| Pages:
1
2 |
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Benzyl and Benzal chloride, Benzaldehyde and Benzoic Acid - Illustrated Practical Guide
This article is part of a series of Illustrated Key Syntheses in Chemistry which I intend to post, with the purposes and guidelines outlined in the
thread of the same name in the General Chemistry section.
I shall attampt to ensure that all the relevant information is contained in a single post. If sufficient reason arises for something to be changed or
edited, I shall edit the entire post.
Aim
The major aim of this project is to present a clear and reliable procedure for the preparation of pure benzaldehyde from toluene in good yield, based
on original work and using readily available chemicals. Benzaldehyde is a gentle chemical, easily oxidized by air - in moist exposed layers one can
observe all the benzaldehyde converted to benzoic acid crystals and traces in the course of a few hours. Benzaldehyde is not foreign to the human
organism - its adduct with hydrogen cyanide (mandelic nitrile) is present bound to sugar as amygdlain in the kernels of almonds, apricots, apples,
cheries and the like, from which it is easily released by enzyme-catalysed hydrolysis. It has a pleasant vanilla-like odour, and is used as an
artificial source of that flavouring. It is also produced in great quantities in industry as an important ingredient in the synthesis of many organic
materials. Some of the routes used in industry are shown in the following figure
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig0.JPG">
All methods except the second require either high temperatures or pressures, or are very low yielding except in a continuous process. The second
method is essentially the one employed here, except the often repeated base hydrolysis is found to be very inefficient and is replaced by concentrated
acid hydrolysis.
While benzaldehyde is a gentle compound, the same can not be said of almost all intermediates and by-products obtained in its preparation. This
applies first and foremost to benzyl chloride, which had been prepared in the course of this work due to erroneous procedures reported and repeated in
preparotary texts. This compound was not isolated due to its tendency to polymerize, and more importantly, due to its physiological effects, which
have been found to be different to a certain extent, and in a sense worse, than had been reported - more on this will be said below. I present a
procedure to minimise the amount of this material isolated, that said, if benzyl chloride is the aim - for instance to be used in Grignard synthesis,
then a simple and efficient method for its preparation can easily be extracted from the text.
Benzal chloride, resulting from the di-chlorination of toluene has been found, surprisingly, to be much more stable to polymerisation, and does not
have the same immediate ill effects on the organism - presumably due to its much greater rate of hydrolysis. It had been prepared in good yield, and
identified by its IR spectrum and density. A little known but surprisingly effective procedure is demonstrated for hydrolyzing benzal chloride to
benzaldehyde - in almost 100% yield. Benzoic acid, a common preservative E210, is produced as a by-product in this procedure from the benzal
trichloride impurity and by air-oxidation of benzaldehyde - it separates as fine crystals in the purification step.
Reaction Scheme
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig1.gif">
The top line of the reaction scheme represent products and conditions of the chlorination step. The second line represents the subsequent hydrolysis
phase of the chlorination products. The third line represent purification, with the formation of a benzaldehyde-HSO3 adduct, removal of the residual
benzyl chloride trapped in the adduct with ether, subsequent liberation of the aldehyde by shifting the HSO3 equilibrium with a weak 10% Na2CO3
solution (strongly alkaline solution must not be used as it will destroy the benzaldehyde via the Cannizzaro reaction), removal of the benzoic acid
impurity by the same Na2CO3 solution, and recovery of benzaldehyde by steam distillation and drying with anhydrous MgSO4.
Findings
The present report is perhaps the clearest example of why illustrated practical guides are needed - one can not rely fully on preparative texts such
as Vogel. Surprisingly for what is by professional standards a simple procedure, a number of important omissions and downright falsehoods exit in the
common literature regarding the present set of transformations, knowledge of which is essential for anyone interested in repeating the work. This
presumably occurs because the authors of books are often compilators who do not have direct first-hand knowledge of what they are describing.
Furthermore the literature tends to copy itself. Certain other literature has been found which is 'not entirely candid', to use the legalistic
euphemism.
Here are the main conclusions of this work.
A 71% benzaldehyde yield based on toluene for 2-step process - this product contains 5-10% benzyl chloride impurity as determined by IR, but is
suitable in many syntheses where benzaldehyde is an intermediate.
A 52% benzaldehyde yield based on toluene for 3-step process - this product is entirely free of any IR detectable impurity.
A 75% benzal chloride yield based on toluene - this product contains 5-10% benzyl chloride and benzyl trichloride impurities as determined by
IR.
No IR detectable traces of ring-chlorinated impurities were found with the present procedure: strong blue-light source, dry and metal-ion
free toluene - middle 80% of distilled fraction, 110C reflux.
The present method uses only cheap easily obtainable chemicals: toluene, TCCA, HCl (most of which can be easily recycled if just to maximise
yield), Na2CO3, and NaHSO3 (sold as the anhydrous salt metabisulphaite/metabisulphate Na2S2O5). The most expensive chemical used is diethyl ether -
however only a small amount is required if the purification step is performed. It might be possible to substitute with DCM - but the solubility
ratios of adduct:benzyl chloride in this have not been investigated.
A temperature vs time graph of blue-light/UV catalysed chlorination of refluxing toluene is presented and shows the optimum end-point
temperature for benzal chloride is 195-197C, and not 187C as is frequently reported (eg. Heinrich Wieland, Laboratory Methods of Organic Chemistry).
The latter produces a 60-70% (IR determined) benzyl chloride by-product which is exceedingly hard and hazardous to separate due to its very low
hydrolysis rate, and the same 1atm b.p. as benzaldehyde (179C). Moreover the weight end-point determination frequently repeated in the literature, of
40gms gain for 50gms initial toluene, is not only an impractical method, but also hopelessly wrong. It has been clearly calculated
post-facto by someone who hadnt done the experiment, rather than actually used as a useful experimental aid as this number is presented,
because allowing for conversion to simple integer ratios, 50:90 corresponds to the stoichimetric weight gain 92:161 of toluene:benzal chloride. In
reality, due to polymerization of the benzyl chloride accelerated by UV light, and different chlorinated toluene derivatives present simultaneously,
at 187C the weight gain is only about 25:33. Even at the optimum end-point of 196C which I establish the weight gain is 184:291 - below the
stoichiometric amount reported, due to some of the toluene being converted to useless polymer at the benzyl chloride stage. Similar ratios were
observed for different batch sizes and light sources, from Hg vapour to halogen lamps.
The maximum chlorine flow rate for a 1L round flask reaction vessel and with delivery of chlorine sub-surface to the boiling toluene by a
simple tube was found to be 400ml Cl2/min corresponding to an adjustment of 1 drop HCl / 2sec in the chlorine generator of http://www.sciencemadness.org/talk/viewthread.php?tid=9713. This rate corresponds to the delivery rate at which no chlorine is evidenced in the
exhaust gases as determined by bubbling in conc. NaCl solution, which completely absrorbs HCl, but in which Cl2 is insoluble (see pictures below).
The corresponding reaction rate is about 0.5 mole benzal chloride per hour of chlorination. The limiting factor is attributed to reaction geometry
rather than intensity of blue-near UV light used because the latter was clearly present in excess, being provided by a 1KW Hg high-pressure vapour
lamp (see pictures below), which generates almost 60% of its energy in the interval between 350-488nm corresponding to the wavelength range where
toluene does not absorb significantly on the one hand, and there is enough energy in the light quanta to dissociate the chlorine on the other.
The maximum chlorine flow rate under the same conditions but with the weaker source of 2 50W halogen lamps with reflector mirrors and removed
front glass plates (see pictures below) was found to be 130ml Cl2/min, which translates to 1 drop HCl / 6 sec for the generator, or 0.16 mole benzal
chloride per hour. These lamps are a factor of 6 less efficient in the requisite wavlength range over the Hg light source, further proving that the
reactor geometry is the limiting factor with Hg lamps, however light intensity is the limiting factor with 100W halogen lamps. Attempts at enhancing
the reaction rate with a 500W diffuse reflector halogen light did not produce detectable improvements in the reaction rate over the 2 50W bulbs due to
the wide beam-widths of such lamps and lower quality of their reflectors - it did however generate far more stray heat.
It is found that no wash-bottles or H2SO4 bottles are needed to dry the chlorine, a single CaCl2/CaSO4 U-tube is used, as per the Cl2
generator in http://www.sciencemadness.org/talk/viewthread.php?tid=9713.
An attempt to use benzoyl peroxide as a free-radical initiator with gaseous chlorination showed it to be much inferior to even the halogen
lamps.
Alkaline-catalyzed hydrolysis (Na2CO3 or CaCO3) of benzal chloride often reported in the literature (see Wieland) is slow (about 6hrs) and only
about 50% efficient at that for conversion to benzaldehyde as evidenced by IR spectra.
