Document Yr1L71yZw7k1NJMav8mrZryNK

Determination of the Process-Volatility of Plasticizers A. C. POPPE Essochem Europe, Inc. MacHelen (BruseeU), Belgium INTRODUCTION Testing vinyl for real-life performance involves not only the end-product properties of the for mulation but its processing characteristics as well. The latter is important since it has a considerable influence on the overall economics of the product. Volatile loss of some of the vinyl formulation components is an important consideration, particu larly in processes such as spread coating. During fusion, the product is exposed to large volumes of fast moving hot air (180-200*C or 360-400*F) re quired to heat up the vinyl layer. These conditions promote considerable evaporation of the volatile components, especially plasticizer. Testing plasticizers and plastisols in conven tional laboratory ovens may lead to a correct rank ing as to process volatility but it usually does not give adequate information regarding the mag nitude ofthe evaporation losses in industrial opera tions. In order to simulate plastisol fusion practice, we believe the laboratory method should use condi tions similar to those applied in the actual indus trial oven. Thus, high air velocity (forced convec tion) and high air temperature should be applied. In the following paragraphs the required condi tions will be more precisely defined and a labora tory procedure to determine the "processvolatility" ofplasticizers will be described. Finally, some results obtained with this procedure will be discussed. PLASTISOL FUSION CONDITIONS IN INDUSTRIAL HOT-AIR OVENS The following data indicate the range ofrelevant conditions that are found in industrial practice. The ovens are relatively long, exposing a large surface of vinyl to the circulating hot air. The air tempera ture typically is 180-200*0 (360-400*F). The resi dence time of the vinyl in the oven ranges from about 20 s to a few minutes, depending on several factors, including coating thickness (typically 0.1- l mm, 4-40 mils). The relatively short residence time indicates that the heat transfer rate is high, since in less than this short period the vinyl layer must be heated up from room temperature to a level of 170*C (340*F) or higher. This is achieved via a high air velocity, which also promotes evaporation. The air is purged at a rate of 10,000-50,000 m*/h (350,000-1,500,000 SCF/h), thus removing vol atilized material. The purge rate is dictated by the requirement to prevent drip at "cold spots'* in the oven. On the average, however, the circulating air is far from being saturated. The plasticizer concen tration in the purged air is typically of the order of 1 g/m* (0.03 g'SCF), with considerable variation from case to case, depending on the conditions. It thus follows that plasticizer losses ranging from 10-50 kg/h (20-100 lb/h) may occur. Table 1 is de rived from a paper by H. Delorme (1), supporting the above data. An analysis ofthe heat-up ofthe vinyl layer on its course through the oven has been made using, clas sical theory as described in the literature (2). In the heat-up process two major steps can be defined. Firstly, the heat transfer rate from the bulk ofthe air to the vinyl surface plays a role of importance. Sec ondly, the transfer of heat from the vinyl surface through the inside ofthe layer may be determining. It can be shown that with a relatively thin vinyl coating and air-heating, the air-to-vinyl heat trans fer rate is the main rate-determining step. Thus, the vinyl layer has an approximately uniform tempera ture across its thickness. This greatly simplifies the mathematical approach. With sufficient pr cision for the present purposes, the heat-up of the vinyl layer in the oven can thus be described as: AT/AT. - -*** (1) The meaning of the symbols, together with some typical values for the parameters are shown in Table 2. (Equation 1 should be used applying con sistent units, as shown below. For example the Table 1. Plasticizer Losses In Fusion Ovens Application Air purge mVh (SCF/h) Plasticizsr cone. In purge g/m* (g/SCF) Calculated loss kg/h (lb/h) Coated fabric Cushioned flooring Bottle caps 12.000 60.000 30.000 (420.000) (2.100,000) (1,050,000) 3.0 0.78 0.28 (0.085) (0.022) (0.008) 38 47 8.4 (80) (100) (18) JOURNAL OF VINYL TECHNOLOGY, SEPTEMBER 1981, VOL. 3, NO. 3 UCC 037441 173 A. C. Poppe Table Z Symbol Moaning Typical value AT Temp, difference hot Variable air--vinyl at time f. AT. initial temp, difference 160*C (290*F) f Time (seconds) Variable elapsed after entrance into the oven. k Heat transfer coeffi 20-80 J/(m*) (s) (*C) or 4-15 cient air-to-vinyl (de BTU/(ft>) (h) (F) pends on degree of air convection) 3 Density 1200 kg/m* or 75 Ib/tt* C Specific heat, vinyl 1800 J/(kg) CO or 0.42 BTU/ (lb) (*F) a Thickness (or half 0.1 - 1 x 10-* m (4-40 mils) thickness for two- sided heating) of the vinyl layer. coating thickness A given in mils, should be recal culated in feet.) The heat transfer coefficient k is mainly deter mined by the degree of air convection. The typical value range as shown above reflects the variation in forced air convection that occur in practice, from oven to oven. Equation 1 closely describes the heat-up of the vinyl layer in the course of time. Substituting the indicated typical values for the parameters (taking k = 40 and A = 0.3 mm) and calculating the point in time where the vinyl still is 8C (14.4F) less than the hot-air temperature (thus, AT/AT, " 0.05) we find: t = 50 s. This value is reasonable in the sense that it falls within the range of industrial practice. The above excursion into theory is useful since it yields quantitative criteria to define a laboratory procedure approaching industrial practice. Any laboratory procedure intended to investigate phe nomena associated with industrial fusion operations--such as the process evaporation losses--should use similar forced air convection conditions as those described above. In hot-air ovens the heat transfer rate as well as the evapora tion rate of the plasticizer in the plastisol may be expected to be directly related to the degree of air convection. PROCE S S-VOLATILITY: LABORATORY EQUIPMENT AND PROCEDURE Description The equipment used in this study is essentially a knife-over-roll spread-coating frame and a preci sion air-heated oven. The geometry of the oven is similar to that of industrial scale ovens. With the spread-coating unit a coating is applied on a 30 x 30cm(l x 1ft) substrate (release paper, fabric, etc.) held in the frame. The frame with coated substrate slides into the oven under pushbutton control and returns automatically after a preset exposure time that can be varied at will. Air temperatures can be controlled over-a wide range. The air circulation rate in the oven is high and can be varied between 500 and 1400 mVh (17,500-50,000 SCF/h). The sample is heated by the hot-air stream from the top and bottom of the oven. The assembly used is commercially available (Werner-Mathis AG, Zurich/Switzerland; type LTSV/LTF). Any other equipment simulating real-life coating processes may De used, of course. The experimental procedure is simple. A known amount of plastisol is coated on the substrate. The coated weight can be found by weighing the plas ticizer container and the tools used before and after coating. Alternatively, the plastisol can be coated on paper and lightly pregelled. After removing the pregelled sheet from the release paper, a sample can be cut and weighed. The sample can then be placed back on the release paper to be exposed to the desired conditions. After cooling, the exposed sample is removed from the paper and weighed again. The observed weight difference is equal to the amount ofplasticizer--and other volatile mate rial, if present--that evaporated during fusion. Discussion The procedure has been successfully applied to investigate the difference in process volatility be tween various plasticizers, including DOP, DINF and DIDP. The loss rates as found with the proce dure appear to be comparable to those experienced in industrial practice. The results have been re ported (3) and a summary of the results will be discussed further below. In addition it will now be shown that the equipment also closely simulates typical industrial heat-up conditions. Figure 1 shows the heat-up curves obtained in the oven used. A thermocouple was placed in the vinyl layer to obtain the results as shown. The coating thick ness in these experiments is relatively high in order to facilitate temperature measurements. The 1.2 mm (47 mils) coating heated-up in about 2 min. The heat-up times are seen to be approximately proportional to the thickness of the coatings with reasonable precision. This is in agreement with Eq 1. From the curves a value for the heat transfer coefficient k can be calculated, using Eq I. Values ofaround 40 J/(m*) (s) (*C) are found at maximum air Fig. 1. Heat-up curves for different coating thicknesses. DINPplastisol; oven temperature 160*0 (320'F); 1.2 mm - 47 mils, 2,15 mm - S3 mils, 4.0 mm -- 157 mils. JOURNAL OF VINYL TECHNOLOGY, SEPTEMBER 1481, VOL. 3, NO. 3 ucc !1774<a Determination of the Proem*Volatility of Plasticuers circulation rates. This range is also representative of industrial ovens. Overall, it appears that the test-assembly approaches real-life conditions well. Further experimental results will be described in the following sections. SOME PRACTICAL RESULTS The results shown below all refer to basic plastisol formulations: PVC100, plasticizer 50(DOP) or 53 (DINF) or 55 (DIDP), organotin stabilizer 0.5. Exposure