White Paper Risks from Rising Temperature

Safe Chemical Development Practices

Authors: Urs Groth, Fabio Visentin, Mettler-Toledo GmbH, Switzerland

Many chemical processes are exotherm and release a defined amount of energy. If the
energy released can't be removed instantaneously, the temperature will rise. Even those pro-
cesses that are intended to be run isothermally will show a small deviation from the target
temperature, which may have important implications on reaction kinetics and safety of the
process. Temporary temperature changes are due to physical effects, and vary with the reac-
tant addition rate, the heat release rate, the process dynamics, and the reactor vessel. Such
behavior may also be caused by limitations of the temperature control or the heating and
cooling capacity. This occurs when reactions are fast and strong and the heat release is
larger than the heat removal capacity. When this happens, a defined amount of heat is accu-
mulated temporarily, and released again over time. Consequently, the temperature changes
initially, but returns to the defined target temperature at the end of the reaction. For any
researcher involved in chemical process scale-up, understanding this temperature change,
and the associated heat that is accumulated by the reaction, is critical to understand the
safety of the process.

In order to truly understand the risks, the following three questions below must be addressed.
1. "Why and when does thermal accumulation occur?"
2. "Is accumulation important to consider, and how big is it?"
3. "What is the impact of an incorrect calculation of the accumulation?"

This paper will provide examples from both the lab and pilot plant which highlight methods
to answer these questions and assess the implications of changing temperature regimes
under which chemical reactions are carried out.

Contents
1 Introduction
2 Why and When Does Thermal Accumulation Occur?
3 Is Accumulation Important to Consider, and
How Large Is It?
4 What is the Impact of an Incorrect Calculation
of the Accumulation?
5 Experiments at the Lab Scale
6 Conclusions
7 Appendix
White Paper
1 Introduction
There is no question that having powerful and fast cooling is a huge benefit when studying reactions under iso-
thermal conditions on the gram scale and above. However, as chemical development moves towards the liter
scale, every system will exhibit increasing degrees of non-isothermal behavior, even when the temperature is
supposed to remain constant. From a scalability point of view, investigations should answer not only the question
of how much heat is released, but also how fast the heat is released. Even when enthalpies, the heat transfer
coefficient, and the specific heat of the reaction mass are known, the real heat release pattern may not necessarily
be understood. Thus, it is critical to quantify thermal accumulation for reactions where the temperature changes
slightly, or the reactions are executed under changing temperature conditions.

Whenever a process is assumed to be fully isothermal, but is in fact not, the heat of reaction trend, simply
calculated as heat flow, represents the heat removal behavior, but not the true heat release based on the chemical
conversion. This is because heat is temporarily absorbed (accumulated) by the reactor and its contents, and
subsequently released and dissipated into the jacket over time. As a result, the heat flow curve does not allow
researchers to draw correct conclusions on the course of the reaction making it impossible to determine reaction
kinetics.

Most commercially available calorimeters are based on a double-jacketed reactor connected to a cryostat, and
are therefore capable of measuring reaction profiles correctly for true isothermal situations only. However, the
thermal accumulation is often neglected or compensated for by an approximation, resulting in inaccurate heat of
reaction calculations. Consequently, the heat trend does not represent the true progress of the reaction, nor can
it be used to estimate reaction kinetics. By answering the three questions posed above, researchers understand
the risks of accumulation early in development, and quickly make decisions to develop the safest process.

2 Why and When Does Thermal Accumulation Occur?

The majority of chemical processes are studied under isothermal conditions in the lab, but then they are run under
pseudo- or non-isothermal conditions at scale. Although a reaction might be assumed to operate under isothermal
conditions, all reactions comprise some share of non-isothermal behavior which causes energy to accumulate.
Accumulation can happen intentionally or unintentionally, and may have significant implications to process safety.
The examples that follow in this paper highlight how reaction temperature regimes will impact thermal accumula-
tion and its relevance for reaction safety.

What Is Isothermal Calorimetry?