Alkaline-catalysed hydrolysis (Na2CO3 with final stage NaOH addition) of benzyl chloride to benzyl alcohol with a 98% yield, US patent
2221882, is a fanciful invention - no traces of benzyl alcohol where evident in IR spectra of the reactor contents even after 6 hrs - this compound is
exceedingly difficult to hydrolyze. A benzaldehyde content was however found in the final product due to the much readier hydrolysis of the benzal
chloride impurity in weakly alkaline solution.
Strong-acid catalysed hydrolysis (25% HCl, 10-fold H2O excess) of benzal chloride to benzaldehyde reporting 94% yield, US patent 4229379, was
found to be absolutely true (restoring some credibility in patents). The bubbling of inert gas (nitrogen/CO2) reported in the patent was however
found to be counter-productive due to the greatly increased volatilisation of the benzaldehyde product and its loss during reflux. It is found that
the steady flux of HCl evolved during hydrolysis, and with the exhaust vented in a permiably stoppered (to prevent convection) bottle containing H2O
absorbent, is adequate to maintain high yield. As usual a conc. salt solution is used to determine end-point of reaction (by cessation of NaCl
precipitation and heat evolution).
The hydrolysis rate for the side-chain chlorides goes in the sequence: benzyl chloride < benzal chloride < benzyl trichloride. Benzyl
chloride goes essentially unchanged through the hydrolysis stage (while the latter two disappear completely from the IR traces of the products).
Hence it is advantageous to go past the benzal chloride maximum in the chlorination stage (see pictures below) if the level of benzyl chloride
impurity in the benzaldehyde is to be minimised without resorting to the 3rd (lossy) purification stage.
Bisulphite-adduct purification of benzaldehyde with optimum proportion of reagents and optimum procedure, which are hard to find in the
litertaure, are presented. The quantities are non-stoichiometric, since the adduct equilibrium is sensitive, and excess bisulphite is needed for
completion of reaction. However the excess must be moderated to prevent loss of product due to its dissolution in the water when the aldehyde is
liberated.
Strong physiological properties were only evidenced by the benzyl chloride. Contrary to everything stated in the literature it is not by
itself lacrimatory in concentrations built up while pouring between bottles etc. I never felt anything in my eyes while doing the chlorinations or
hydrolysis reactions. It does however have a most peculiar effect, at least on me. After breathing air containing benzyl chloride a peculiar
chemical taste develops in the mouth which can take several hours to disappear after reverting to fresh air. The effect is evident even without ever
smelling the benzyl chloride directly - its an indication that sub-threshold levels of benzyl chloride have been inhaled. If one does happen to
detect the odor of the latter (which is typical of a light chlorinated hydrocarbon), the taste appears after a few seconds and takes longer to wear
off. I found this effect so unpleasant that I would rate benzyl chloride as the most unpleasant chemical I have worked with in my many years of
scientific work. With the chlorination taking place in a fume hood and wearing a gas mask the taste effect still developped after spending 20 mins or
so in the lab, and without ever smelling anything directly. It was greatest when transfering liquids between the chlorination and hydrolysis stages -
especially in the first run where I used the faulty 187C end point from the preparative text (Wieland), producing a 60-70% benzyl chloride impurity.
I subsequently performed chlorination to go rapidly past the singly chlorinated phase so one is exposed mainly to benzal chloride, which does not have
this effect (presumably because its much more readily hydrolysed in the olefactory system).
A touch of benzyl chloride on the skin (I touched a condenser on which a tiny bit precipitated) does not lead to stinging - but one can smell
its odour at that spot for many hours despite washing. On standing, initially clear solutions of benzyl chloride become yellow, then brown due to
polymerisation. It attacks most plastics (goes through caps in bottles) and must be stored in glass.
Benzyl chloride became a totally different animal when as a 5-10% impurity in the benzaldehyde adduct it was mixed with ether in the final,
purification, stage. Initially while washing the adduct in a hood I felt an increasing stinging action at one spot on the palm of my hand, light
stinging in my eyes, and a bitter taste in my mouth. I was wearing latex gloves, observation showed no pinholes, while taking them off and smelling
the hand gave no hint of either ether or chlorinated toluenes. I left the area, flushed the hand and after a while the effect passed, there was no
peculiar taste hence I was at a loss to explain what happened. The reason became apparent the next day. After drying the washed adduct at 50C in an
oven for 1hr, I opened the oven (still wearing gas mask) and was overcome by an intense stinging in the eyes. After I left the lab the tearing
subsided I washed my eyes and took a shower, but the stinging continued for a couple of hours. The eyes did not redden, and there were no after
effects. The explanation to all this I believe is that ether acts as a permiator - benzyl chloride molecules being non-polar get attched to it, and
being somewhat soluble in water, it helps them pass latex gloves, the polar skin layer and enter the living cells, making the benzyl chloride more
'available' and thus potent. I have not read this effect anywhere - I believe it should have been mentioned by any serious work involving mixtures of
benzyl chloride with solvents such as ether - as a precaution, so people would not have to find out the way I did.
Theory
Chlorination - stage 1
Chlorination of the toluene side chain proceeds by a free radical mechanism, generating one mole of HCl for every mole of chlorine substituent -
hence twice as many atomic chlorine moles are required to perform a chlorination as actually get attached to the benzyl radical.
C6H5-CH3 + Cl-Cl -> C6H5-CH3 + Cl. + Cl. -> C6H5-CH2Cl + HCl
Chlorination by electrophilic substitution on the benzene ring, an unwanted reaction, is a much faster reaction, but it requires the presence of Lewis
acid catalysts such as Fe3+
C6H5-CH3 + Cl-Cl + FeCl3 -> C6H5-CH3 + Cl+ FeCl4- -> C6H4Cl-CH3 + HCl + FeCl3
It is essential that this reaction is hindered, and it is found that it can be essentially eliminated by the following four steps:
Strong light with wavelength < 488nm needed for Cl2 dissociation, but > 350nm at which toluene is opaque
Reflux at T > 110C, this slows ring substitution much more than it does side chain chlorination. This is most likely due to the fact that
ring chlorination must occur in the liquid phase where the Lewis acid is present, while the solubility of chlorine in toluene drops drastically with
temperature as shown by the graph below.
Absence of metal ions, this is achieved by distilling the toluene
No H2O in reaction, achieved by discarding the first 10% of the toluene distillate. Toluene and water do not mix, hence their mixture boils
at T < 100C and contains a constant fraction of water, until all water is removed.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig2.GIF">
The side chain chlorination now proceeds in stages of singly/doubly/triply chlorinated side chain. The stages overlap, so all three products are
present at any given time, but at widely different levels depending on the level of chlorination. The variation of composition with amount of
chlorine added is shown in the following graph (note that the amount of chlorine added assumes no polymerization of benzyl chloride, and 100%
chlorination efficiency, so it doesnt correspond to the amount of Cl2 actually passed through the toluene).
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig3.JPG">
One can see that the maximum amount of benzyl/benzal chloride which can be attained is about 70%, with the other constituents roughly evenly split,
this is in accord with concentrations measured here from IR graphs.
Reaction rate is also strongly dependent on the dissociation of the chlorine molecules, which as can be seen is part of the two-step process of
side-chain chlorination. This will be the rate determining step at low illumination (as seen here with halogen lights). The Cl-Cl bond is about
245kJ/mol, and the light has to be sufficiently energetic to break this bond. At the atomic level light is bundled, and each molecule can receive
energy only by colliding with this bundle - a photon. The amount of energy in the bundle depends on the wavelength, the Planck equation E = h c/lamba
= 245kJ/mol * Avogadros number. Putting in the numbers we get lambda = 488 nm. This is a resonance process so the amount of dissociation peaks
around 488nm. Toluene while clear in the visible becomes opaque to UV light, hence the most favourable light source has its energy centred at around
488nm, i.e. in the blue. The following figure shows the variation of the three reaction constants for side-chain chlorination with light wavelength
at reflux temperature and room temperature is as expected.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig4.GIF">
The figure below shows the effectiveness of various light sources at generating energy in the required range of 350-488nm. In the present case we are
interested in comparing halogen and mercury vapour lights as these are the two sources used here. There are two types of Hg vapour lamps - low
pressure (left figure) and the high pressure used here (right figure) and they produce slightly different spectra. The basic spectrum is that of the
low pressure lamp and consists of a set of narrow emission peaks - as is the case for all atomic emission spectra. In the high-pressure lamp the
density of ionised Hg atoms is much greater and so basic Hg emission lines are collision broadened - hence the peaks on the right correspond to those
on the left - but are much more filled-in in the high pressure spectrum. The broadening is only a few tens of nanometers and so the energy emitted in
a wide wavelength range is essentially unchanged by going from low to high pressure Hg.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig5.GIF">
Hence basing analysis solely on the graph on the left, and calculating graphically the ratio (energy emitted 350-488nm)/(energy emitted over all
wavelengths) we get about 0.6 for the Hg lamp and 0.1 for the halogen lamp. Assuming radiation efficiency of both lamps of similar power is about the
same (i.e. heat loss by conduction as opposed to radiation is similar in both cases) the former is 6 times more efficient than the latter in puting
out energy capable of exciting chlorine atoms. For the present experiment the 1000W Hg lamp puts out 60 times more useful energy than the 100W
halogen bulbs (which however are better focused). Since the observed increase in reaction rate is only a factor of 3 this provides support that in
the former case reaction rate is limited by the radical reaction rate, rather than by radical rate of formation.
Hydrolysis - stage 2
Hydrolysis proceeds by the following reaction
C6H5-CHCl2 + H2O -> C6H5-CHClOH + Cl- -> C6H5-CHO + 2Cl-
and as is often the case for hydrolysis reactions is catalysed by either acid or base. Base catalysis is probably initiated by OH- nucleophilic
attack, while the acid catalyzed reaction, which has been found here to be much more efficient most likely involves the formation of a carbonium ion.
Purification - stage 3
The main impurities contained in the benzaldehyde generated by the previous two steps is benzyl chloride, which does not hydrolize to any great extent
in step 2, it is fully soluble in benzaldehyde and hence is present to the full extent of its formation in step 1, as shown by IR graphs. Since
benzyl trichloride is almost all converted to benzoic acid in step 2, which is insoluble in benzaldehyde, it is advantageous to carry out chlorination
beyond the benzal chloride maximum if one wants to avoid the purification step. In order to separate the benzaldehyde it is converted from a polar to
an ionic compound using the propensity of aldehydes to form bisulphite adducts. Sodium metabisulphite dissolved in water generates the bisulphite
anion
Na2S2O5 + H2O -> 2Na+ 2HSO3-
which form the water soluble ionic adduct with benzaldehyde
C5H5-CHO + NaHSO3 -> C6H5-CH(OH)(SO3).
This reaction is a delicate equilibrium and an excess of bisulphite is needed to run it to completion and avoid loss of benzaldehyde in the next
phase. The adduct is quite soluble in water - hence it must not be washed with water - but sparcely soluble in conc. NaHSO3- solution - hence excess
bisulphite must be used so a substantial portion of the adduct is precipitated. However since the liquid portion contains almost no benzyl chloride
impurity it is inessential that all adduct is precipitated, it is however essential that no benzaldehyde is left in free form, as evident by the
smell. The non-polar benzyl chloride is now separated from the ionically bound benzaldehyde by washing in diethyl ether, and the free aldehyde
regenerated by using a buffered alkaline solution in the form of 10% Na2CO3, by consuming the bisulphite in the following reaction
HSO3- + OH- -> H2O + SO3--
and so shifting the adduct equilibrium to the left. This equilibrium can also be shifted left by acids converting the bisulphite to SO2, but this has
been found to be much less efficient. Use of strong bases is unfortunately precluded by the tendency of benzaldehyde to undergo Cannizzaro
disporportionation
2C6H5-CHO + H2O -> C6H5-COOH + C6H5-CH2OH
Benzoic acid formed by benzyl trichloride hydrolysis, and by air oxidation of benzaldehyde
2C6H5-CHO + O2 -> 2C6H5-COOH
a process which occurs with remarkable rapidity whenever benzaldehyde is exposed to the air, such as shaking with Na2CO3 solution which inevitably
lowers the pressure inside the reaction vessel (and hence the purification operation due to its many phases involves a loss of 10-15% of
benzaldehyde), is removed concomitently with the liberation of the aldehyde from the adduct during the reaction with Na2CO3 by the formation of its
sodium salt which is 200 times more soluble than benzoic acid.
Estimation of fraction of compounds in a mixture by the IR method.
This involves the use of Beers law. The transmission coefficient of light of a definite wavelength through a thickness l of compound of
concentration c and extinction coefficient e(lambda) is
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_form1.bmp">.
Now compounds such as substituted benzyls have common absorption lines corresponding to various vibration modes of ring C-C and C-H bonds. Thus ring
C-H stretching vibrations are shown by all aromatics in the 3030 cm-1 region, while aromatic C-C stretching modes appear at 1450 cm-1. If T0 and
T0ref are observed and reference transmission intensities of such a common peak, while T1 and T1ref are the respective transmission intensities of a
peak differentiating compound 1, then the relative concentration of compound 1 c1 is given by
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_form3.bmp">
In the second factor the extinction coefficient for the common absorption peak and the concentration both cancel, giving us the necessary correction
factor between the reference spectrum and our cell.
Method
All operations described below were carried out in a fume hood, wearing a gas mask, latex gloves, and with a
draught established in the laboratory.
Ambient temperature in the lab varied in the range 30C - 40C.
Step 1 - Halogenation
Production of 70% benzyl chloride - 30% benzal chloride by procedure of Wieland, Laboratory Methods of Organic Chemistry
300ml of commerical toluene where placed in the distillation setup below and slowly distilled, with the first 35ml discarded, and subsequent 210ml
collected. I find it best to start the distillation with no water in the condenser, allowing its inner tube to heat up and be completely coated with
toluene. This purges H2O, present either if the condenser was not perfectly dry, or condensed from the toluene, and sticking preferentially to the
glass walls, while the stream of toluene below bypasses it.
The boiling point rose somewhat towards the end of the distilation and the liquid being distilled yellowed somewhat showing that xylenes and iron
impurities were definitely present in the commerical sample. The 210ml gathered corresponded to 2 moles of toluene, and weighed 184gms.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig6.JPG">
Next the chlorine generator was assembled, attached to a 5mm tube inserted in a 3-neck flask containing the toluene, and reaching to the bottom as
shown in the figure below. Exhaust from the reflux condenser was passed through a 40cm air condenser, and then down some polyethylene tubing into a
glass wool plugged bottle containing a stoichimetric amount of NaOH solution to absorb the HCl formed. A bottle with conc. NaCl solution was also at
the ready to determine whether any unreacted chlorine was passing the reactor. Two 50W halogen bulbs were aimed at the chlorine delivery tube, the
toluene was refluxed, the chlorine generator and U-tube purged of oxygen, and a drip rate of HCl established which led to no substantial bubbling in
the NaCl bottle - subsequently the tube exhaust was led into the NaOH bottle - lying above the surface of the solution. The rate of HCl feed so found
was 1 drop / 6 secs, which roughly corresponds to 130ml Cl2 / min. The reaction was continued till the temperature of the boiling mixture reached
187C - as required by Wieland - this took roughly 8 hours, although operator action was only required to replenish HCl in the pressure equalised
dropping funnel. The flask contents substantially darkened during the course of the reaction to a dark yellow colour. Distillation of the mixture
produced a transparent product, which however rapidly yellowed on standing. This was found later to be due to polymerization of the benzyl chloride
component. The reaction products were not weighed at this stage, however the weight gain of a smaller 25gms batch chlorinated in a similar fashion
was found to be only 8 grams - far short of the 50:90 gain stated in the text. The density of the liquid measured 1.1, consistent with benzyl
chloride, as opposed to 1.25 for benzal chloride.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig7.JPG">
An IR spectrum of the product is shown below, together with reference spectra for benzyl and benzal chloride. It is clearly seen that we have a
mixture of these two components, with benzyl chloride component being about 70% by Beers law, using 700 cm-1 as the common peak and the peaks at 560
cm-1 and 588 cm-1 as the benzyl chloride and benzal chloride individuating peaks. There are no other peaks (save for CO2) and none corresponding to
ring chlorinated products.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig8.GIF">
Since benzyl chloride was the major product any hope of hydrolysis to benzaldehyde was lost - however patent US2221882 reports conversion of the
former to benzyl alcohol in up to 98% yield by an intricate procedure involving reflux with a stoichiometric amount of the weak base Na2CO3, followed
by the addition of the stronger base NaOH 'to complete the reaction', all in a current of CO2 gas. This hydrolysis was carried out and IR spectra
taken at regular intervals. After six hours the following spectrum was obtained.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig9.GIF">
Clearly many of the peaks are those of the reagents - indicating a poor yield, however new product peaks are clearly identifiable, the most obvious
one being at 1701 cm-1 which is clearly due to a carbonyl bond, but others at: 745, 832, 1025, 1205, 1311, 1388, 1597, 1654, 2736, and 2818 cm-1.
None of these correspond to the putative product, benzyl alcohol, of which there is not a trace, and I wrecked my brain for a long time comparing to
the spectra of benzoic acid, chlorinate phenols and the like, until I got an unexpected perfect match with benzaldehyde. The minor benzal chloride
component got hydrolysed to the aldehyde, as evidenced by the clear diminution of its individuating peak at 586 cm-1 compared to the same peak in the
reagent mixture. The benzyl chloride peak is almost unchanged. The result was surprising considering Na2CO3 and especially NaOH is considered too
strong a base for benzaldehyde to survive, which is why the more complicated hydrolysis with CaCO3 is normally recommended.