time is 2 min, unless otherwise men tioned. The coating thickness is 0.25 mm (10 mils), unless otherwise mentioned. Since the evaporation process is largely surface-determined, the results have been expressed in terms of weight loss per unit of time and area. itumiu c Fig. 2. Loss rates at different temperatures (05 mm coating thickness). Each point is the average ofat least two test results. Influence of Plasticizer Type and Temperature on Process Volatility Figure 2 reveals the profound influence of plas ticizer type. Also the temperature dependence of the volatility is considerable: approximately 4% per C or a factor 1.5 per 10*C (2.2% per *F or a factor 1.25 per 10*F). Still, at 200C (392*F), DIDP has lower losses than DOP at 180*C (356*F). The absolute level of the evaporation losses is high. For DOP at 200*0 (392*F), for example, 8.7 g/(m*)(min) or0.8 g/(ft*)(min) has been found. Fora 1.8 x 18 m (6 x 60 ft) oven, having 32 m* (360 ft*) exposed vinyl surface, this means 360 x 0.8 " 290 g/min of DOP evaporated (max.), equivalent to 17 kg/h or 38 lb/h. Since the product is at lower tem perature during the heat-up period, the calculated loss represents a high figure. Nevertheless, the cal culated loss is in good agreement with the loss data shown in Table 1. The relative process volatility DOP/DINP/DIDP is approximately 1/0.55/0.35. It is interesting to note that this ratio also holds for the vapour pressure of these plasticizers in the temperature range 350400F. This observation strongly suggests a direct relationship between process volatility and vapour pressure. Influence of Coating Thickness and Exposure Time The influence of coating thickness is mainly through longer required heat-up time. The appar ent absolute weight loss over the same exposure period is slightly lower for thick layers than for thinner ones, due to the lower average tempera tures at the beginning of the exposure period. Fig ure 3 shows plots of the1 observed cumulative weight loss vs exposure time, for two plasticizers and for two different coating thicknesses. The weight loss for the thicker layers lags initially but parallels the curves for the thin layers after heat-up has been achieved. Thus, plasticizer evaporation is essentially surface-controlled since the thickness of the layer does not have an influence (apart from the initial lag). Another effect is noticeable from Fig. 3. Cumulative weight loss os time for different coating thickness. Fig. 3. The curve for DOP (0.25 mm, 10 mils), starts dropping off after 2-3 min. This is probably due to depletion of the plasticizer in the vinyl layer. The average loss rate for DOP is 8.7 g/(m*) (min) or 0.8 g/(ft*) (min). For a 0.25 mm (10 mils) coating-- weighing approximately 300 g/m* or 28 g/ft*--this * means 2.9 percent weight loss per min, or 8.7 per cent of the plasticizer lost per min. Thus, after 3 min, approximately 25 percent of the DOP has evaporated from the sample. This is visible in the curve for DOP (0.25 mm). The depletion rate ofthe thicker layer is less on a percentage basis, ofcourse, while the absolute loss rates (g/ft*) are closely simi lar. Influence of Air Circulation Rate The air circulation rate in the oven as used can be varied between 500 and 1400 m*/h (17,500-50,000 SCF/h). In most experiments the maximal rate has been used, obtaining conditions representative for industrial practice. Some experiments have been made at the minimum rate. The loss rates appear to be reduced to approximately half of those at maximum rates. Also, temperatures have been re corded for the two cases. Figure 4 shows the data for DINP. It is clear that a low air rate leads to slow heat-up. The loss rates are also considerably re duced. It is also clear, however, that the longer residence time required counteracts this lower loss rate. CONCLUSIONS The described technique to study process volatility--using a laboratory oven with consider able air convection, comparable to industrial JOURNAL OF VINYL TECHNOLOGY, SEPTEMBER 1981, VOL. 3, NO. 3 uec 037443 177 A. C. Poppe perature, and other factors can be investigated in a simple way. The relevance of these experiments to real-life performance is obvious. By judicious choice of the plasticizer system, money can be saved and en vironmental measures be facilitated. Fig, 4. Influence of air circulation rate (high - 1, low - 2) on heat-up and weight lose. DINP plastisol, 1 mm (40 mils). practice--yields useful data otherwise not easily available. The influence of plasticizer type, tem REFERENCES 1. H. Delorme, GFP-conference on spread coating, Lyon/ France (June 19-20. 1979). 2. M. Jakob, "Heat Transfer," Vol. 1, J. Wiley te Sons, New York (1949). 3. A. C. Poppe, Kunttstoffe, 70, 38 (1980). ucc 037444 178 JOURNAL OF VINYL TECHNOLOGY, SEPTEMBER 1981, VOL. 3, NO. 3