By Elizabeth Raines, Chemical Engineer
....and Further Understanding the Differences between Reaction and Adiabatic Calorimetry
This article expands on the information presented in “Reaction Calorimetry vs. Adiabatic Calorimetry: Which Method is Right for Me?” and “Scale Up Aspects of Phenol-Formaldehyde Reactions” available on the Fauske & Associates, LLC website.
Thermal hazards screening is a quick and cost-effective tool used to obtain the required data for the safe scale-up of new or altered chemical processes. One parameter often used to understand the specifications of equipment required for large-scale manufacturing is the heat of reaction. While general studies found in literature can be useful for screening thermal hazards, oftentimes the characteristics of a particular reaction and reaction conditions are not available, and must be determined experimentally. Fauske & Associates, LLC (FAI) employs many tools to collect this necessary data including both reaction and adiabatic calorimeters. It was previously discussed in “Reaction Calorimetry
As discussed in “Scale Up Aspects of Phenol-Formaldehyde Reactions,” producing phenolic resins is a very old but still very active process in the industry. The chemistry has a wide variety of uses such as ablation, abrasives, coatings, composites, wood bonding, and much more. This example explores a generic semi-batch phenol formaldehyde process where a controlled addition of 37% aqueous formaldehyde was added to a mixture of phenol and 50% caustic (catalyst). The same recipe was tested in both the FAI VSP2TM and Mettler Toledo RC1. The VSP2 test was performed with a 50°C starting temperature while the RC1 was performed isothermally at 70°C.
Figures 1 and 2 show the temperature and temperature rise rate profiles from the adiabatic calorimeter experiment.
Figure 1: Temperature vs. Time for the VSP2 Phenol Formaldehyde Reaction
Figure 2: Temperature Rise Rate vs. Temperature for the VSP2 Phenol Formaldehyde Reaction
For this experiment, the phenol and 50% caustic (catalyst) were preloaded to the test cell. The mixture was heated to 50°C and held isothermally to equilibrate. A 2 hour controlled addition of 37 wt.% formaldehyde was performed beginning around 40 minutes. The formaldehyde was heat traced so that the formaldehyde was added at 50°C. For this base catalyzed recipe, the formaldehyde to phenol ratio was greater than one (2.2) as is typical for a resole resin. Typically, these reactions are intentionally not reacted to completion. The resole product carries pendent methylene hydroxyl moieties that allow the resin to be self-crosslinking at higher temperature. The phenol-formaldehyde reactive system is very complex and difficult to understand solely from thermal data, however, the results from the test do show a constant temperature rise rate from 50°C to around 90°C. The first thermal profile (“Feed Limited Addition Reaction”) appears to be feed-rate limited and is likely primarily the addition reaction between formaldehyde’s methylol groups and phenol’s aromatic ring. The second thermal profile could be a combination of the addition reaction and condensation reactions between aromatic rings to form a higher molecular weight species with additional crosslinking by condensation within the higher molecular weight polymer to form a three dimensional network. After the maximum temperature of 186°C at 99 minutes is achieved, the temperature is observed to decrease during the remainder of the addition. This is due to a combination of a significant decrease in the reaction heat generated along with adding the formaldehyde at a cooler temperature to the reaction mixture. At 157 minutes the addition was complete and the mixture was held adiabatically without any additional reactivity observed. At 200 minutes the heaters were disabled and cooldown data were collected.
This data can be used to determine many important parameters such as the required relief area (vent sizing), characteristic time of adiabatic runaway, the moles of noncondensable gas generated, kinetic parameters, temperature of no return, self-accelerating decomposition temperature, heat of reaction, and others.
In order to calculate the heat of reaction, it is assumed that the heat of reaction and heat capacity are constant throughout the reaction, and that there is zero conversion of the limiting reagent at the onset temperature of reaction. Together with these assumptions, the following equation can be utilized to determine the heat of reaction.
ΔΗr = the heat of reaction (J/g)
Φ = the phi-factor or thermal inertia for the experiment (-)
Cs = sample heat capacity (3.2 J/g-K)
ATR = the adiabatic temperature rise (K)
m = the mass fraction of limiting reagent (0,3 based on the mass of phenol)
The phi-factor is specific to the experimental design. To calculate the phi-factor, the following equation can be utilized.
Cs = sample heat capacity (3.2 J/g-K)
Cb= sample holder heat capacity (0.5 J/g-K)
ms = sample mass (74 g)
mb= sample holder mass (50 g)
For this experiment, the phi-factor is calculated to be 1.11
When external heating is not applied during the reaction, the adiabatic temperature rise, ATR, is the difference between the maximum temperature and the onset temperature. In this experiment, the reaction begins at 50°C when the addition is initiated. The maximum temperature observed in the experiment was 186°C, but the addition continued after it was realized. A cooling effect was observed bringing the temperature to 165°C at the end of the addition. Because this was the final temperature at the end of the reaction at known conditions (mass and heat capacity), 165°C is to be used as the maximum temperature. The ATR or difference between the maximum temperature and the onset temperature (165- 50°C) is 115°C. The phi-corrected ATR, (“ФATR)” , is found to be 128°C.
The results of this calculation indicate that for the measured ATR (128°C when phi-corrected) the heat of reaction for the phenol-formaldehyde reaction (both the initial and secondary reaction) is calculated to be 1,362 kJ/kg phenol.
The same reaction was studied in the RC1 (not shown in this article) and the measured heat of reaction was determined to be 982 kJ/kg phenol. The calculated ATR for the reaction was 95°C. Because the RC1 experiment was conducted at 70°C as opposed to the 50°C starting temperature in the VSP2, the sensible heat correctioin for the formaldehyde was adjusted in the RC1 to be representative of the VSP2 experimental conditions.
Why the difference? The RC1 was capable of controlling the reaction allowing only the desired reaction to occur. The VSP2 showed the total thermal potential when the reaction cannot be controlled (a loss of cooling scenario) resulting in more heat and a higher temperature rise. It is invaluable to have both pieces of this information to best prepare for how the desired process needs to be run and what could happen when an upset scenario occurs.
For more information or to discuss, please contact Elizabeth Raines at firstname.lastname@example.org, 630-323-8750.