Since the benzyl chloride remained unreacted throughout the procedure, it was redistilled from the reaction mixture, resulting in a clear solution.
The solution was bottled, however it darkened in the course of several days due to polymerization, and attacked the polyethylene capped bottle.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig10.JPG">
Synthesis of 75% Benzal chloride - the optimal mixture
It is quite apparent the the end point determination in Wieland is quite mistaken. The reagent was clearly never weighed to give the 50:40
toluene:weight gain, while the end-point temperature of 187C is much too-low and gives a mixture of singly and doubly chlorinated toluenes in the
unfavourable ratio of about 70:30. Repetition of the chlorination using the vapour reflux temperature instead of the liquid temperature as the end
point criterion, led to the same results, with the difference that keeping track of the reflux temperature is much more finnicky as the point of
condensation in the reflux collumn keeps descending as the temperature is rising, so that the thermometer has to be constantly lowered.
The literature shows that benzyl and benzal chloride boil at 179C and 205C respectively. Expecting the maximum concetration of the latter to be
achieved at an end point of 205C is incorrect since as noted in the theory section all three chlorinated derivatives are present simultaneously, and
at the optimal point for benzal chloride its ratio to benzyl chloride is expected to be about 70:15, leading roughly to a b.p. for the mixture of 0.85
* 205 + 0.15 * 179 = 201C, while the presence of polymerisation impurities is expected to lower the temperature by about 5C or so, giving an estimated
end point temperature of 196C.
Since to achieve the desired objective the chlorination needs to be continued about 1.5 times further than in the previous section 100W halogen bulbs
were deemed too slow for a 2 mole experiment (it would take about 12hrs) and a 1kW high pressure mercury vapour bulb with 60% radiation of the correct
wavelength (see theory section) was used instead. This has two added advantages:
Chlorination goes rapidly past the benzyl chloride point where the liquid rapidly darkens due to polymerisation, lowering the yield.
Exposure to benzyl chloride is minimised, benzal chloride was found not to have the adverse physiological properties of the former.
The new setup is shown below, the main difference being the presence of an Hg vapour lamp (operated through a ballast) as close as possible to the
reaction flask to maximise light intensity in the reaction area.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig11.JPG">
There are two points with this setup
The Hg lamp must be close, but not too close else hot spots form in the area of the flask adjacent to the bulb, where toluene is pyrolised by
the high temperature - this was observed in one run, but the charred products amounted to less than 1% of yield loss. However in order to avoid flask
cleaning problems it is best to locate the lamp no closer than 8cm, however if forced air cooling is used separations down to 2cm can be used.
The ground glass junction must be protected from the heat of the Hg lamp by glass-wool lagging, else matter will carbonize in it, and it will
freeze.
The chlorine generator was purged as before, and Cl2 introduced into the refluxing toluene. The Hg lamp was ignited - at this stage dark glasses and
an aluminium screen to shield from direct rays was employed in addition to the safety measured mentioned above - the exhaust tube was attached to a
pipette and dipped into the flask containing the conc. NaCl solution (seen in photo above) and the drip rate adjusted to a maximum consistent with no
bubbles being evident from the end of the pipette. This coincided with 1 drop HCl / 2 sec corresponding to about 400ml Cl2/min, or three times faster
than with the halogen bulb. The appearance of the exhaust from the pipette at this stage is shown below, it spits streaks of NaCl due to the
disolving HCl gas precipitating NaCl out of its solution. This effect of HCl gas on NaCl solutions was mentioned by Stefan http://www.sciencemadness.org/talk/viewthread.php?tid=9728&a....
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig12.JPG">
Having so adjusted the optimum rate of Cl2 delivery the exhaust must not be left to bubble through the NaCl solution as the salt formed will
eventually block any tube. The tube is now placed in a different bottle containing NaOH solution, without touching the solution, and lightly plugged
with glass wool, to contain the HCl until it is absorbed, and to prevent the entry of oxygen. The tube can be periodically replaced in the NaCl
solution and the drip rate adjusted, since the Cl2 flux will slowly decrease due decreasing drip rate with lowering head pressure in the dropping
funnel. Alternatively one can accept the variable reaction rate and replace the HCl in the funnel every 50mins or so - in that case the entire
reaction time until a reflux temperature of 196C is reached is about 6hrs or so, and uses up about 1L of 16% HCl. The figure below shows how the
reflux temperature rises with time which correlates with the reaction rate, with the knee in the graph being the point where I returned to the lab and
refilled the dropping funnel with HCl.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig13.GIF">
With halogen bulbs when a temperature of 140C degrees or so is reached the liquid yellows considerably see figure below, while if an Hg lamp is used
it becomes quite dark yellow. The reason for this, as mentioned above, is benzyl chloride polymerisation which, as expected, is accelerated by UV.
The idea is to chlorinate quickly at this stage to minimize loss of product at the benzyl chloride stage.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig14.JPG">
Quite a bit of heat is released in the NaOH absorber bottle during the reaction and it becomes hot to the touch. If a sub-stoichiometric amount of
NaOH is used the solution becomes acidic towards the end due to HCl dissolution re-forming hydrochloric acid (which still heats the solution
substantially) and precipitates large chunks of NaCl previously formed by neutralization of NaOH out of solution (see figure below). After the
end-point is reached, the solution is let to cool somewhat, and when safe, the reflux condenser is changed for downward distillation, a boiling chip
is dropped in the flask and distillation - which produces a clear viscous liquid (see figure below) is commenced. The distilation can be continued
until the liquid in the distilling flask becomes quite dark and viscous - at which point it must be stopped to avoid solidification.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig15.JPG">
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig16.JPG">
The distillate obtained from 184gms of toluene was found to weigh 265gms, and occupy 214mls (see figure above). Its density is thus 1.24, compared to
1.25 quoted for benzyl chloride. The very viscous remnants in the flask were found to weigh 26gms. Hence the correct weight gain,
determined experimentally rather than calculated, is 184:291. The yield assuming benzal chloride is 82% based on toluene.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig17.JPG">
This distillate, as opposed to the benzyl chloride in the previous procedure was found to be quite stable, it did not visibly darken over the course
of several days, its odour - which was sweetish typical of more heavily chlorinated hydrocarbons - did not produce the physiological effects of benzyl
chloride.
This time the IR spectrum shows a clear predominance of benzal chloride (individuating peak at 586) as compared to benzyl chloride (individuating peak
at 561). Note that there is some artificial difference in the appearance of the spectra due to the reference using x2 stretching in the fingerprint
500-2000 cm-1 region, while I do not use that stretching. Beers law gives an approximate content of 5% for the latter (using 0.96 transmission, with
correction for background, at 561 for benzyl chloride compared to 0.3 for reference below). The peak at 629 individuates benzyl trichloride, and
being more substantial than the benzyl chloride peak shows we have achieved our aim of going past the dichlorination maximum slightly in order to
minimize the amount of monochloride present. We shall be more than repaid for the slight loss of yield by the ease with which the trichloride is
purified.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig18.GIF">
Step 2 - Hydrolysis
Reflux with CaCO3 - unsuccessful
Most texts describe this as being carried out in a weakly basic environment for catalysis while avoiding Cannizzaro disproportionation of the
benzaldehyde and loss of yield. The procedure (in the now tainted) Wieland, which is frequently repeated elsewhere (as well as here) is to use reflux
with a heavy excess of H2O with freshly precipitated CaCO3 under a soft current of CO2 gas to remove the HCl formed and protect the benzaldehyde from
O2 oxidation. This was arranged as shown in the figure below
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig19.JPG">
This 'hydrolysis' was continued for 6 hours with IR readings taken every two hours. There was a very disappointing yield which did not approach even
30% after 6 hours, despite the mixture being boiled quite violently.