Isothermal calorimetry applies when a constant reaction mass temperature can be maintained at all times during
the reaction. In a heat flow calorimeter, the temperature of the reactor content (Tr) is controlled by varying the
temperature of the heat transfer liquid (Tj) in the jacket. When an exothermic reaction occurs and heat is gen-
erated, the temperature of the heat transfer medium in the reactor jacket is constantly adjusted to keep Tr constant.
As a result of the variable jacket temperature, any heat evolved during the process is removed immediately and
dissipated into the jacket. Thus, the temperature of the reaction mass exactly follows the set temperature and no
heat is accumulated (Figure 1).

2 White Paper
METTLER TOLEDO
 For a perfectly isothermal reaction, the heat of reac-
qr
tion can be calculated as follows:

q r = q flow = UA (Tr−T j ) [eq. 1]

Temperature

Tr (qflow = heat removed by the jacket)

Heat of reaction (qr) Realistically, a completely isothermal behavior with

Reaction mass temperature (Tr) no change of Tr or heat accumulation cannot be
Tj Jacket temperature (Tj) accomplished due to heat capacities, controller
Time
delays, and time constants of the reactor system.
Consequently, every commercial reactor behaves non-
Figure 1: Perfectly isothermal reaction
isothermally to a certain degree.

What Is Non-Isothermal Calorimetry?

Non-isothermal calorimetry applies when the temperature in the reactor changes during the process. Temperature
variations during a reaction can either be intentional or unintentional.

A temperature change of the reactor content accumulates a defined quantity of the reaction energy, known as
thermal accumulation (eq. 2). The degree of accumulation is relative to the temperature change, the reaction mass
(mr), and the reactor contents' heat capacity [∑(mx cpx)]. •

qaccu = dTr/dt ∑(mr cpr + mi cpi)

• • • [eq. 2]
(qaccu = heat accumulated by the reactor contents)

If the evaluation of the heat, with respect to time is to be representative of the true heat of reaction, the heat accu-
mulation is an essential component and must be included in the overall heat balance (eq. 3).

q r = UA (Tr−T j ) + dTr/dt ∑(mr cpr + mi cpi)

• • • [eq. 3]
(heat of reaction = heat removed by the jacket + heat accumulated by the reactor contents)

The key elements of the thermal accumulation calculation are the mass balance and the specific heat of the reac-
tion contents. While the major part of the energy is stored in the reaction mass (mi cpr), an unknown fraction •

is absorbed in the reactor wall (glass, metal) and the inserts (probes, stirrer, baffle). Therefore, the fractions of
energy required to heat or cool the inserts (mi cpi) and the reactor cannot be neglected and must be included in
•

the calculation of the specific heat capacity.

(Tr−Ta) 1
cpr = (UA • − mi cpi) • • [eq. 4]
dTr /dt mr

White Paper
METTLER TOLEDO
3
White Paper
3 Is Accumulation Important to Consider, and How Large Is It?
Example 1 – Reaction with Minor Non-Isothermal Behavior
In the reaction below (executed in an OptiMax™ HFCal with a 1 L reactor), the reactant is added at a target tem-
perature of 50 °C. As shown in Figure 2, the temperature of the reaction mass (Tr) deviates by only 0.4 °C from
the target temperature. However, there is still a significant accumulation of unreacted material as seen in qr. The
reaction continues after dosing is completed (02:46:00). Since the reaction itself is not very strong (qr max =
35 W) and the cooling system fast and powerful, the heat of reaction is removed instantaneously. Thus, the
reaction temperature is nearly identical to the target temperature. This slight change of the reaction temperature is
unintentional and typically considered as isothermal.

40 840
85
Maximum heat of reaction = 35 W 35
qr 820
80
30 800
75 Heat of reaction (qr) 25 780
Reaction mass temperature (Tr)
20
Tset (°C), Tj (°C), Tr (°C)

70 Jacket temperature (Tj) 760

15
Tset 740
65

qr_hf (W)
Reactant addition (Mr)

Mr (g)
10
720
60 5
700
55 0
Maximum reaction mass temperature = 50.4 °C 680
Tr -5
50 660
Tset = 50 °C -10
45 Tj 640
-15

Mr 620
40 -20

-25 600
02:40 02:45 02:50 02:55 03:00 (hh:mm)
Figure 2: Pseudo-isothermal reaction (only little change of reaction temperature)

Example 2 – Reaction with Significant Non-Isothermal Behavior

The reaction in Example 2 is intended to be an isothermal reaction at set-point of 40 °C, but becomes non-
isothermal as the heat produced is much larger than the heat removal capacity of the jacket. As a result,
the remaining energy accumulates in the reactor, causing the reaction temperature (Tr) to rapidly rise from
40 °C to 96 °C (Figure 3).

