SYNTHETIC STRATEGIES - FROM LABORATORY TO INDUSTRY

<< NEWER REACTIONS AND PROCEDURES: CATALYTIC AND NONCATALYTIC

SYNTHESIS OF CHEMICALS FROM CARBON DIOXIDE:Carbon dioxide - Dry Ice >>

Chapter - 13

SYNTHETIC STRATEGIES - FROM LABORATORY TO

INDUSTRY

S. Chandravathanam

INTRODUCTION

Scale-up is an act of transferring a laboratory process to the larger equipment typical of a

commercial plant, or designing a piece of commercial equipment based on research-scale

models. This is often a complex matter in which, for some processes, trail and error still

has a significant foothold. Even with careful planning and strict methodology, scale-up

can be fraught with difficulty and unexpected problems. The reasons for this are

numerous; many common laboratory methods cannot be applied at the large scale,

equipment may exhibit unexpected behaviour at sizes never used before, or critical heat

or mass transfer phenomena may not be discernible at laboratory scale.

The design of a new plant or commercialization of a new chemical process represents

a tremendous investment of time and money. The risk is considerable and the economic

penalty, if the plant or process fails to produce as expected is severe. To minimize such

risks, industries undertake lengthy and expensive process research and development

programs.

An invention might sometimes lay unused for ages without paying back in terms of

industrial realization and of profitable business for its lack of technical know-how. For

example, the pigment indigo was extracted from the sea animals during the 13th century.

One gram of the pigment could be obtained from almost 10,000 numbers of the animal.

So that it was the very expensive dye material, and the colour was restricted to the Royal

family only, and so named as the Royal purple. It was the situation till Bayer in the 19th

century studied the dye material, and started synthesizing it from the easily available raw

materials.

This text deals with the technical knowledge and on the tools that are necessary to

change an invention into a true innovation. The term scale-up has usually been explained

as how to design an industrial reactor able to replicate the results obtained in the

laboratory. These are done with most of the time innovative ideas, and sometimes with

13.2

Synthetic Strategies From Laboratory to Industry

many mistakes.The scale-up usually means that the scaling facor of 1000 times from the

laboratory process to the pilot plant scale, and further 1000 times from the pilot to the

industrial scale. Typical laboratory scale preparation lies with in the production of 100

g/day. A pilot plant will typically produce 1-50 Kg/day, and the industrial plant will be

aimed with the production of tons/day.

Few of the indispensable steps which are undertaken while going from the laboratory

scale to the commercial or industrial scale are, cost estimation, process design, pilot plant

studies may or may not be accompanying with the dimensional analysis.

COST ESTIMATION

The cost estimation for the scale-up of production is a very crucial step. For this process,

chemical engineers long been relying as a rule a thumb on the use of power law. As per

the law the investment is proportional to the scale of a production facility raised to some

constant power, characteristic of the particular process.

The so called power law of investment takes the form,

I2/I1 = k (Vt,2/Vt,1)1/n

Where I is investment, Vt is production rate and k and n are constants.

The power law is based on the `minimization law', which states that people minimize

their efforts per unit of dimension whenever a change of scale or volume is required. The

constant n is the number of dimensions of the production rate-limiting activity, typically

is either one, two or three corresponding to the linear, area and volumetric dimensions.

PROCESS DESIGN

Process Design is the heart of the process of scale-up. When the research department

discovers a new reaction to make an existing product or a new material, the process

department will have to translate these discoveries into a new process which could be

commercially and technically feasible.

The few of the factors of importance during this process design stage are,

- Production rate

- Temperature and pressure conditions of the reaction

- Every available information about operating variables, thermodynamics and kinetics of

the chemical reaction

- Stoichiometry of the main reaction

Synthetic Strategies in Chemistry

13.3

- Desired yield of the product

- Desired product purity and the possibility of the feed recycling

The importance of the process design step can be explained with one of the case studies

with the process design for the manufacture of Grignard Reagent. Grignard reagent is one

of the important reagents in organic synthesis for the introduction of alkyl groups on

carbonyl carbon atoms. It's requirement is vast but its manufacturing in large scale was

not undertaken for a long time for some reasons like the high exothermicity of the

reaction, the high inflammability of the solvent diethyl ether, and the high sensitivity of

the reagent for water, which needs to be used for cooling purpose.

