By: Donald J. Knoechel, Ph.D., Senior Consulting Engineer, Fauske & Associates, LLC
The acid-catalyzed hydrolysis of epichlorohydrin to make monochloropropanediol (MCPD) has been chosen as the round-robin reaction for the DIERS community for the year 2019-2020. As a lead-in to that exercise, we at Fauske & Associates, LLC, carried out a series of reaction calorimetry (RC) experiments together with thermal screening by differential scanning calorimetry (DSC) of the post-reaction mixtures and analysis of the decomposition kinetics using AKTS [1] to determine a TMR24ad (temperature where the time-to-maximum rate is 24 hours under adiabatic conditions) for the runaway scenario.
We recently presented part of this work at the 2019 DIERS Users Group Meeting held in Burr Ridge, IL, September 16-18, 2019, titled "Reaction Calorimetry Data and Scale-up Considerations for the Hydrolysis of Epichlorohydrin." In part one of this series of articles, we present a portion of the reaction calorimetry and thermal screening results from this study. Part two will explore some simple calculations to show how these thermal challenges offered by the epichlorohydrin hydrolysis are handled with scale. Part three will look at how our main adiabatic calorimeters (ARC, VSP2) characterize the polymerization/decomposition given the previous thermal screening and RC results.
A simplified depiction of the chemistry is shown below.
Epichlorohydrin (EPI), catalyzed by aqueous acid, reacts with water to form monochloropropanediol (MCPD). This post-hydrolysis reaction mixture can polymerize and ultimately decompose on heating. Reaction calorimetry is used to characterize the desired process heat of reaction and, in turn, project an adiabatic potential to help identify what temperature could be reached in a loss-of-cooling scenario or all-in addition. Thermal screening of the post-hydrolysis reaction mass by DSC gives us a first look at where in temperature that secondary reaction activity may initiate. Kinetic modeling by AKTS [1] definitively identifies the extent of overlap of the projected loss-of-cooling/all-in addition scenario and the secondary polymerization/decomposition. Finally, the data from RC, DSC and the results of the AKTS modeling, all of which characterize the energies involved in a generic Epichlorohydrin hydrolysis recipe, will be used to identify a criticality class per Stoessel [2,3] for the process.
Our generic laboratory recipe has a composition of epichlorohydrin 27.1% wt., 72.4% wt water, and 0.5 % wt acid catalyst (69.0% nitric acid). Using this recipe, Figure 1 shows the heat flow profile for a semi-batch reaction (380 g scale) conducted at 80°C with an EPI addition rate of 3 g/min (35 min) addition to the starting mass of acidified water. The area under the heat flow profile curve represents the total heat of our semi-batch reaction and integrating yields -87.0 kJ or a normalized heat of reaction of –78.5 kJ/g-mole EPI. Dividing the total heat by the thermal mass (mass x heat capacity) projects a theoretical temperature rise under adiabatic conditions due to the intended heat of reaction. For this semi-batch reaction, the calculated adiabatic temperature rise is +60.9°C.
We note that at the end of the addition, 76% of the heat evolution has been realized, meaning 24% of the total energy is accumulated. If cooling was lost at the end of the addition, for example, the dynamic adiabatic potential would decrease to +14.6oC.
So here just from the intended heat of reaction we see a sizeable adiabatic potential as from 80°C a loss of cooling/all-in addition scenario could result in a temperature rise to 140.9°C (maximum temperature of the synthetic reaction [MTSR] per Stoessel [2, 3]) where the vapor pressure of water would be 39 psig.
But that's not all of the story. If we take the post-hydrolysis reaction mass from this 80°C RC run and look at it in the DSC, we see the scan shown in Figure 2.
In the DSC scan, we see the secondary reaction potential due to polymerization/decomposition at elevated temperatures. Taking the integrated energy from the scan, -508 J/g, and dividing by the heat capacity of the reaction mass tested (3.758 J/g°C), we calculate an additional +134.9°C adiabatic temperature rise potential. Adding this higher temperature-activated energy to our previous intended reaction energy predicts a total possible temperature rise of +195.8°C!
To confirm that the intended heat of reaction under loss of cooling/all-in addition conditions can raise the temperature of the reaction to a temperature where the secondary reactivity can initiate, adiabatic calorimetry is needed. Adiabatic testing using the vent sizing package (VSP2) and the accelerating rate calorimeter (ARC) of this reaction system will be the subject of the third article in this series.