340 Heat of reaction (qr) Maximum heat of reaction

= 1280 W 2200
320 Heat absorbed (qaccu) 1200
300 Heat removal (qflow) 2000
Reaction mass temperature (Tr) 1000
280 1800
qaccu (W), qr_hf (W), qflow_hf (W)

Reactant addition (Mr) Maximum heat absorbed

260 800 1600
240
1400
220 600
200 1200
Tr (°C)

Maximum heat removal

Mr (g)

180
qr 400
1000
160
qflow 200 800
140
120 qaccu 600
0
100 400
Maximum reaction mass -200
80 Mr temperature = 96 °C 200
60 Tr
-400 0
40

03:25 03:30 03:35 03:40 03:45 03:50 03:55 04:00 (hh:mm)

Figure 3: Unintentional non-isothermal reaction due to strong exotherm

4 White Paper
METTLER TOLEDO
 Consequently, the stored heat is released and dissipated into the jacket once the heat production becomes less
than the heat removal capacity. The qflow trend represents the heat removal across the reactor wall, and is 430 W
at its maximum. This decreases once the heat production becomes less than the heat removal capacity. The area
under the qflow curve reveals the total heat released during the reaction (∆Hr). When isothermal calorimetry is
applied, only the qflow trend is measured. However, by measuring the heat of accumulation (qaccu) and adding it
to qflow, the true heat of reaction profile (qr) is obtained, with its maximum at 1280 W. The area under both curves
remains the same, since the total amount of energy released is unchanged. However, the true qr graph reveals
the impact of the heat accumulation, as well as the importance of measuring and understanding both the isother-
mal and non-isothermal effect, with respect to time. The maximum heat flow in this example corresponds to
approximately 260 W/L. This is important to consider because the plant heat removal limit is typically < 35 W/L
and as presented, this process would not be safe to scale up.

Intentional Non-Isotherm Reaction Regime

Unlike the previous examples, reactions are often carried out intentionally under changing temperature conditions
to improve reaction kinetics, eliminate by-product formation, or improve conversion rate and yield. It may be
beneficial to run a specific reaction in a temperature ramp, where the reactant addition takes place partially or fully
during the temperature ramp. In such a process, the heat of reaction is often used to increase the temperature,
improving the economy of process.

Temperature control in such cases can vary and increase either linearly (Figure 4) or pseudo-adiabatically
(Figure 5). Sometimes however, the temperature of the reaction mass is maintained during the reactant addition,
and increased towards the end of a process (Figure 8). In the semi-batch reaction below (Figure 4), the reactant
is added during a temperature ramp. Subsequent to the start of the temperature ramp, the jacket temperature (Tj)
is significantly above the reaction temperature. Once the reactant addition begins and the reaction starts, the
jacket temperature is automatically reduced by the temperature control system, such that the reactor temperature
(Tr) follows the defined temperature ramp. After the temperature reaches its target value, the jacket temperature
is reduced further to control the reactor temperature at the target (50 °C).

Heat of reaction (qr) 220

18
110 qr Reaction mass 200
temperature (Tr) 16
100 Jacket temperature (Tj) 14 180
Reactant addition (Mr) 12 160
90
10
140
80 8
Tr (°C), Tj (°C)

6 120
qr_hf (W)

Mr (g)

70
4 100

60 2 80
0
50 60
Tj -2
-4 40
40 Tr
-6 20
30
Mr
-8
0
-10
20 -20

00:50 01:00 01:10 01:20 01:30 01:40 (hh:mm)

Figure 4: Semi-batch reaction carried out in temperature ramp

White Paper
METTLER TOLEDO
5
 The overlapping effects (temperature ramp, reactant addition, and the reaction itself) require an accurate deter-
White Paper
mination of the heat flow across the reactor wall, the heat accumulation, and the energy needed to adjust the tem-
perature of the added reactant (qdos).

q r = UA (Tr−T j ) + dTr /dt ∑(mr cpr + mi cpi) + dm/dt cp dos (Tr−Tdos)

• • • • • [eq. 5]
(heat of reaction = heat removed by the jacket + heat accumulated by the reactor contents + heat associated with reactant addition)

This places high demands on the measuring system accuracy and data evaluation, since the dynamic compen-
sation of the reactor wall capacity, the reactor time constant, the specific heat capacity, and the heat capacity of
the inserts significantly influence the quality of the result.