Reflux with Na2CO3 - inefficient
Next the use of the stronger base Na2CO3 was attempted - this is soluble so it was hoped that better mixing would be effected between the two phases,
as can be seen in the figure below the oily benzal chloride layer became cloudy while vigorous boiling (with use of boiling chips) ensured good
mixing.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig20.JPG">
However after 6hrs of such reflux the IR trace was still disappointing. A clear sign of benzaldehyde can be seen by the appearance of a new peak at
1701cm-1, however all the reagent peaks such as the benzal chloride peak at 586cm-1 are still evident. Beers law indicates less than 50% conversion,
which is very inefficient. The reaction really is not worth continuing further as with O2 oxidation being a competing process we reach a point of
diminishing returns.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig21.GIF">
Reflux with conc. HCl - near 100% yield
This method was taken from patent 4229379. While I have learned to take patents with a grain of salt - and this one had all the hall-marks of
'fairlytale' patents, such as inert gas cover, extremely high yields where previously low yields were reported, extreme accuracy in numbers - I was
astounded to find that this patent is absolutely true. This is the more so, since thermodynamically basic catalysis seemed favoured, as in that case,
with HCl generated the OH- concentration shifted the equilibrium to the right. With acid catalysed hydrolysis there is no such enhancement. Indeed
the high HCl catalyst concentration, with HCl being also the product formed, should shift equilibrium to the left. Nevertheless, my yield of
benzaldehyde was 96% based on benzal chloride. I initially did a very small run with about 90mins worth of reflux, as I didnt want to lose all my
benzal chloride to what I feared was a bogus patent - this is what I got
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig21.JPG">
The initially heavy clear oily layer (density benzal chloride 1.24 as compared to 1.12 for conc HCl) turned yellow and floated to the top (density
benzaldehyde 1.05) of the acid layer. There was suddenly a strong smell of benzaldehyde in the air.
I did find some changes to the patent were necessary to optimize yields in the laboratory. This included
Dispensing with the inert gas - I found the HCl flux liberated in the hydrolysis was quite sufficient to prevent the influx of any oxygen if
the exhaust gases were absorbed in a bottle semi-permiably stoppered with glass-wool. Moreover a flux of CO2 led to a loss of product since the
benzaldehyde is very volatile in steam. This leads to the second point.
A single condenser, whether of the Leibig or West variety proved inefficient in prevention of benzaldehyde efflux with the escaping HCl. The
HCl parts from water only very reluctantly, while the benzaldehyde clings to the water, and tends to get carried along into the absorber bottle, even
through a fully cooled condenser. I therefore introduced a second condensation stage, consisting of a straight Hempel tube half full of rings - this
gave a large surface area for water/benzaldhyde droplets to condense and equilibrate with the liquid, and prevent loss into the exhaust. At the end
of the reflux the collumn is of-course flushed.
It is very important to know when the hydrolysis is essentially complete, without relying on IR spectra, as continuing further refluxing the
benzaldehyde - its most active form - without HCl efflux leads to exposure to air and immediate loss of yield. The floating of the oily layer to the
top is part of the indication. However full hydrolysis is not yet achieved at that point. This is best gauged by venting the exhuast HCl at the end
into a conc. HCl solution instead of NaOH (or what is equivalent, venting into a sub-stoichiometric NaOH solution). Crystals of NaCl form on the
surface of conc. NaCl solutions while HCl is being evolved, while the solution itself is almost hot to the touch. Cessation of these effects
indicated that no more hydrolysis is occuring.
Here is a picture of the final setup I adopted
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig22.JPG">
I can not emphasise enough how much overhead stirring, such as shown in the picture, is essential here. Initially the high b.p. layer is at the
bottom - where the mantle is hottest. With such an arrangement this layer gets heated to a much higher temperature than the 110 b.p. of the HCl layer
above. When bubbles of the organic layer so superheated rise past the nonpolar-polar interface, they superheat the HCl leading to violent erruptions
into the reflux collumn. Eventually the entire reflux can get ejected through the reflux collumn. The mixing must be quite vigorous - about 3
turns/sec.
I used the x10 H2O excess suggested in the patent - using more will lead to loss of aldehyde due to its dissolution in the H2O (0.6%). Hence 214ml of
the benzyl/benzal chloride/trichloride was placed in the 1L 3-neck flask above, 405 ml 32% HCl was added, followed by 135ml of distilled H2O, which
leads to about a x10 excess of H2O in 25% HCl. The contents were refluxed with the reflux temperature at 106C - this is the same temperature
mentioned in the patent, but what is important is that it self adjusts- there is no need for a fixed bath. The reflux took 6hrs at which stage HCl
evolution ceased, as evidenced by cooling of the absorber bottle and no more NaCl crystals forming on the surface. Here are the results, the picture
on the left as observed just before cessation of mixing, and on the right upon cooling
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig23.JPG">
The mixture was now quickly steam/HCl-distilled to effect separation of the benzaldehyde and minimize oxidation.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig24.JPG">
As expected for non-mixable components, all the benzaldehyde distilled at a fixed temperature of 106C (below the b.p. of both aqueous component and
benzaldehyde), when this was through the temperature rose somewhat to 110C and white needles of benzoic acid started forming in the condenser, it has
quite a solubility differential in H2O, it is fairly soluble in hot water ~ 6.9gms, but quite insoluble in cold. At this point distillation was
ceased.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig25.JPG">
The liquid left in the flask consisted of a tarry layer due to benzyl chloride polymerisation during the hydrolysis and an upper HCl layer saturated
with benzoic acid (left picture below). On cooling the benzoic acid crystalysed on the surface in fine white needles, while sending a beautiful
magical 'snow storm' of shining benzoic acid crystals, to which unfortunately the picture doesnt do justice, into the air in the flask.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig26.JPG">
The benzaldehyde layer was separated in a separation funnel from the heavier - due to HCl - aqueous layer in a separation funnel, and
yielded 151 gms. This gives a yield, based on 184gms of toluene of 71%.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig27.JPG">
The IR spectrum of this benzaldehyde is shown below
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig28.JPG">
All the di- and tri- chlorinated toluenes have hydrolysed and the only evident impurity is benzyl chloride evidenced by the peaks at 561 and 1266cm-1.
Beers law give their content at 5-10%. If this small impurity can be tollerated this is the end of the synthesis - apart from drying the
benzaldehyde with some anhydrous MgSO4.
Step 3 - Purification
The aim of this procedure is to eliminate the benzyl chloride impurity as well as any traces of benzoic acid and water from the benzaldehyde formed in
step 2. Purification by distillation isnt appropriate as both compounds have the same b.p. at atmospheric pressure, and they are both non-polar.
However the aldehydes reacts with bisulphite to form an ionic solid, whereupon the non-polar benzyl chloride can be separated by washing with a
non-polar solvent such as ether. Other low b.p. non-polar solvents such as DCM are probably also suitable.
This purification is a high loss procedure - about 15% of the benzaldehyde is lost, it is also long winded and hazardous, and so should only be
attempted if the impurity in step 2 really cant be tollerated (such as if you intend to drink the stuff). The loss mechanisms are as follows:
Adduct formation is a delicate equilibrium requiring excess HSO3-, some benzaldehyde remains unconverted and is lost with the ether wash.
The ether dissolves some of the adduct and it is also lost in the wash
Heavy excess of water must be used during the alkaline liberation of the benzaldehyde (to prevent Cannizzaro) and the steam distillation. In
all about 1L H2O is used to purify about 75gms benzaldehyde. This leads to dissolution of 3-5 grams in the water i.e. a loss of about 4-6%.
The many operations, especially the steam distillation require exposure of the benzaldehyde to the air, leading to loss by air oxidation to
benzoic acid. Indeed 3ml benzaldehyde in a non-tightly closed bottle was completely lost to benzoic acid in a matter of 3-5hrs.
There is a requirement to dry the adduct of ether, this is done at 50C in an oven. This leads to loss of adduct by its breakup, and this
equilibrium is further pushed to the right by air oxidation of benzaldehyde.
If all this doesnt put you off heres the procedure.
To a 72gm portion of the benzaldehyde from step 2, 86gms of Na2S2O5 were added dissolved in 190ml of H2O. After 5-10 minutes shaking, and 10 minutes
further standing heat evolution ceased, and a white adduct formed in a slight pool of excess bisulphite solution.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig29.JPG">
If one looks carefully immediately after shaking, small bubbles of oily benzyl chloride can be seen on the surface of the adduct (this colour is not
due to excess bisulphite as its concentration has substantially diminished due to absroption into the adduct. The contents of the bottle should no
longer smell of benzaldehyde, rather of benzyl chloride whose odour it has previously been masking.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig30.JPG">
The contents of the bottle are now carefully, with rotation, slid onto a filter funnel and filtered and patted down thoroughly at the pump. The
filtered solution is kept - the benzyl chloride is insoluble in it, and has mainly stuck to the adduct. Next the adduct is transferred to a bottle
and cooled in a fridge. 70ml of ether is then used to wash both the original bottle and the funnel, subsequently being transferred to the bottle with
the adduct, shaken, and left to stand. The contents of the bottle are then transferred to a filter and dried at the pump as much as possible. The
filtrate is discarded. The adduct still has some ether clinging to it (smell!), hence it is dried in an oven for several hours at 50C.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig31.JPG">
The adduct is then transferred to a 3-neck flask, and its aqueous filtrate, after being used to wash the funnel is added. Next 86gms of Na2CO3
dissolved in 700ml of H2O and heated to about 50C is added to the flask, stoppers inserted, and the flask vigorously shaken, until all the adduct
dissolves. It is then left to stand until a clear oily layer has formed at the top and the aqueous solution has become almost clear.
