glycol glucoside (EGG) has been achieved at atmospheric pressure at 130-1 0°C in N-methyl pyrollidone (NMP).

The new process has the following advantages over that described in UK-A-2106506. a) A single-stage, more easily controlled process is employed. b) The overall time for production of the final, purified/dry polyol is approximately 1/20 that required in the previous method. c) An autoclave is not required. The process can be carried out in a conventional reactor at atmospheric pressure. d) There is more control over the molecular weights attainable; the amount of alkylene oxide (e.g. propylene oxide (PO)) added controls the final molecular weight obtained. e) The molecular weight and the unsaturation levels of the PE polyols produced are similar to those obtained previously. f) The process is more reproducible. g) The catalyst (e.g. KOH) is in solution and present at for example 4 mol% cf 20 mol% used previously, thereby lessening corrosion problems in the case where the catalyst is KOH. h) The process may be easily adapted for continuous production.

FORMATION OF POLYOLS

The present invention relates to the formation of polyols by the oxyalkylation of polyhydric alcohols.

According to the present invention there is provided a method of producing a polyol comprising reacting a polyhydric alcohol selected from glycerol, pentaerythritol, sucrose, sorbitol, a glucoside, and 1, 2, 6 - hexane triol with an alkylene oxide in nitrogen-containing heterocyclic solvent in the presence of a basic catalyst. The alkylene oxide may be, for example, ethylene oxide or propylene oxide.

The catalyst is preferably used in an amount of 1-10 mole^more preferably 4-8 mole , based on the polyol.Examples of suitable catalysts are KOH and amines. Preferably the solvent is N-methylpyrollidone.

Preferably the reaction is effected at a temperature of 130-1 0°C.

•A principal objective of the oxyalkylation reaction is to liquify intractable solid polyhydric alcohols to make them suitable for example for use in polyurethane (PU) forming reactions with diisocyanates; their liquid nature making them suitable for PU production including their use in Reaction Injection Moulding (RIM) and Reinforced Reaction Injection Moulding (RRIM).

As an illustration of the invention, solution oxypropylation of peutaerythritol (PE) and ethylene

4. of PE and KOH.

The solution was heated under an atmosphere of nitrogen stirred and maintained at 130-140°C. A predetermined amount of PO (850-1750cm ^-5 ) was then 5 added dropwise from a pressure-equalising funnel at an approximate rate of 11 in 3hr. (As the following results show the molar mass of the oxypropylated Pintaerythritol (OPPE) product was closely related to the amount of PO added.) 0 After addition of the PO the reaction flask and contents were allowed to cool to room temperature. They were weighed to determine the amount of PO added, some of which was present as dissolved, unreacted material. § The catalyst was then neutralised (to pH7) using about 70g of a cationic exchange resin (Zeocarb 226) or by addition of 0.1M HC1. The contents of the flask were filtered and subjected to rotary-film evaporation (3 hr/ 80°C/ 1mm Hg) to remove volatiles. 0 MP was then removed by distillation at 145°C/ 1mm Hg. The OPPE product was a dark-red coloured liquid. It was mixed with its own volume of water and shaken. The pH of the mixture was again checked and if necessary adjusted to a value of 7 by the addition of 5 0.1 M HC1. The water-OPPE mixture was poured into a separating funnel and sufficient P0 (about an equal volume) added to produce the formation of an aqueous layer at the bottom of the funnel. The aqueous layer

3. i) The solvent and unreacted alkylene oxide may be recycled.

The invention will be further described by way of examply only with reference to the following Examples. Example 1

Oxypropylation of Pentaerythritol Materials

Propylene Oxide (PO) was subjected to reflux distillation over potassium hydroxide for three hours and then fractionally distilled at atmospheric pressure. The fraction boiling at 34°C was collected. It was stored in a desiccator in a freezer.

N-Methyl Pyrrolidone (NMP) was subjected to fractional distillation at atmospheric pressure and the fraction boiling at 202°C was collected. It was stored in a desiccator.

Peutaerythritol (PE) was dissolved in hot water at 85-95°C and filtered. It was then recrystallised from the aqueous solution by cooling to just above 0°C and dried in a vacuum oven at 100°C/0.05mmHg. The procedure was repeated until a melting point of 260 C was obtained.

Potassium Hydroxide (KOH) pellets were crushed under PO and dried in a vacuum oven. Procedure

Known amounts of PE (50-80g) and KOH (4-5g) were added into a tared flanged reaction flask and dissolved in NMP (400-600cm^). The amount of NMP used was approximately the minimum required for the dissolution

was removed and the remaining PO/OPPE layer washed three times with equal volumes of water.