Adiabatic Control Regime – a Special Case of Non-Isothermal Operation

The adiabatic, or pseudo-adiabatic, control regime

qr is a special case of the non-isothermal reaction
control, where all heat released by the reaction
Temperature

accumulates in the reactor. The reactor temperature

increases propor­tionally to the conversion of the
reactants (Figure 5). In such situations, the jacket is
Heat of reaction (qr)
Tr usually controlled to behave like insulation.
Reaction mass temperature (Tr)
Tj Jacket temperature (Tj)

Time
Figure 5: Adiabatic reaction control

Adiabatic control can be dangerous if the reaction and all possible side-reactions are unknown or not sufficiently
known. However, a pseudo-adiabatic control regime is commonly used to improve the conversion, due to
increasing reaction temperature. In some cases, adiabatic control in a lab reactor is applied to test reactions at
lab scale with regard to safety, simulating cooling failure scenarios, or situations of poor cooling at plant scale.

4 What Is the Impact of an Incorrect Calculation of

the Accumulation?
The quality of the final heat flow results depends on various factors, such as the accuracy of the measurement,
the correctness of the mass balance, the calculation of the specific heat capacity, and the degree to which a
process is non-isothermal. As the percentage of accumulation in the total heat becomes larger, the requirement
for precise and high-quality calculations of heat capacity and accumulation become more important.

If a reaction is entirely isothermal, the specific heat (cpr) is not taken into account at all, and thus incorrect
determination or potential errors in cpr have no impact on the final result. However, as a reaction is performed
non-isothermally, the precise measurement of reactor and reactor content heat capacity becomes increasingly
important. When reactions are run in temperature ramps, adiabatically or pseudo-adiabatically, the accumulation
is the predominant, critical factor to be taken into account to calculate the heat of reaction and determine the
fundamental safety of the process. In such cases, the heat flow may become almost zero and is therefore,
negligible.

Precision reaction calorimeters (see Appendix 1) determine the specific heat capacity, with the highest possible
accuracy, compensating for heat accumulation in the inserts and the reactor wall.

6 White Paper
METTLER TOLEDO
 Through reactor-specific models, dynamic effects occurring during temperature ramps can be corrected, thus
accounting for the reactor wall material and thickness. This ensures that the jacket temperature applied for the
calculation always represents the true jacket temperature, under all temperature regimes throughout a chemical
reaction.

Case Study – Upscaling a Non-Isothermal Process from Lab to Plant

The following industrial example deals with the sulfonation of phenol and compares data obtained in the lab
with pilot plant data. 2-hydroxybenzene-sulfonic acid is widely used for dyes, medical intermediates, photo-
developing chemicals, and cosmetic preparations.

Major Product

OH OH OH
SO3H
+ H2SO4 +

SO3H
2-hydroxybenzene-sulfonic acid 4-hydroxybenzene-sulfonic acid
Figure 6: Reaction scheme

The sulfonation of phenol is carried out by adding concentrated sulfuric acid (98%) into liquid phenol at 60 °C
(Figure 6). The reaction is initiated instantaneously with the acid addition, kinetics is extremely fast, and it
releases large amounts of energy. However, as the reaction does not go to completion at 60 °C, it is run under
pseudo-adiabatic conditions in the plant reactor, where the heat of reaction is used to increase the reaction
temperature. Because the enthalpy and the heat release profiles are unknown, a number of lab experiments were
carried out at 1 L scale using an OptiMax HFCal reactor under different conditions.

Pilot Plant Data

The temperature and dosing profiles in Figure 7 illustrate the pseudo-adiabatic reaction regime and the reaction
progress of the plant reactor. After starting the addition of the sulfuric acid, the temperature of the reaction mass
increases quickly in a pseudo-adiabatic manner due to the energy released by the reaction. The jacket does not
cool down, leaving the inside temperature rising pseudo-adiabatically. Upon reaching the target temperature of
100 °C, the reaction temperature is subsequently controlled to maintain 100 °C until the reaction is completed.