At this point if the Na2CO3 is pure it might be possible to conserve some benzaldehyde by separating the layers in a funnel. If the Na2CO3 is
slightly impure (such as mine) or the solution has'nt completely clarified, steam distillation is required. If that is considered a hassle, it is
possible to distill the benzaldehyde directly without the addition of steam, however some loss will occur due to the concentration of the alkaline
solution as the distillation proceeds. I used a common kettle to generate the steam - although an evacuated flask is better, since it introduces less
oxygen into the mixture. It is best to use a splash head, the 3-neck flask is heated almost to boiling point, and steam is introduced at such a rate
that the bubbling is vigorous but there is no foaming. A kettle run at 100% power generally generates too much steam, and I adjusted the power using
a triac controller to produce a 20% duty cycle.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig32.JPG">
This time the aldehyde gathers as the lower layer (since the salts which made the aqueous layer dense do not co-distill). The cloudiness of both
layers can vary here because the density differential at this stage is small and is a function of temperature, also the water is saturated with
benzaldehyde and vice versa. Clarification can generally be attained by heating the contents of the flask to 60C or so and letting cool. The layers
are separated on a separatory funnel, and the aqueous layer reused for any further purification runs, if they take place immediately. The
benzaldehyde is now dried with anhydrous MgSO4 which immediately clears it - if it was cloudy. The 72gm batch gave 52.5 gms of pure benzaldehyde at
the end of the procedure, corresponding to a 73% yield for the purification step, and a 52% yield based on the initial toluene. The benzaldehyde was
bottled in a tightly closed bottle and stored in a dark place.
The following graph shows that the benzyl chloride trace has been entirely elliminated and the benzaldehyde is essentially pure.
<IMG src="http://www.sciencemadness.org/scipics/Len1/Benz_fig33.GIF">
Health hazards - documented physiological properties of halogenated
toluenes
In the introduction I had already listed some of the unpleasant aspects of the chemicals in this synthesis which I had encountered. The question is
will the exposure to benzyl chloride here likely impact my lifespan. In answer to that question I briefly reserached the physiological properties of
halogenated toluenes.
Halogenated toluenes are considered probable carcinogens - this means there is no adequate statistics with humans that they initiate cancer, but there
is with animals. Here however is some fairly interesting human data, which would be conclusive if not for the small bin size
| Quote: | | A retrospective mortality study was reported on a cohort of 697 male workers who were exposed to benzyl chloride, benzoyl chloride and
benzotrichloride at a chlorination plant in Tennessee (Wong and Morgan, 1984). The length of employment at the plant ranged from 1 year to >35
years. Seven deaths from respiratory cancer were found in the total cohort compared with 2.84 deaths expected, based on U.S. mortality rates for
males. Five of these deaths occurred in workers employed for at least 15 years. This was significantly greater than the 1.32 deaths expected for this
subgroup. The results of this study were confounded by exposures to several chemicals and lack of data on smoking. |
One should note that no attempt has been made to differentiate benzal and benzyl chloride, due to the fact that these chemicals are almost always
encountered together, and the statistical significance for even the combined human data is already small.
Here is some more remarkable work on mice
| Quote: | | Fukuda et al. (1981) conducted two skin-painting studies on specific- pathogen-free ICR mice, using benzyl chloride dissolved in benzene. Benzene-
only controls were included for vehicle comparison. In the first study, no tumors were observed in 11 mice treated with 10 uL benzyl chloride 3
times/week for 4 weeks, followed by 2 times/week until termination at 40 weeks. In the second study, 2.3 uL benzyl chloride was diluted to a final
volume of 25 uL with benzene and applied to the skin of 7-week-old mice 2 times/week for 50 weeks. Two of 20 control animals developed lung adenomas,
while 5/20 treated mice developed tumors, including 2 lung adenomas and 3 skin carcinomas. Two of the skin carcinomas metastasized to the primary
lymphatic organs, liver, or kidneys. Although these tumor incidences are not statistically significantly greater than controls, the authors considered
benzyl chloride to be a weak carcinogen when applied topically. The short duration of the studies limited their sensitivity. |
Prior to having done this synthesis I would have concluded that the evident carcinogenicity of benzyl chloride only in a mixture with benzene is a
property of benzene, which is a known carcinogen. Now I conlcude otherwise - the carcinigen is most likely the benzyl chloride, the benzene
acts as the permiator, much as ether did in my undesired experiment.
Here however is some comforting data
| Quote: | Efforts to assess the potency of benzyl chloride as a carcinogen and skin tumor initiator provided predominantly negative results. Coombs (1982a)
applied 1.0 mg benzyl chloride in toluene to the backs of 40 T.O. (Swiss- Webster derived Theiler's Original) mice, followed by twice weekly
treatments of croton oil in toluene for 10 months. While 8/19 positive controls treated with 0.4 mg benzo[a]pyrene developed skin tumors, none (0/37)
of the benzyl chloride-treated mice did. In a second initiation-promotion test, Coombs (1982b) topically applied 10, 100, or 1000 ug benzyl chloride
in acetone, followed by twice weekly applications of the promotor 12-O-tetra-'3-decanoyl- phorbol-'3-acetate. At the end of 11 weeks, all of the
positive controls treated with (7,12-dimethylbenz[a]anthracene) had skin tumors, whereas at 6 months (approximately 12 weeks later), only 20% of the
mice treated with benzyl chloride showed similar changes. Ashby et al. (1982) topically treated groups of 20 Swiss mice with 100 ug benzyl chloride in
toluene twice weekly. After 7.5 months, none of the treated mice had skin tumors compared with 18/20 of the positive controls treated with
benzo[a]pyrene.
Druckrey et al. (1970) administered benzyl chloride in peanut oil via weekly subcutaneous injection to BD-strain rats for 51 weeks. Local sarcomas
were produced in 3/14 rats given 40 mg/kg/week and in 6/8 rats given 80 mg/kg/week. The average induction time was 500 days and metastases to the lung
occurred in the high-dose group only.
Groups of 20 strain A/H mice were injected intraperitoneally over a 24- week period with benzyl chloride in tricaprylin (total doses of 4.7, 11.8, or
15.8 mmol/kg). No differences in pulmonary adenoma formation between treated and vehicle control mice were observed (Poirier et al., 1975).
|
Quotes are taken from http://www.epa.gov/iris/subst/0393.htm.
Conclusion
A very efficient two-step method has been presented for the preparation of benzaldehyde with 5-10% benzyl chloride impurity. A purification stage
reduces the yield to 51% but eliminates all traces of the impurity from IR spectra. Indications from health studies are that the health risk posed to
those carrying out this procedure are not too high. Nevertheless all operations were conducted with: fume hood, 2-filter gas mask, draught in lab,
latex gloves, dark glasses and screen during Hg lamp operation.
[Edited on 21-4-2008 by len1]
|
|
|
Klute
International Hazard
Posts: 1378
Registered: 18-10-2006
Location: France
Member Is Offline
Mood: No Mood
|
|
Beautifull work!! Very complete and detailed, it a pleasure to read!
thumbs up
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Thank you.
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Very nicely done, Len.
Caught one misspelling/finger slip "cooling of the abrorber bottle ", unless abrorber is something only Oz has.
If you very little monochlorotoluene and benzyl alcohol, a wash with saturated NaHCO3 will remove the benzoic acid from the aldehyde (in ether or DMC)
nicely, this may give a clean enough product without the bother and cost of the bisulfite addition product.
When I did this, I did not irradiate the liquid in the pot. Instead I did that in a fractionation column loosely packed with largish pieces of broken
labware (I save everything). I used an elliptical reflector around the column and lamp. Monitored the temperatures at the bottom and top of the
column, regulated the temperature to keep the top no higher than the boiling point of benzal chloride and for most of the reaction not above that of
the monochoride. Chlorine in the bottom, HCl off the top after a full reflux condenser for the organics.
Hydrolysis will NaHCO3 in water, with some of it solid, previously swept with CO2, and added 1,2-dichloroethane as an organics solvent. Just stirred
and kept refluxing, with a slight positive pressure of CO2 and low light level; left it running for several days. All the workup was done as much as
possible in CO2 atmosphere.
I was not nearly as painstaking as you in record keeping, and IR spect was still outside of hobbyist range back then. Can't remember the yields, but
were decent if not quite as high as yours. The HCl hydrolysis sounds interesting, if I ever get a lab set up again I may try it.
|
|
|
12AX7
Post Harlot
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Nice but do you think you could make it less wordy? I really hate academic papers that are wordy for the sake of wordy, know what I mean?
Tim
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Not important, thank you, I will correct the error as Im still within the generous 24hr edit time.