The resulting solution was a pale yellow colour or colourless. If coloured, it was treated with activated charcoal and filtered several times, finally using a Millipore filter (FALP 0025 - 1nm). The solution was then subjected to rotary - film evaporation at 80°C/ mm Hg. The rotary - film evaporation was repeated several times and the product was finally heated to 40°C/lmm Hg under nitrogen for 8 hrs (If traces of water and other volatiles can be tolerated in the product, the repeated rotary- film evaporation and final heating under nitrogen can be omitted.) _• The OPPR product obtained was a colourless or very pale yellow liquid. The weight of PO reacted was found by weighing the final product.

Table 1 specifies the reaction mixtures investigated using the foregoing procedure. The amounts of PO added (columns 6 and 7) i.e. was determined by weighing the reaction flask before and after the reaction. The amount of P0 reacted (column 8) was determined from the amount of OPPE product finally obtained. In table 2 the amounts of reactants are given in units of moles (columns 2, 3 and 4). The ratio of P0 reacted to base moles of PE (column 5) gives the average number of moles of P0 reacted with each mole of PE added. The values of Mn in column 6 were obtained

as (column 5 x 58) + 136. They may be compared with the values of Mn in column 7, which were determined independently by the chemical determination of OH groups in the product. Differences between the values in columns 6 and 7 may be attributed to innaccuracies in the two determinations and loss of material during purification. The mol of unsaturated chain ends (column 8) was determined by Wij's titration.

TABLE 1

Example T/°C PE/g KOH/g NMP/cm^ PO άdded/cm ³ PO added/g PO reacted/ car % PO reac PO adde

1 130 68 4.48 500 860 714 478 67

2 140 80 5.20 500 1190 988 656 66

3 130 87 5.80 600 1750 1452 935 65

4 140 68 4.48 500 1440 1195 806 67

5 130 54 4.50 400 1740 1444 940 65

TABLE 2

Example PE/mol mol KoH/OH ! POreacted/mol

1 0o 50 4 8.2 16.4 1087 1025 O.30

2 0. 59 4 11 .3 19.2 1250 1248 O.63

3 0.64 4 16„ 1 25.2 1598 1598 0.71

4 0. 50 4 13.9 27.8 1748 1755 0.61

5 0.40 3.8 16 ₀ 2 40.5 2485 2090 1 .03

Remarks

The procedure used and the results in table 2 may be compared with the procedure of UK-A-2706506 and the results contained therein for the further chain extension of OPPE.

1. The catalyst level was 4 mol % KOH/OH compared with 20 mol % in the patent. The lower catald t level reduces problems of corrosion which may be encountered in metal or glass reaόtors. 2. An approximately constant % of the added - PO undergoes reaction the value (65-67%) used and on the length of time the reaction mixture is maintained at the reaction temperature after addition of all the PO. In the reported examples, the mixture was immediately allowed to cool to room temperature. The % reacted can be increased by maintaining the reaction temperature for a long time.

3. Following from 2, Mn increases approximately in proportion to the moles of PO added. Thus, control. of Mn is easily achieved. Due to side reactions, there will no doubt be some limit to the maximum value of Mn which may be realised. However, it should be possible to match the range of molar masses achieved in the patent (up to 5600) by the addition of further PO, extension of the reaction time, and/or variation of the reaction temperature.

4. The fractions of unsaturated chain end;-,

( ≥ 1 mol %) are similar to the values found using other

methods of preparation. Previous studies indicate that the values can be decreased by using lower reaction temperatures. Correspondingly longer reaction times will then be required. 5. The preparation is a one-stage, solution polymlerisation. The total time to obtaining crude product is first the reaction time (say 4 hr). In the patent, a two stage preparation was necessary to achieve products of similar molar masses (say 4-5 days including the purification of the first-stage product).

Solvent and PO can be recycled. In the patent method, PO could be recycled.