125 Reaction mass temperature (Tr)

115 Jacket temperature (Tj)
Reactant addition (Mr)
105

95 Tr = 100 °C
85
Temperature (°C)

75 Tr
65 Tr = 60 °C
55 Tj

45

35

25
Mr
15
07:26:24 07:40:48 07:55:12 08:09:36 08:24:00 08:38:24 (hh:mm:ss)

Figure 7: Temperature and dosing profiles from the plant reactor

White Paper
METTLER TOLEDO
7
 Due to concerns with heat accumulation and risks of exceeding the temperature limits, a more thorough analysis
White Paper
was conducted in the laboratory at a 1 L scale. A number of lab experiments are carried out in an OptiMax HFCal
reactor to assess the thermal data and process parameters, and to simulate different operating regimes, such as
isothermal, non-isothermal, and pseudo-adiabatic reaction. This allowed researchers to understand and optimize
the effect of process parameters such as temperature control and dosing on the system.

5 Experiments at the Lab Scale

Assessing the Reaction under Isothermal Conditions
A vessel was filled with phenol and kept at temperature of 60 °C. Dosing of sulfuric acid took place in four stages
at rates from 9.2 g/min to 16.6 g/min. With the first addition of the sulfuric acid, an immediate and violent reac-
tion can be observed. From the heat release pattern it is obvious that the reaction is almost dosing-controlled at
the beginning of the reaction, while the reaction rate diminishes with the decreasing concentration of phenol. After
the final dosing step, the reactor rate slows down and at 1:40:00, the temperature is increased from 60 °C to
100 °C to bring the reaction to completion (Figure 8).

500 qr Heat of reaction (qr)

Reaction mass 80 1800
450 temperature (Tr)
Jacket temperature (Tj) 1600
60
400
Reactant addition (Mr)
1400
350 40
1200
Tj (°C), Tr (°C)

300

qr_hf (W)
20 1000

Mr (g)
250
800
0
200
600
150 Tj -20
Mr
Tr = 100 °C 400
100
Tr = 60 °C -40 200
50 Tr
0
-60
00:40 01:00 01:40 01:40 02:00 02:20 02:40 (hh:mm)
Figure 8: Temperature and dosing profiles of the isothermal reaction followed by a temperature ramp

In the isothermal phase of the reaction, the maximum heat generation rate reaches approximately 50 W,
corresponding to approximately 79 W/kg. The conversion of the accumulated reactants during the temperature
ramp generates approximately 90 W, which corresponds to 143 W/kg. Assuming a maximum heat removal
capacity of 30 W/kg to 35 W/kg at production scale, these values are significantly higher than the cooling
capability of a standard plant reactor.

The total heat of reaction is 104.3 kJ (32.8 kJ/mol), of which approximately 60% (67 kJ) are released during
the isothermal phase. This leaves a significant amount of accumulated reactant (about 40%) corresponding to
approximately 37 kJ released during the heating ramp to 100 °C.

8 White Paper
METTLER TOLEDO
 From this isothermal study, researchers identified that the synthesis of the hydroxy-benzene-sulfonic acid is a fast
and very powerful reaction, with little initial reactant accumulation in the beginning. However, it slows down signif-
icantly after about 50% of the phenol is converted, leaving a substantial reactant accumulation and thus, creating
a safety risk. In this case, a sufficiently high conversion can only be achieved if the reaction mixture is heated to
100 °C after the dosing is completed.

Assessing the Reaction under Non-Isothermal Conditions (Temperature Ramp)

Based on the results of the previous procedure, the same reaction was run while the temperature was ramped in
a controlled manner from 60 °C to 100 °C, with simultaneous controlled sulfuric acid addition (Figure 9).