Alternatively, kinetic modeling by AKTS [1] of several DSC runs at different scan rates of a post-hydrolysis mixture [4] affords a time-to-maximum rate versus temperature plot shown in Figure 3.
Here we see that the projected temperature rise from the all-in scenario from the reaction calorimetry experiment (maximum temperature of the synthetic reaction [MTSR] of 140.9oC) coincides with a time-to-maximum rate of < 4 hours. However, the same plot identifies the TMR24ad temperature as 113.7°C.
The collected and derived data can be used to identify a criticality class per Stoessel [2, 3] for this recipe of the EPI hydrolysis process. Figure 4 below shows the general criticality class diagram (note TD24 = TMR24ad). If we assume the maximum technical temperature (MTT) for this process is the open system reaction mass boiling point (taken as water, 100°C), together with the process temperature (Tp) of 80°C, the TMR24ad of 113.7°C, and MTSR of 140.9°C, the criticality class is 4 (MTSR > TMR24ad > MTT > Tp).
The RC experiment presented above had a 3 g/min-controlled addition of EPI (35 min). Even with this rather moderate addition rate, the adiabatic potential is reduced by the controlled addition as the accumulated energy (24%) dynamically decreases the adiabatic potential to +14.6°C reducing the MTSR to 80 + 14.6 = 94.6°C. Under process conditions utilizing a controlled addition of EPI, the criticality class shifts from 4 all the way to 1 (TMR24ad > MTT > MTSR > Tp), though we note that the margins between Tp (80°C), MTSR (94.6°C), MTT (100°C), and TMR24ad (113.7°C) are small. Furthermore, the process can only be considered criticality class 1 if, when cooling is lost, the EPI addition is stopped so that no more reactant can enter the now uncooled reactor.
The other option defined by Stoessel for defining MTT is for a closed system. That would be the temperature at the maximum pressure, which is the set pressure of the relief device on the reactor. This is important if the intended scale-up equipment has a pressure steam or single fluid jacket system using a heat transfer fluid capable of temperatures beyond that of, particularly for this example process, free steam (100°C). In the closed system, with no relief device, MTT could very well be defined by the maximum possible jacket temperature (utility failure overheating scenario). If this temperature were, say, 150°C, then the original classification could be considered 5 (MTT > MTSR > TMR24ad > Tp) which, with a controlled EPI addition is reduced to 2 (MTT > TMR24ad > MTSR > Tp).
In summary, the information gathered and illustrated using RC, DSC, and knowledge of the scaled-up process equipment definitively shows how important it is to control the intended heat of reaction in this epichlorohydrin hydrolysis process. Not only is there plenty of intended reaction energy to deal with, but there is also secondary polymerization/decomposition energy waiting to be initiated should the reaction runaway. While this reaction is easy to control in a laboratory reaction calorimeter via jacket cooling, just how easy is it to control at scale?
In our next newsletter we, will explore some simple calculations to show how these thermal challenges offered by the epichlorohydrin hydrolysis are handled with scale.
Reaction Calorimetry was performed in a Mettler-Toledo RC1eMidTemp, and Differential Scanning Calorimetry in a TA Instruments Q2000 using 20mL Swissi pressure crucibles [5].
If you have process scale-up concerns or reaction calorimetry needs, please contact Don Knoechel at (knoechel@fauske.com or 630-887-5251) to discuss your process.
References
- AKTS AG, http://www.akts.com (AKTS - Thermokinetics software and AKTS - Thermal Safety software).
- Stoessel, F. (1993) What is your thermal risk?. Chemical Engineering Process, October, 68-75.
- Francis Stoessel, Thermal Safety of Chemical Processes, Risk Assessment and Process Design Book, Wiley-VCH (2008).
- The series of DSC scans using scan rates of 8.0, 4.0, 2.0, 1.0 and 0.5 °C/min were carried out on samples from a post-reaction mass from an RC run at 80°C with a 0.822 g/min EPI addition (126 min). The RC results from this run will be shared in Part 2 of this series.
- Pierre Reuse, Estimation of Time to Maximum Rate under Adiabatic Conditions (TMRad) Using Kinetic Parameters Derived from Differential Scanning Calorimetry – Investigation of Thermal Behavior of 3-methyl-4-nitrophenol. Swiss Safety Institute, Basel, Switzerland.