6. The present method is adaptable to continuous operation as the PO is added continuously throughout a reaction. In a continuous low process, continuous addition of PO/PE/KOH/NMP may be used together with continuous withdrawal of product after a certain residence time, depending on the molar mass required. 7. The method can be used with other polyhydric alcohols which are soluble in NMP, such as sucrose, glucose, serbitol and glycerol. Alternative catalyst systems may have been employed. Example 2 Oxypropylation of Ethylene

Glycol Glucoside (EGG) (I)

EGG is a mixture of compounds, and may be respresented by the idealised structure (I)

ro

( I )

It is composition varies but generally it consists of a mixture of predominantly < and glucosides, some bis glucoside, oligomeric material and residual ethylene glucol and glucose. The material used in the preparation was an amber glass which yielded a syrup. at -~>s 100°C. It had an equivalent weight (by acetylation) of 47.3 which compares favourably with that calculated for the monoglucoside of 44.8.

EGG (200g), KOH (4g) were dissolved in distilled NMP (200ml) contained in a 1 litre flask equipped with a mechanical stirrer, thermometer, pressure equalising dropping funnel and a condenser. A length of rubber-tubing connected the condenser to a mineral-oil bubbler which indicated the rate of loss of unreacted PO. The contents of the flask were raised to 140°C and PO added dropwise at a rate of _' 150ml per hour. Initial addition of PO caused the temperature to rise to *-—-*-155 ⁰ C and the reaction was sufficiently exothermic to maintain the reaction temperature at 130-140° with only minimal external heating. At temperatures below 130°c the PO

failed to react. After PO addition (425ml) was complete, oxygeu-free nitrogen was passed through the reaction solution to drive off unreacted PO. The flask was then cooled and weighed to determine PO addition (323g 90% conversion of PO).

Methyl ethyl Ketone (MEK) (400 ml) and Zerolit DM-F resin (70g) were then added and the slurry mechanically stirred until pH 7 was attained. The suspension was filtered and MEK removed from the filtrate using a rotary film evaporator (water bath 30°C) under water-pump vacuum. The solution was then transferred into a flask for vacuum distillation to remove NMP. (0.01 mm Hg; maximum flask temperature 150°C). The product (513g) w s a clear, amber, viscous liquid at room temperature, mobile above 40°C. It had the following characteristics: a water content of 0.11% w/w (Karl Fischer) a residual NMP concentration of 0.1% w/w (GLC) An equivalent weight of 118.8 1.1 was determined by acetylation compared with a theoretical value of 123.7 calculated on the basis of PO absorbed. An effective equivalent weight of 254 towards MDI was obtained suggesting a functionality close to 2 for the polyol. (MDI reacting at the more accesβible sites on the ends of the propylene oxide chains). This suggests that chain extension occurs mainly at the primary hydroxyl groups in EGG leading to the

idealised structure (II) for the oxypropylated

( I D

GPC analysis using polypropylene oxide calibration standards yielded Mn = 514 Mw = 560 and Mn/Mw = 1.09. Note

1. The object of the preparation was: a) to prepare a liquid polyol from solid EGG. (EGG although producing a syrup at •-*»—100°C is not mixible with MDI and its viscosityis too high to process easily by RIM) b) to maximise the carbolydrate content of the resultant polyol by the minimum addition of PO.

The method may however be used to prepare a range of EGG/PO adducts by varying the amount of PO added as in the case of PE.

2. Sucrose has also been studied but is not readily soluble except at high temperatures. KOH is not an effective catalyst in NMP solution for this substrate. Trialtrylamine aid other less basic catalysts are more acceptable but some degradation/charring still occurs.

Polyurethane Formation

OPEGG has been employed alone and or in admixture with the commercially available polyether, polytetra- methylene oxide (PTHF) to produce polyurethanes by reaction with 4,4' diphenyl methane diisocyanate (MDI). Viscosity studies on these polyol blends in the temperature range 20-60°C indicate that their

5 2 viscosities lie between 10 -10 ceutipoise making them suitable for normal fabrication processes (for surface coatings, adhesives, cast polymers and RIM processed materials).

The polyurethanes produced ranged from a clear amber glass (from OPEGG/MDI) to clear yellow elastomers.

Properties have been studied by DSC, dynamic mechanical analysis, tensile and impact testing; solubility and water-uptake studies have also been made.

Some typical data are shown below:

Properties of Polyurethanes from OPEGG/PTHF blends

Polymer Polymer Composition Characteristics _Tg (a)o _c % OPEGG % PTHF^

100 0 Clear amber glass 57

75 25 Clear orange tough 23 leather 50 50 Clear yellow stiff -15 rubber 25 75 Clear pale yellow -37 rubber

0 100 Clear colourless -44 rubber

(a) DSC

(b) a transition at 30°C also observed

(c) PTHF of nominal molecular weight 750.