400 Heat of reaction (qr) 50 900

qr Reaction mass
temperature (Tr) 40 800
350
Jacket temperature (Tj)
30 700
Reactant addition (Mr)
300
Tset (°C), Tj (°C), Tr (°C)

20 600

250 500
10

qr_hf (W)

Mr (g)
200 0 400

-10 300
150
Mr
Tr = 100 °C -20 200
100 Tr
-30 100

50
Tj
Tr = 60 °C -40 0

00:50 01:00 01:10 01:20 01:30 01:40 (hh:mm)

Figure 9. Sulfuric acid addition and concurrent temperature ramp

Similar to the procedure described in the previous experiment, the reaction starts instantaneously with the sulfuric
acid addition, releasing a large amount of heat. Due to the large release of energy and despite the powerful
thermostat control, the actual temperature deviates slightly from the desired ramp profile. However, unlike in the
previous experiment, there is hardly any reactant accumulation at the end of dosing (approximately 4%), as
the temperature is constantly raised to 100 °C.

Compared to the previous experiment with an isothermal phase followed by a temperature increase, the maximum
heat release rate is nearly identical (approximately 50 W equivalent to 79 W/kg). This may be caused by the
mass transfer limited reaction kinetics, dictated by the two-phase reaction mass and the mixing regime. The total
heat of reaction is 51.8 kJ (32.6 kJ/mol). However, thanks to the accurate temperature ramp control and the non-
isothermal reaction regime, the reaction is better controlled, significantly more efficient, and safer than the original
isothermal process.

White Paper
METTLER TOLEDO
9
 Assessing the Reaction under Pseudo-Adiabatic Conditions
White Paper
From the two previous experiments it is apparent that the non-isothermal process control is more efficient and,
because of the much smaller accumulation, also safer. As a result of these findings, the process was further
examined under pseudo-adiabatic conditions, followed by isothermal temperature control at 100 °C, mimicking
the plant operation.

Heat of reaction (qr)

320
Reaction mass 50 900
300 temperature (Tr)
qr
280 Jacket temperature (Tj) 40 800

260 Reactant addition (Mr)

30 700
240
220 20 600
Tj (°C), Tr (°C)

200

qr_hf (W)
500

Mr (g)
10
180
160 0 400

140
-10 300
120 Mr

100 -20 200

Tr Tr = 100 °C
80
-30 100
60 Tj
Tr = 60 °C
40 -40 0

01:00 01:10 01:20 01:30 01:40 01:50 02:00 (hh:mm)

Figure 10: Reaction under pseudo-adiabatic conditions

The results in Figure 10, show a similar pattern as already observed in the non-isothermal experiment, based on
the temperature ramp with one major difference; while the reactant accumulation at the end of dosing is only
minor, the heat accumulation upon reaching the final temperature of 100 °C is significant (approximately 43%).
This occurs because the heat production is much larger than the heat removal capacity. Thanks to the ability of
the OptiMax HFCal to independently calculate and display the energy stored during the pseudo-adiabatic tempera-
ture increase, the true heat release pattern (qr) can be characterized precisely, providing essential process details.

Results of the Lab Scale Experiments and Comparison with the Plant Data
From the overview in Table 1, it can be noted that any of the non-isothermal approaches reduce the reactant
accumulation to a point where the reaction almost becomes dosing-controlled. The reaction however, appears to
be mass transfer limited, independent of the thermal regime.

Results Determined by the Three Different Approaches

Reaction 1 Reaction 2 Reaction 1
Isothermal Non-isothermal Non-isothermal
60 °C with T-ramp to 100 °C (T-ramp) (pseudo-adiabatic)
qr 32.8 kJ/mol 32.6 kJ/mol 35.6 kJ/mol
qr max approximately 50 W equivalent to 79 W/kg
Heat accumulation at reaching 100 °C 5% 3% 41%
Reactant accumulation at end of dosing 40% 4% 3%

Table 1: Summary of experimental results

Based on the lab experiments in OptiMax, it becomes obvious that the most appropriate reaction conditions are
very similar to the current operating conditions in the plant.

10 White Paper
METTLER TOLEDO
 Comparing the OptiMax temperature profiles with the plant data presents a very good match across a wide range
(Figure 11). The two temperature profiles deviate only slightly upon reaching the target temperature of 100 °C
which is due to the plant's temperature control systems.

