stability of obtained foams. The evolution in foam microstructure supported the standpoints above. SEM images of the various samples at cross-sections perpendicular to the direction of rising of the foams were presented in Fig. 4.
Scanning electron micrographs of RPUFs prepared from: a 0%; b 30%; c 50% and d 70% COP.
It was found that the cell morphology of the foams was commanded by the foam formulation. For example, neat RPUF (Fig. 4 a) was made up of spherical and polyhedral closed cells with well-defined cell wall. The average pore diameter is about 400 μm. With adding progressively larger amounts of COP until 50% (Fig. 4 b and c), the homogeneity of the cells deteriorated slightly and the average pore diameter became larger gradually. These are destined to decrease compression strength and dimensional stability to a certain degree. As COP content in polyol mixture reached 70%, the cells shrank acutely and average size of the cells reduced markedly and a large number of broken cells were observed. These observations are consistent with the data given in Fig. 3. It appears that the COP content in polyol mixture should not be higher than 50% in order to acquire NFF with better comprehensive performance.
Flame retardancy of foams
Effect of EG/DMMP content on the LOI values
In order to further ameliorate flame retardancy, combined use of EG and DMMP was attempted. As shown in Table 3, when DMMP loading remained constant (10 wt-%), increasing amount of EG induced higher LOI values. This implied that these hybrid fillers posed stronger flame retardant effect due to the synergistic effect associated with the presence of both additives. A material with LOI of 26 or higher is commonly rated as a flame retardant material. 21 Foams filled with EG/DMMP was thus considered to own certain anti-flammability property.
Thermal stability of foams
Fig. 5 shows the TGA behaviour of neat RPUF and FF20 (RPUF filled with 20 wt-%EG/10 wt-% DMMP) at a heating rate of 5°C under a flow of air.
TGA profiles of unfilled and filled RPUFs.
Similar shape of the two curves implied that both the processes followed the same staged degradation mechanism. First peak (30–210°C) was excluded from the discussion because it was the result of moisture absorption by foam. The second peak (210–420°C) can be connected with the thermal pyrolysis of the polymers, forming alcohol and isocyanate groups. The third one (420–680°C) originated from the combustion of the foams.
20
21
,
22
Compared with neat foam, FF20 showed lower weight loss rate. Besides, the later had higher charred residue at high temperature section, forming a thermal barrier to prevent further decomposition These facts validated that the incorporation of EG/DMMP endows RPUF with better thermal stability. Fig. 6 depicts the dependence of thermal stability on the flame retardant content. With the increase of EG content in EG/DMMP systems from 10 to 40%, better improvement in thermal stability of filled RPUF samples was observed, witnessed by gradually diminished weight loss rate and steadily increased charred residue at high temperature section.
TGA profiles of RPUFs with various amounts of EG/DMMP.
Conclusions
CO transesterification can elevate OH number and depress the viscosity of CO availably. The combination of COP and petroleum-derived polyether can be used as polyol component for producing RPUF. The experimental results showed that the mass fraction of COP not more than 50% gave good comprehensive performances including fast foaming rate, acceptable compression properties and better dimensional stability of foam. Incorporation of EG/DMMP into NEF50 conduced obvious improvement in flame retardancy of foam. TGA/DTGA data revealed that EG/DMMP addition greatly contributed to thermal stability of foam.
Footnotes
Acknowledgements
This work was financially supported by the Youth Talents Foundation on innovation 2020 of Institute of Coal Chemistry (grant no. 2011SQNRC04), Chinese Academy of Sciences, China.