Grignard reagent is prepared in situ when alkyl halide reacts with Mg metal scraps in

diethyl ether.

Through careful process design studies, these drawbacks had been overcome; with the

use of alkyl chlorides or bromides instead of iodides, the exothermicity of the reaction

can be reduced as a result of the reduction in the reaction rate of chlorides and bromides

compared to their iodide counterpart. Along with that, it could also reduce the cost as the

chloride is cheaper than the iodide. The risky diethyl ether was replaced with

tetrahydrofuran (THF), which has atleast 30 °C higher boiling point than the former. Inert

gases which can play both as the cooling system and blanket replaced water, as the

product Grignard reagent is more sensitive to water, if there is any leakage.

PILOT PLANT

A pilot plant is generally a collection of equipment designed to allow operation of a novel

process at a scale small enough to be safely manageable but large enough to provide a

realistic demonstration of operations and physical principles as they might apply in a

commercial facility and to allow the collection of meaningful engineering data for a

further scale-up.

Pilot plants provide important information on the best ways to handle reactants,

intermediates, products and waste streams, on energy transfer, on the best choice of

13.4

Synthetic Strategies From Laboratory to Industry

separation technologies and an operating procedures. Some of the many areas of process

development in which a pilot plant can play an important role are,

- Confirming the operational feasibility of a new process

- Identifying scale-up effects on yield and selectivity

- Collecting kinetics and design data

- Testing materials of construction

- Testing the operability of control schemes

- Assessing process hazards and safety issues

- Commercial viability of the raw material

- Process troubleshooting and optimization

- Testing process recycle streams

Pilot plants can be classified according to numerous criteria. Foremost is the fundamental

distinction between a pilot plant for a continuous process versus one for a batch or

semibatch process. In the continuous process, pilot plants tend to be single-purpose,

product dedicated facilities that are generally smaller. Batch pilot plants, typical of the

fine chemical industry, tend to be multipurpose. The requisite flexibility to handle a wide

variety of products and processes can add considerably to the complexity and cost of a

plant.

DIMENSIONAL ANALYSIS

Dimensional analysis is a process by which the dimensions of equations and physical

phenomena are examined to give new insight into their solutions. It is a powerful

technique for spotting errors in equations. It shows how physical dimensions of the

variables that govern a problem can be used to find physical laws. The main requirement

for dimensional analysis is dimensional homogeneity it states that all parts of an

equation must have the same dimension. In other words, we can add or compare

quantities that have similar dimensions only.

Dimensional analysis results in many advantages like

the reduction of number of variables in a variable set; if there are `n' variables or

parameters of concern, then the total number of dimensionless numbers is reduced

to n - r, where `r' is the number of basic dimensions of all the parameters of

concern (Buckingham pi theorem).

Synthetic Strategies in Chemistry

13.5

Gives the guidelines for scaling the results from model test to the full-scale.

Otherwise dimensional analysis sets the rule under which full similarity in model

test can be achieved.

Nondimensional parameters are more convenient than dimensional parameters

since they are independent of the system of units.

In the following passage two case studies of application of dimensional analysis are

given.

CASE STUDY 1: Correlation between Meat Size and Roasting Time

What is the roasting time for two times the mass of the meat?

The main parameters of concern for this problem are listed below along with their

dimensions.

Symbol

dimension

Physical quantity

roasting time

meat surface

thermal diffusivity

L2T-1

surface temperature

temperature distribution

The higher the heat conductivity k, of the meat, the faster is the cooking.

The higher the heat capacity ρCp, the slower is the heat transfer.

∴Thermal diffusivity a = k/ρCp

The total number of dimensionless numbers are 5-3 = 2.

∏1 = T/T0

(T0-T)/T0)

∏2 = aθ/A

When the temperature distribution T/T0 is achieved throughout the meat then it can be

said that the meat is cooked.