120 Heat of reaction (qr)

Tr lab temperature profile 60 1000
qr
110 Tr plant reactor temperature profile
50 900
Reactant addition (Mr)
100 40 800
Tr = 100 °C
90 Tr plant Tr lab 30 700

qr_hf (W)
20 600
Tr (°C)

Mr (g)
80

10 500
70

0 400
60
Tr = 60 °C
-10 300
50
Mr -20 200
40
-30 100

01:00 01:10 01:20 01:30 01:40 01:50 (hh:mm)

Figure 11: Comparison of plant data with data from the lab reactor OptiMax

In conclusion to this sulfonation example, a pseudo-adiabatic control regime, combined with an adjustable
dosing rate, provides the shortest cycle time and the safest reaction conditions. Further investigations, however,
could become necessary to examine the impact of the stirrer type and speed on mixing and mass transfer.

6 Conclusions
A non–isothermal reaction regime is frequently used in production to reduce the risk of a large accumulation of
reactant or heat, and to achieve a higher selectivity, yield, and quality. In order to truly understand safety risks, the
degree and impact of thermal accumulation must be considered and addressed. Therefore, it is essential that a
reaction calorimeter allows the precise investigation of qr based on all components, not just heat flow, and under
all temperature regimes. On the basis of a series of reactions, the capabilities, performance, accuracy, and preci-
sion of the METTLER TOLEDO calorimeters were demonstrated using OptiMax HFCal. The experiments and their
results confirmed that the Reaction Calorimeter is capable of running non-isothermal calorimetry, due to the ability
to compensate for dynamic effects of the reactor system (reactor-specific Ta model), and the determination of the
specific heat and the accumulated heat.

The heat of reaction, the reaction enthalpy, and the temperature profiles of the sulfonation of phenol were studied
under three different conditions at laboratory scale, and compared with the same reaction performed at production
scale. In this case study METTLER TOLEDO reaction calorimeters clearly provide answers to the three critical ques-
tions posed in this paper. The results prove the precision of OptiMax reaction calorimeters, and highlight the fact
that the equipment can easily deal with fast and highly exothermic reactions under challenging and problematic
non-isothermal conditions for scale-up purposes.

White Paper
METTLER TOLEDO
11
 While calorimeters based on commercially available jacketed lab reactor-cryostat combinations may provide
White Paper
sufficiently good information for fully isothermal reactions, they can provide false or inappropriate data in the event
of a non-isothermal reaction regime. As seen in the examples in this paper, comprehensive overall assessment
of all thermodynamic and system-related effects is critical so, reactions which are carried out in a non-isothermal
manner can be effectively analyzed and interpreted with confidence. This allows the calculation of the heat of
reaction with the highest possible accuracy, and subsequently supports use of that data for the calculation
of reaction kinetics and process simulations. The best thermostat design also ensures fast response, precise
control and immediate cooling. See Appendix 1 for more information.

7 Appendix
Appendix 1: Need for an Excellent Thermostat
METTLER TOLEDO thermostats are equipped with unique instant cooling principles to provide fast response and
accurate control to guarantee the best and safest working conditions. The use of 'designed for purpose' systems is
critical for the delivery of quality calorimetric data. Heating

Cooling

The RC1e Reaction Calorimeter

The heating and cooling system of the Reaction Calorimeter RC1e® (Figure 12) is based on the concept of circu-
lating heat transfer fluid with a high velocity (approximately 1.2 L/sec) and a large amount of pre-cooled oil. While
Compressor

the high velocity ensures fast response on temperature

Heating changes, the reservoir with the pre-cooled oil guarantees
Condenser
instantaneous cooling in case of a large exotherm or emergency. Alongside the algorithms that ensure the highest
precision of calculation, the thermostats are designed to quickly respond to changing conditions in the reactor in
order to maintain the desired temperature accurately. Expansion
valve

Cooling

RC1e
Oil system to control temperature
in reactor
Heating
Instant cooling − cold oil bath
used to perform cooling when needed
(5 L are ready to use)

Cooling
External cooling (water or cryostat)
Figure 12: RC1e Reaction Calorimeter thermostat concept
www.mt.com/rc1

12 White Paper
METTLER TOLEDO
The EasyMax and OptiMax Reaction Calorimeter
The temperature control system of the EasyMax™ and OptiMax™ Reaction Calorimeter (Figure 13) uses the con-
cept of electrical heating and Peltier-based solid-state cooling technology. Both heating and cooling are extremely
fast due to the compact thermostat design, and enable the system to achieve temperatures well below the freezing
point without the use of a cryostat. As a result of the design, the space requirements are extremely modest.