It sounds like you had a more intricate and original setup - it would be great if you ever decided to do practical work again. I realised that much
of the reaction is probably taking place outside the liquid toluene at the bottom of the flask, and tried to illuminate as much of the toluene-Cl2
contact area. That said its clear that the quality of the mixing is the rate determining step with the Hg illumination and effort towards enhancing
that arent in vain as you mention. If the chlorination could be done in 1/2 hr instead of 4 that would be a major improvement. SO2Cl2 chlorination
according to Vogel is on that timescale - but for the amateur any time gained that way would be lost in producing it. My standard supplier doesnt
carry it - its un-Australian to stock any chemical that cant be used by the farmer.
NaHCO3 hydrolysis sounds interesting - I would indeed be worried with leaving the aldehyde in contact with Na2CO3 for that long. But I think that HCl
procedure from the patent is a winner - hydrolysis is so rapid you dont even need protector gas, the HCl evolution does that for you.
Tim I also hate lofty wordiness for the sake of appearing erudite. But I believe most people i deal with are far more snobish that way than I am.
Ive been in academia for most of my life and dont use the phraseology of your average ocker. But is that necessarily bad? On the other hand I wrote
the article with me in mind - if several years down the track I wanted to improve the experiment, for instance along the lines of not_important, I
dont want to start by reinventing the wheel, because ive forgotten something that ive already found out before. If there is any unnecessary,
repetitious or self-indulgent information in the text please point it out. I value straightforwadness but dont know what I can change Len
|
|
|
Magpie
lab constructor
Posts: 5939
Registered: 1-11-2003
Location: USA
Member Is Offline
Mood: Chemistry: the subtle science.
|
|
Len, this is an outstanding piece of work on many fronts. I especially like the fact that you are doing your own research to find optimum methods,
not just accepting time honored publications.
This work is timely to what I am trying to accomplish. My efforts are in a stall due to a mild case of intestinal flu. But I intend to read your
document carefully before proceeding. To date I have only read part of it. The one question I have is on what you are using to absorb the byproduct
HCl gas. You say you use an NaCl solution. Did you mean NaOH?
Sorry I won't be able to provide a complete critique in your 24-hour window. Perhaps Polverone will provide an extension for such an extensive
document.
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Thanks Magpie. Im using two different absorber bottles for the HCl, NaOH solution (in the glass-wool plugged bottle) and conc NaCl solution (in the
flask). Most of the time I run the HCl over the NaOH solution, but when adjusting the Cl2 flow rate and when trying to gauge end of hydrolysis, I
buble the exhaust into the conc. NaCl in the flask.
In the former case you can see when all Cl2 is consumed by the absence of bubbles (just NaCl spit), if the Cl2 flow rate is adjusted beyond that point
bubbles begin to appear which is both wasteful and hazardous.
In the latter case HCl is absorbed in conc. brine at the end to indicate, by lack of NaCl crystal formation when no more HCl is being formed. Neutral
H2O is actually quite good at absrorbing HCl and I felt none of its odour (although this was of course in a fume hood - but I had to put my head in
when adjusting things). Both bottles get quite hot to the touch when HCl is absrobed, and this is an additional indication.
I do hope you get over your stomach flu - I prefer almost any other thickness. When I was doing the sodium and ether work I had what they think was a
stomach virus which then prob. driggered a duo ulcer. Took a while to get over, even now looking at the sodium pictures I feel sick in the stomach.
Still it gave me lots of time doing nothing to write those posts.
I do hope maybe Polverone, if he reads this, will consider lifting the 24hr edit limit on the prepublication section. After all being able to edit in
response to critisism is the whole point of that section. Len
|
|
|
Fleaker
International Hazard
Posts: 1252
Registered: 19-6-2005
Member Is Online
Mood: nucleophilic
|
|
Excellent work as usual Len!! I very much like your format and how straightforward it is to read. I've never really had a profound need or desire for
benzaldehyde (I can readily make it via persulfate and benzyl alcohol), but it is convenient to have a reliable and practical guide
to making it.
I think any future posts of large projects that I make will be made in this style or something like it. It really is a good format and I think you
were right to suggest it way back when.
And Tim, wordy is how I write, not Len! 
The HCl and 3M brine trick is clever use of the ''common ion effect''. This was also the first time I saw an accurate representation of benzyl
chloride's effects--in particular you mention the ether ''carrying'' the BzCl; I think that is an important note, and looking back through my lab
experiences, I now feel the same way about some nasties I work with!
Neither flask nor beaker.
"Kid, you don't even know just what you don't know. "
--The Dark Lord Sauron
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Thanks Fleaker, if people like this and find it useful, then the time spent postingg it hasnt been in vain.
I had benzaldehyde on my list of desirable syntheses because I can remember it as a 'nifty' chemical from my 2nd year organic chemistry course. If
benzyl alcohol was freely available here, I guess I might have been lazy and done it that way - but it isnt. Its not a chemical readily used in
agriculture and theres no other chemical industry to speak of here which means most chem suppliers dont stock it. On the other hand attempts to buy
it off the one or two specialized suppliers will be viewed with suspicion unless bought on the work account (which I cant honestly do). Thats because
its a precursor to benzaldehyde (whose sale is controlled, because its a precursor to a precursor). While we are lagging the world here in our use of
chemicals, we are leading it in our fear of them - at least we are leaders in something.
I remember a few years back ordering some chemical from the specialized supplier I use at work for my hobby. The sales girl on the line first
established that I was asking them for a non-account sale, and to my quote of their stock number of H2O2 replied that this very suspicious - I asked
what she means - and she replied that I should have access to their stock numbers. I found that treatment offensive, and asked to speak to the
manager. She recognized me and asked why Im not ordering on the account as per usual. I explained because its not for work - at which stage she
couldnt help me. They have now lost most of my work orders - and I never want to go thru a similar experience.
On the other hand this is one of the most interesting syntheses ive done - and so another example where adversity actually leads to benefits in the
long run.
[Edited on 29-3-2008 by len1]
|
|
|
ziqquratu
Hazard to Others
Posts: 385
Registered: 15-11-2002
Member Is Offline
Mood: No Mood
|
|
Hey len, great write-up. A couple of ideas that might improve yields in the purification steps:
First off, to dry your adduct, why not simply put it under vacuum for a couple of hours? You might avoid the decomposition if you keep it at room
temperatue, and the ether won't last long even under a comparatively low vacuum.
The other place you might be able to easily save yourself some yield is in the work-up. Instead of steam distiling, why not go the classical route -
extract the benzaldehyde with ether or DCM, dry (sodium or magnesium sulfate) and remove the solvent and then vac distil the aldehyde? Unless you're
trying to avoid any residual benzyl chloride, which there should be none, this might be a better method if possible. The other advantage to this is
that you should minimise contact with air, thus reducing losses by oxidation.
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Exactly what I was thinking. Next time one should just use vacuum. I realise the purification procedure is where most improvements can be made. The
steam distillation - though recommended in Wieland, can also prob. be avoided as you say. I had no choice in the matter as my Na2CO3 obtained from
calcining NaHCO3 contained organic particles which stuck to the benzaldehyde-water interface and would have made clean separation in a funnel
impossible.
It might be possible to avoid using ether altogether by dissolving all the adduct in water and separating from the BzCl. I have not been able to get
data on its solubility - or the extent to which it decomposes on solution. Maybe someone knows?
Anyway its too late for me to repeat this as I have my pure benzaldehyde now, and many other interesting adventures are awaiting.
|
|
|
MagicJigPipe
International Hazard
Posts: 1554
Registered: 19-9-2007
Location: USA
Member Is Offline
Mood: Suspicious
|
|
Surprisingly, I have found that most people don't view benzyl alcohol as suspicious more so than other chemicals. I can easily obtain benzyl alcohol
in my area from pharmacies and a small "head shop". Apparently it is used in perfumes (I assume to make esters).
When I first asked the pharmacist for it, he said he had to "ask the state board" to make sure "it can't be used to make drugs". I got the okay the
next day. It's extemely expensive, though.
Would it be easier to make BnCl from BnOH and then chlorinate that to BnCl2?
"There must be no barriers to freedom of inquiry ... There is no place for dogma in science. The scientist is free, and must be free to ask any
question, to doubt any assertion, to seek for any evidence, to correct any errors. ... We know that the only way to avoid error is to detect it and
that the only way to detect it is to be free to inquire. And we know that as long as men are free to ask what they must, free to say what they think,
free to think what they will, freedom can never be lost, and science can never regress." -J. Robert Oppenheimer
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
You could, but I dont know why, if youre going to chlorinate you might as well start from toluene. Ive seen plenty of BzOH -> benzaldehyde
methods. Some are posted on this forum i believe, Fleaker was referring to one. But once you have the alcohol the hard part is done, and its the
fact that so many reactions are occuring (plus the simple starting materials) that interested me in this synthesis.
