Table of Contents Table of Contents
Previous Page  149 / 390 Next Page
Information
Show Menu
Previous Page 149 / 390 Next Page
Page Background

Figure 3:

Survey (left) and close-up SEM images of solid precipitates from filters,

survey scan showing rust particle (white)

Scattered in the samples a few rust particles were found accounting for a fraction of the iron

contents of the solids as found in ICP-AES. The rust particles show very different morphologies

and compositions compared to the rest of the sample and are therefore easily identified (Figure

4).

Since aluminum is a major constituent of the precipitates, solid state

27

Al-NMR was employed

to identify the actual aluminum species present. For this, only rust-free samples were used. The

peaks in the spectra were assigned by reference measurements. Samples from filters early in

the distillation process contain two different aluminum species which could be identified as

tetrachloroaluminate and aluminum chloride. Later in the process, only tetrachloroaluminate

was found in samples from filters and evaporators (Figure 5).

There are a few metals that can act as possible cations for the tetrachloroaluminate anion, with

copper and iron being the most abundant in the precipitates. Generally, all tetrachloroaluminate

salts are scarcely soluble in the product mixture of the Müller-Rochow direct synthesis, so

precipitation of the identified compounds is not surprising. Rather unexpectedly, the

precipitates comprise a high carbon content. The nature of the carbon species in the precipitates

could not be elucidated; thermogravimetric analysis in combination with mass spectrometry

(TGA-MS) shows the elimination of chloroalkanes and alkanes at elevated temperatures (>

175°C). Since the precipitates are solids, the carbon species must be of high molar masses or

polymeric or both. Since no evidence for elemental carbon was found and the hydrogen content

of the precipitates suggests a saturated linear organic compound of high molar mass, polymers

are most likely. They have to originate from reactions of the alkanes, alkenes and chloroalkanes

in the product mixture. Such reactions could be those of chloroalkanes with the Lewis-acidic

aluminum chloride to form tetrachloroaluminate and carbocations which then further react with

alkenes to form polymers via cationic polymerization.

The silicon species present were found to be mainly highly crosslinked siloxanes probably

resulting from the reaction of silane residues with moisture which occurred while taking the

samples from filters or evaporators in air.

141