EasyMax and OptiMax

Electrically-heated jacket to control
reaction temperature

Instant cooling − Peltier elements to

Heating provide instant cooling when needed

Cooling
External cooling (water or cryostat)
Figure 13: EasyMax and OptiMax solid-state thermostat concept
www.mt.com/HFCal
Compressor

Heating
Issues Using Calorimeters on the Basis of Traditional JacketedCondenserLab Reactor

Jacketed lab reactor systems (Figure 14) are typically operated with commercially available cryostats that are
connected to the jacketed reactor with hoses of undefined length, diameter, and shape. Because the velocity in
Expansion
such a system is much lower compared to an RC1e (about 20% to 60% of an RC1e), the fluid circulation is
valve

restricted which reduces heating and cooling performance. The response Cooling
time for immediate cooling in case of
a strong exotherm, may be insufficient making temperature Heating
control difficult and the system sluggish.

Cooling

Other Jacketed Lab Reactor

Compressor
Calorimeters – Cryostat based
Heating
Heating
Condenser Oil circulating into the jacket

Expansion
External cooling via cooling system
valve
Cooling
(refrigeration system), needed to cool
Cooling
all oil (orange) before cooling
Figure 14: Concept of a jacketed lab reactor based calorimeter

Due to the system concept and the fact that the reactors are not accurately characterized with regard to the ther-
modynamic properties, the determination of the parameters relevant for non-isothermal calorimetry becomes more
difficult and inaccurate. While this is not an issue if the reaction is run fully isothermally, the heat of reaction
calculation is compromised in the event of a non-isothermal
Heating
reaction regime. As temperature may vary during the
reaction, running a reaction isothermally is virtually impossible and requires an understanding of the thermal
behavior of the reactor contents, as well as the reactor itself.

Cooling

White Paper
METTLER TOLEDO
13
 Appendix 2: Literature References
White Paper
1. Becker F., 1968, Thermokinetische Messmethoden, Chemie Ingenieur Technik, Vol. 19, pp. 933 – 980.
2. Blackmond D.G., Rosner T., Pfaltz A., 1999, Comprehensive Kinetic Screening of Catalysts Using Reaction
Calorimetry, Organic Process Research & Development, 3, pp. 275 – 280
3. Gigante L., Lunghi A., Martinelli S., Cardillo P., 2003, Calorimetric Approach and Simulation for Scale-Up
of a Friedel-Crafts Reaction, Organic Process Research & Development, 7, pp. 1079 – 1082.
4. Regenass W., 1997, The Development of Stirred-Tank Heat Flow Calorimetry as a Tool for Process
Optimization and Process Safety, Chimia, Vol. 51, No. 5, pp. 189 – 200.
5. Stoessel F., 2008, Thermal Safety of Chemical Processes, Wiley-VCH, Weinheim
6. Stoessel F., 1997, Applications of Reaction Calorimetry in Chemical Engineering, Journal of Thermal
Analysis, Vol. 49, pp. 1677 – 1688.
7. Stoessel F.; Ubrich O., 2001, Safety Assessment and Optimization of Semibatch Reactions by Calorimetry,
Journal of Thermal Analysis and Calorimetry, Vol. 64, 61.74
8. Visentin F., Puxty G., Kut O.M., Hungerbühler K., 2006, Study of the Hydrogenation of Selected Nitro
Compounds by Simultaneous Measurements of Calorimetric, FT-IR, and Gas-Uptake Signals, Ind. Eng.
Chem. Res. 2006, 45, pp. 4544 – 4553.
9. Groth U., Mettler-Toledo GmbH, 08/2012, Understanding the Risks (White Paper)
10. Groth U., Mettler-Toledo GmbH, 11/2013, Safety by Design (White Paper)

Mettler-Toledo GmbH, AutoChem www.mt.com/HFCal

Sonnenbergstrasse 74 For more information
CH-8603 Schwerzenbach, Switzerland
Phone +41-44 806 7711
Fax +41-44 806 7290

E-Mail autochem@mt.com
Internet www.mt.com/autochem

Subject to technical changes.

© 11/2016 Mettler-Toledo GmbH, AutoChem