[Edited on 1-4-2008 by len1]
|
|
|
evil_lurker
National Hazard
Posts: 767
Registered: 12-3-2005
Location: United States of Elbonia
Member Is Offline
Mood: On the wagon again.
|
|
www.lemelange.com
1 gallon USP (it was at one time, can't call it that now since it was repacked under non-USP conditions) =$35
Cheap as dog shit.
Not all chemicals are bad. Without chemicals such as hydrogen and oxygen, for example, there would be no way to make water, a vital ingredient in
beer.
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
I could only find one online supplier of this dog s**t here. Heres what he has to say about it
BENTONITE Wine Grade Powder
BENZALKONIUM CHLORIDE 10% Solution
BENZALKONIUM CHLORIDE 50%Solution
BENZENE AR
BENZENE Tech
BENZOCAINE - Restricted
BENZOIC ACID Powder
1,2,3-BENZOTRIAZOLE
BENZYL ALCOHOL - Restricted
BENZYL BENZOATE Pure
BIAL'S REAGENT
BISMUTH CHLORIDE
BISMUTH CARBONATE
So looks like you cant buy dog s**t you must use something else. That is no so bad as I keep pointing out - makes one resourceful, and you have more
fun and find out more.
[Edited on 1-4-2008 by len1]
|
|
|
microcosmicus
Hazard to Others
Posts: 287
Registered: 31-12-2007
Member Is Offline
Mood: spin up
|
|
Availability of chemicals varies widely in time and space.
In particular, different countries have different policies ---
While Australia restricts benzyl alcohol, the U.S. restricts
toluene so people in these countries will want to run the
process in opposite directions.
Because OTC availability and legality vary and change, it
is good to know how to prepare chemicals from all sorts
of sources so one does not get caught unawares when
a chemical becomes unavailable. That, of course, makes
one all the more resourceful.
Also, even if all the chemicals were available at good prices
so one didn't need to prepare half of them oneself, this is
still interesting as a reaction in its own right. I found this
interesting reading and maybe will try it some day even if
the toluene I use as starting matrerial will have been prepared
from benzoic acid or benzyl alcohol.
|
|
|
garage chemist
chemical wizard
Posts: 1803
Registered: 16-8-2004
Location: Germany
Member Is Offline
Mood: No Mood
|
|
Your detailed investigations are without equal on this board len1.
It's great that you take the time and actually research the different available methods that can be used, and come up with better procedures than
those found in literature.
I find it hard to believe that benzyl alcohol is actually restricted in australia. Have you tried to get it from a pharmacy? It is used in lotions and
cosmetics, you can tell them that you want to make those yourself.
Alternatively, purchase an ester, like benzyl acetate ot benzyl benzoate. Those are used in fragrances and you can hydrolyze them.
Here in germany, this can be done for phenylacetic acid, which is regulated, but its esters aren't.
A special question: Do you know whether the sodium bisulfite can be replaced by potassium bisulfite? The latter is available for an extremely low
price from winemaking suppliers, while the former has to be bought from chemical suppliers and costs several times as much.
I could do some experiments to the solubility behavior of the potassium salt of benzaldehyde bisulfite adduct, e.g. whether the adduct crystallizes
out when benzaldehyde is shaken with saturated potassium bisulfite solution.
If the potassium salt turns out to be too soluble, one can always make SO2 from the potassium bisulfite and bubble it into NaOH, but it would be good
to know whether this is actually necessary.
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
@microcosmicus
I cant believe anyone would be so silly as to restrict toluene. Petrol is about 30% toluene. All it takes is a few hours specnt with a good
fractional distillation setup and youre done. If toluene was restricted here, thats how I would have started - and it wouldnt have added much to the
time since you have to distil the toluene anyway. Its in fact cheaper that way, I pay about $12/L for toluene, whereas petrol is $1.35/L.
@garage chemist
Thank you for your kind words. I really didnt expect that Ill have to do so much work when I started out on this - I thought the stuff in the
literature was all correct. I was quite surprised.
I think you are right in what you said before, we have to move on, get P done, get all these syntheses that have been dragging in threads here for
years finished, and get on to new things. That way well bridge the gaps to the prof chemistry sites.
Its sodium rather than potassium metabisulfite thats sold here for wine making for some reason. And its very cheap, $6/kg. I wonder why they use the
K salt in germany - perhaps it gives a softer wine (here the wines are quite harsh). I have not been able to find anything on solubility of the Na
adduct, let alone the K adduct, in the literature - except what I could observe myself - that is that the Na salt is quite soluble - perhaps 100g/L.
Its the common ion effect that gets it percipitated. So I would expect on that basis there should not be much difference between the Na and K salt
solubilities (especially as the anion is much larger than either so atomic radii seem not so important).
I can only imagine the faces in the pharmacy if I should ask for benzyl alcohol - today I heard on the radio that you can expect to get the third
degree if you ask for sudafed there - because of a big drive against illigal activities. So given its restricted at the chem supplier I wouldnt even
try. You are lucky to live in a country where much of this was invented, and theres a completely different culture and understanding of that sort of
thing there. I did want to ask you whether its pharmacies in germany that sell red phosphorus, or is it chem suppliers? Len
[Edited on 2-4-2008 by len1]
|
|
|
Magpie
lab constructor
Posts: 5939
Registered: 1-11-2003
Location: USA
Member Is Offline
Mood: Chemistry: the subtle science.
|
|
Concerning the availability of benzyl alcohol, Evil_Lurker says:
| Quote: |
www.lemelange.com
1 gallon USP (it was at one time, can't call it that now since it was repacked under non-USP conditions) =$35
|
They don't sell it anymore. Is there a message here?
Just goes to reinforce the old adage: "The time to stock up is now."
[Edited on 1-4-2008 by Magpie]
|
|
|
evil_lurker
National Hazard
Posts: 767
Registered: 12-3-2005
Location: United States of Elbonia
Member Is Offline
Mood: On the wagon again.
|
|
| Quote: | Originally posted by Magpie
Concerning the availability of benzyl alcohol, Evil_Lurker says:
| Quote: |
www.lemelange.com
1 gallon USP (it was at one time, can't call it that now since it was repacked under non-USP conditions) =$35
|
They don't sell it anymore. Is there a message here?
Just goes to reinforce the old adage: "The time to stock up is now."
[Edited on 1-4-2008 by Magpie] |
Really, did you contact them today?
Its still on their website.
Not all chemicals are bad. Without chemicals such as hydrogen and oxygen, for example, there would be no way to make water, a vital ingredient in
beer.
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
| Quote: | Originally posted by len1
...
I cant believe anyone would be so silly as to restrict toluene. Petrol is about 30% toluene. All it takes is a few hours specnt with a good
fractional distillation setup and youre done. If toluene was restricted here, thats how I would have started - and it wouldnt have added much to the
time since you have to distil the toluene anyway. Its in fact cheaper that way, I pay about $12/L for toluene, whereas petrol is $1.35/L.
|
I think you will find that the aromatics content of automotive fuel varies from region to region. In Hong Kong the BTEX rations are 2:15:2:15, in
North America they're roughly 1:2:1:4. The amount of BTEX in N American gasoline is about 18%, so that's less than 5% toluene and a bit over 2%
benzene (they may have pushed that benzene level down towards 1%, NZ has nearly 50% BTEX.
| Quote: |
...
Its sodium rather than potassium metabisulfite thats sold here for wine making for some reason. And its very cheap, $6/kg. I wonder why they use the
K salt in germany - perhaps it gives a softer wine (here the wines are quite harsh).
...
[Edited on 2-4-2008 by len1] |
Germany tend to use potassium salts more frequently than some other countries, this is because of the large deposits of potassium salts there.
|
|
|
Magpie
lab constructor
Posts: 5939
Registered: 1-11-2003
Location: USA
Member Is Offline
Mood: Chemistry: the subtle science.
|
|
Evil_lurker: I took another look at the website. I was mistaken. I hadn't scrolled over far enough to the right and only saw the first two empty
columns. My apologies.
This is good news!
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
| Quote: | | The amount of BTEX in N American gasoline is about 18%, so that's less than 5% toluene and a bit over 2% benzene (they may have pushed that benzene
level down towards 1%, NZ has nearly 50% BTEX. |
Here its pretty much the same as NZ - if I recall correctly from my benzene extraction from petrol data, toluene was about 30%. Sounds like US petrol
is completely different. Still toluene is the best solvent for contact adhesive, Im sure its used that way in the US as well.
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Most of the larger, more densely populated countries seem to be pushing to reduce aromatics in vehicle fuel, if for no other reason than ground water
contamination. Countries with low population densities seem to go for high BETX, because that is an easy way to get the octane rating up.
|
|
|
| Pages:
1
2 |
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