T/T0 = f (F0)

13.6

Synthetic Strategies From Laboratory to Industry

aθ/A = idem

T/T0 = idem

θ α A idem = identical

This equation relates the roasting time with the area of the meat. But in reality the

roasting time needs to be calculated with respected to the mass of the meat.

mass is related to area with the following equation.

m = ρV α ρL3 α ρA3/2

the density ρ remains the same irrespective of the meat size. Therefore,

ρ = idem

and

A ∝ m2/3

θ2/θ1 α (m2/m1)2/3

That is,

θ α m2/3 = m0.67

∴doubling the mass of meat, the roasting time increases by, 22/3 = 1.58 times. The final

equation contains only two parameters, instead of the initial five parameters, therefore,

the number of supplementary experimental runs is getting drastically reduced.

CASE STUDY 2: Homogeneous Irreversible 1st Order Reaction in a Tubular

Reactor

How much is the volume and residence time of the reactor to be increased for the

increase of volume throughput by a factor of n (qT = nqM)?

The important parameters of concern are, v - flow rate, d, L diameter and length of the

tubular reactor, ρ,µ - fluid density and viscosity respectively. T0 inlet temp.

cin, cout inlet and outlet conc. keff. effective reaction rate constant

keff. = k0 exp(E/RT)

The parameters of mass and heat transfer are, D - Diffusion coefficient, Cp heat

capacity, k - thermal conductivity, cin ĆHR - heat of reaction per unit time and volume,

T0 - inlet temp., ĆT - temp. difference between fluid and tube wall.

The complete relevant list of parameters is,

v, d, L, ρ, µ, cout, cin, k0, E/R, D, Cp, k, cinĆΗR, T0, ĆΤ

Synthetic Strategies in Chemistry

13.7

From the pi theorem, the number of nondimensional parameters should be, 15 - 6 = 9.

The obvious five nondimensional numbers are,

L/d, Cout/Cin, E/RT0 and ĆΤ/Τ0

The remaining four nondimensional numbers are derived using the following

combination of the parameters.

= (k0 τ)-1 (mean residence time τ

L/v at pipe flow)

Π1 = v/Lk0

Π2 = µ/ρd2k0

= (k0 τ Re L/d)-1

Π3 = D/L2k0

= (k0 τ Re Sc L/d )-1

= Da-1 (Da = Damkohler number)

Π4 = (ρCpΤ0) / cin ĆΗR

Π5 = (kΤ0) / cin ĆΗR d2 k0 = (k0 τ Re Pr Da L/d)-1

∴The nine dimensionless numbers are,

L/d, Cout/Cin, E/RT0, ĆΤ/Τ0, κ0τ, Re, Sc, Pr, Da

During scale up

· No change of reaction temp. ∴T0 and k0 are constant

· No change in the physical and chemical partners of the reaction ∴the kinetic and

material numbers, remain unchanged.

· L/d = identical

· To attain specified degree of conversion, cout/cin = identical

Therefore, the following two are the only two parameters need to be adjusted.

(k0 L)/v

Re = (v d ρ) /µ and k0τ

But it is impossible to have both,

v L = idem and L/v = idem and L/d = idem

For the given condition of qT = nqM, , dT = ndM,

q ∝ vd2 and vTdT2 = n vMdM2

13.8

Synthetic Strategies From Laboratory to Industry

as, Re α v d = idem

VT = (VM)/n

VT = n3 VM and hence τΤ = n2 τΜ (τ = V/q)

Where,

M - model scale

T - technological or industrial scale

q volume throughput

n factor

V volume of the reactor

The results state that for the increase of volume throughput n times the volume of the

reactor needs to be increased by n3 times and the residence time needs to be increased by

n2 times.

REFERENCES

1. Kirk-Othmer, Concise Encyclopedia of Chemical Technology, Fifth Edition, Vol. 2,

John-Wiley & Sons, 2007.

2. John J. Mcketta (Ed.,), Encyclopedia of Chemical Processing and Design, Vol. 49,

1994.

3. Marko Zlokarnik, Scale-up in Chemical Engineering, , Wiley-VCH, 2002.

4. Keld Johansen, `Aspects of scale-up of catalyst production', Studies in Surface

Science and Catalysis, 143 (2002) 1.

5. Gianni Donati, Renato Paludetto, Catalysis Today, 34 (1997) 483.

Table of Contents: