Note: Descriptions are shown in the official language in which they were submitted.
`
:: :
~0674~9
, :
MDZ6142 ~:
The presont in~ention relates to the oper~tion of mercury-
.~ cathode cells rOr the electrolysis of alkali-metRl chloride
solutionB~ More particularly it relate~ to an improved method
for the operstion Or mercury-cathode cells.
Many countries throughout the world h~ve large instsllations
;
: Or cells rOr the ~anuracture o~ chlorine ana caustic alkaii by
. the electrolysis Or alkali-~etaI chloride solution, wherein the
, ~:
: said 801ution is electrolysed while flowine between the lover
faces Or an array Or graphite os metal anode plates and a flowing
liquid cathode, which i9 maint~ined by ~eeding in mercury or
-2-
'
:: :
~7449
- dilute slkali-metal amalgam at one end or one side of the cell
and ~ithdrawing amalgam enriched in alkali-metal at the opposite
end or side o~ the cell. Chlorine li,berated at the anodes is
continuou~ly removed from the top o~ the cell and the liberated
alkali-metal~ ~hich collects in the ~lowing amPlgam c~thode is
continuously removed in the enriched amalgam and converted to
caustic alkali by reaction of the enriched amal~2m with water in
a soda cell~ U~Ufilly cslled a denuder, fr3~ which denuded a~algam
is recirculated by means o~ a pump to the electrolytic cell~
io A well-known problem i~ operàtin~ such~cells is the build~un
Or deposits Or thick mercury~ sometimes re~erred to as "mercury
butter~ on the baseplate o~ the cell. The m~chanism o~ thiok
~ercury formatio~ i8 not ~ully understood. It i8 thaught to be
influenced by the presence o~ trace impurities in the brine
electrolyte5 but even ~ith the be~t practicable attention to
purification o~ the reed-brine thick mercury depo~it3 can build
up ~ith prolonged operation of the eell and the problem has tended
- - to bec e now acute with the hiBh current-density operation which
has been practised in recent ti~es.
Thick mercury deposit~ can cause current shorts between the
anode~ and the cathode amsl~a~. Besiaes reducing the current
efficiency o~ thc pro~es~ 3uch shorts can cause serious damaRe~
especially to the coated metal anode~ ~hich haYe recently begun
to replace co~ventional graphite anodes.
?5 We have now ~ound that the build-up Or thick mercury on the
cell baseplate can be reduced by dispersinR ~ater or an aqueous
medium in the amal~s~ csthode without interruption Or the electrolysis,
Accordin~ to the present inYention ~e provide a process for the
. ,
~ ' .
. ~
~ ~06~449
.
reduction of thick build up of mercur~ on the cell base-plate
o~ a mer~ury cell ~or the electrolysis of brine which com-
prises dispersing in the amalgam water or an aqueous medium
continuously or intermittently at one or more stations along the
line of the amalgam flow in the cell wit~out interruption of
the electrolysis,
The preferred aqueou~ mediurn i8 brin~ ~or ex~plo the aqueou~ ¦
electrolyte o~ the cell~ f
The accompanying dra~ings illustr~te sehematically ~even
:; suitRole em~odiments o~ apparatus for putting the process of the
in~ention into practice.
Each Or the ~igures 1 to 6 sho~s in isometrie projection as
seen from one side edge Or the baseplate the interior o~ a
portion Or a mercury-osthode cell in whi~h the cathode-amalg~m
rlOW i8 along the length of the cell. Fig 7 show~ a schematic
vertical section along the length of a mercury-cathode cell ~ith
, its mid-portion cut out an~ the external amalg~m circulation.
Fig 8 sho~s a detail o~ Fig 7.
In each Or the figures 1 to 6 the ~ell baseplate 1 carries
the rlowing amalga~ cathode 2 and above this is the flowing
nqueous electrolyte 3~ The anodes ~hich lie above the analgam
I layer in the electrolyte 3 have been omitted fro~ the drawings
for the sake o~ cl~rityO
In Fig 1 an inclined bafrle 4 is fixed across the line o~
amalgam flow ~ith its lower edge spaced from the baseplate 1
and belo~ the normsl lerel Or the Emalgamg and uith its upper
l portion lyin~ in the aqueous electrolyte 3. The barfle 4 ma~
suitably consist o~ B steel core coatea with Q chlorine-
resistant plsstics material. Because o~ the interference to
flo~ of the two liquids by the iDclined baffle and the very
_4-
- 1~67~4~9
high specific gra~ity of the amalgam layer~ a portion of the
aqueous electrolyte i~ drawn under the baffle with the amalgam,
turbulence is created and sufficient electrolyte is cau3ed to
disperse in the smslgam to cause decomposition Or any thick
mercury tending to deposit on the bsseplate ror an appreciable
distance downstreOEm rrom the bar~le. Depending on the length
of the cell and other parameters, such as the slope of the
baseplate and current density~ a plurality of baffles 4 spaced
apart alone the cell may be required to keep the whole length
Or the cell free rrOm thick mercury deposits.
In Fig 2 a roller 5 replaces the baffle 4 o~ Fig 1. The
roller is rotatable about an axle 6 carried in bearinRs (not
shown), ~hich may be fixed to the baseplate or the walls of the
cell. The roller is spaced away from the baseplate and is
immersed partly in the cathode amal~am 2 and partly in the aqueous
electrolyte 3. Rotation Or the rollers caused by the flowing
amalgam layer draws a portion of the aqueous electrolyte beneath
the roller and produces the required dispersion Or electrolyte
in the amalgam. Additional rollers 5 may be fixed at intervals
~20 down the line Or the amalgam flow.
In Fig 3 an electromagnetic tr~nsducer 7, working preferably
at an ultrasonic frequency~ is placed across the line of amalgam
flow to replace the baffle 4 of Fig 1 and the roller 5 Or Fig 2.
The position Or the lower surface Or the transducer i5 adjusted 80
a3 to o cillate about a mes~ position defined by the interface
- betweeD the amalgam layer and the aqueous electrolyte when the
transducer is not in operation. The transducer may be operated
continuously to maintain a dispersion of aqueous electrolyte in
f
7~4~
the amalgsm downstream from the position of the transducer or it
may be operated for short periods, e~ ten-minute periods,
intermittently 80 as to preYent the build-up of thick mercury
deposits to any significant extent. Additional trQnsducers 7 may
be installed at intervals down the line of the amalgam flow as
necessary to maintain any length Or basepl~te free from thick
mercury deposits. Furthermore, each transclucer ~hown in Fig 3 as
a ~ingle transducer 7 may be replaced by a plurality Or smaller
in~ependently sctuated transducers opersting side by side across
the line Or amalgam fl~w. The ~uspension arrangements and power
supply lines for the transducers, indicated as 7a, may suitably
pass through the cover of the cell (not shown).
In Fig 4 is shown an alternative arrangement employing
electromagnetic tran~ducers for dispersing aqueous electrolyte in
the flowing amalgam cathode. Here a transducer 26 is installed
above the cover of the cell (not shown) and the mechanical
vibrations ~enersted by the transducer are conveyed into the cell
by means of a bar 27 passing throu~h the cell cover and reaching
to the inter~ace between the amalgam cathode 2 and the aqueous
electrolyte 3. Although for clarity only one transducer is shown
in Fi~ ~, generally a plurality of transducers 26, each with its
attached bar 27~ will be installed in a spaced row across the cell
80 as to disperse aqueous electrolyte in the amalgam cathode over
the full width of the cell. If desired~ the cell cover may be
provided with a row of openable ports through which the bars 27
may be introduced only when it i~ desired to treat the a~algam
ca~hode for a ~hort period. After the treatment the transducers
may be moved to another station in the same cell, where another
-6-
~La367~L~9
row of openable port8 iB provided~ or to a dirferent cell. The
bars 27 mu~t bo made of chlorine-re~i~tant material or mu~t be
coated with a chlorine-resi3tant material. Most suitably the
bars are provided with an electrically-insulating coating 80 a8
to svoid the possibility of shorting between the anodes and th0
c~thode oY the working cell when the bar~ are being introduced
and re~oved,
Fig 5 sho~ an arrangement for introducing water or brine
into the amalgam cathode by injection from a ~eries af Jet oririces
placed above the amalgn~ cathode 2 in the cell. An inlet pipe 28
carries a serie~ of branches 29~ each Or ~hich i8 terminated by
a jet orifice 30 ~one only shown). The pipe 28 may be fixed within
the cell above the aqueou~ electrolyte 3; alternatively it may lie
outside the cell and the branche~ 29 will then pa98 into the cell
through the cell cover (not shown), The jet orifice~ 3Q may be
placed above the aqueous electrolyte as shown in the drawing or
; alternatively they may be placed within the aqueou~ electrolyte.
The end 31 or~pipe 28 iB closed and water or brine i~ fed in at
the end 32 under adequate pressure ~o that it igeueB from the ~et
oriric~s 30 with sufficient velocity to penetrate through the
aqueous electrolyte 3 and disturb the amalgam cathode 2. It may
suitably be arranged for the high-speed jets Or water or brine
thu3 produced to pass through hole~ or slots provided in the anode
plates or through 6aps provided between neighbouring anodes.
Fig 6 showfl an arrangement for inJecting water or brine
directly into the smalgam cathode. An inlet pipe 33 is fixed in
the cell~ prererably sbove the level Or the rlowing electrolyte 3
as sho~n~ and carries a series Or branche~ 3~ each of which
; -7~
67449
terminates in 8 jet oririce 35 (only one ~hown) uithin the
~louin~ amalg~m cathode layer 2. The end 36 Or pipe 33 i~
closed and water or brine i8 ~ed ;n at the end 37 unaer ~ufficient
pressure to emerge into the amalga~ cathode layer throuRh the Jet
orifices 35.
In general suf~icient water remains dispersed in the mercury
returning to tbe chlorin~ cell ~rom the denuder to pre~ent the
build-up o~ thick ~ercury depo3its ~or a ~hort di3tance downstream
rrom the mercury inlet of the cell. It is not thererore neces~ary
to disperse additional uater or brine in the amalgam cathode near
. the mercury inlet and the ~irst 3tation downstream fro~ the inlet
at ~hi.ch any of the de~ices.shown in figures 1 to 6 i8 installed
: ~ wilI generally be chosen to lie ju~t before the po~ition at which
the build~up Or thick ~ercury deposits i~ known to start during
conventionnl operation o~ the cell.
In Fi6 7 is shown another arrange~ent Or appara~us ror
~: di3persing water or brine in the ~lowing cathode a~al~sm within
: ~ the cell in accordance with the invention. The cell ba3eplate 1
~ : carrie~ the rlowing amslgam cathode 2 and abo~e this i9 the
,. .
: 20 flo~ing brine electrolyte;3. 8 are the end walls Or the cell
:, . .
through Yhich pass the brine inlet 9~ the ~pent-brine outlet 10,
sn inlet 11 Yor mercury or dilute amalgnm and ~n outlet 12 ~or
enriched ~alga~. ~he cell cover 13 is pro~ided with a~ outlet
: 14 for chlorine ~as~ and the elcctrical connection~ 15 pas6
throu6h the co~er to the anodes 16. Enriched amalgam i8
continuously re~o~ed from the cell at 12 to the denuder 17, where
it i8 decomposed by a ~upply of water (not shown), and a stream of denuded
; , .. . .. . . .. .
amalgam flow~ from denuder 17 through water-wash box 18 ¦ -
-8-
1~ - . .
1067~49
and pu~p 19 back to the inlet end Or the cell through pipe 20.
The main ~tream o~denuded a~algam~ returns to the cell ~hrough
vslYe 21 and inlet 11. AB discus3ed hereinberore~ ithe mercury
. stream returning to the cell ~rom the denudler by way Or nipe 20
contains ~ome water dispersed in it. According to the embodiment
of the invention illuatrated in Fig 7~ a small ~raction Or thi~
- ~et mercury stream~ controlled by valve 22 in conJunction with
val~e 21, is fed by way Or pipe 23'to a spreader-bar 24 which lies
~cros~ the li~e Or amalgam rlo~ and dips i~to the amalgam cathode
10 JUst upstrea~ ~rom the position where build~up of thick mercury
depogit5 i5 kno~n to start during con~entionsl operation of the
.
cellO The detail of the ~preader-bar i8 ~hown in side elevation
in Pig 8. The stream o~ wet mercury delivered throu~h pipe 23
pa~se~ into the hollow spreader-bar 24 and enters the flowin~
15 amalgam cathode through a series of orifice~ 25. Additional
spreader-bars 24 connected to branches (not sho~n) rrom pipe 23 ma~
be in~tslled in the cell at interval~ further down the line of
amslga~ ~lo~ to m~intain a~y lenBth o~ baseplate free from thick
mercury depo~its.
. 20 As a modification o~ the embodiment o~ the invention discussea
. in the precedi~ p~ragraph, additional water or brine (prererably
: unchlorinated reed,brine) may be di~persed in the mercury
returning to the cell ~rom the denuderg for i~3tsnce by installing
an e~ulsirier red ~;th the chosen aq~eous medium either in the
25 total mercury ~tream flo~inR in line 20 o~ Fig 7 as indicated at
A or in the ~inor ~tream flo~ing in line 23 B8 indicated at B.
This modi~icatio~ ha~ the advantaRe o~ compensating ~or the
; progre3sive separa~ion o~ di~persed water from a ~ercury stream
'.
' -9-
7449
,
as it moves along the pipelines, caused by the ~reat difference
in specific gravities, and ensures 8 higher concentration of
aqueous phRse at the points Or application. If desired~ when a
plurality of spresder-bars 24 i3 in~talled in the cell, a 3eparate
emulsifier B may be in~talled in each of the branches Or line 23
which feed the individual spreader-bars.
By the term emulsifier, we mean a mecbanical device consisting
`:`
of rotating vanes fitting within a ~tationary perforated screen,
and which is positioned across the inter~ace between the amalÆam
and the water to be added BO h~ to draw water into the Emalgam
when the ~anes ar~ rotated,
The invontion is further illustrated but not limited by
the rollowing Exa~ples.
Exam~le 1
The rate of formation of thick mercury content in a mercury
cell was determined as ~ollows. A vertically adjustable copper
probe (di&meter 3 m~) was inserted through a hole in the cell cover
and ~crewed down towards the baseplate. The probe ~as connected
by m~ans Or an electric circuit including a voltmeter to the cell
baseplate. A rirst reading on a micrometer gauge was taken when
the probe made contact with the upper surface Or the mercury as
indicated by the voltmeter. A second reading on the micrometer
gauge when mechanical contact was made with the cell ba3eplate.
The dirference between the aroresaid readings (about 5 to 8 mm)
indicated the thickness of the mercury rilm and could be measured
to an accuracy of - 0.05 mm. Several mea~urements e~ 20 to 30,
were taken at different positions along the cell and the mean
thickness of the mercury film in each position was calculated. The
--10_
.
-'` " 1~36744'9
measurements ~ere repeated at re~ular intervals over 5 periods
of S to 6 day~ each and the rate o~ increase of the thickness
o~ the mercury fil~ was equated with the r~te of formation of
thick mercuryO
A mercury cell operating with a currerlt of 180~000 amp was
~ed with ~ m3~hour of 25.5% W~w godiu~ chlc>ride brine. The mea~
mercury ~low rate ~as 53 litres/minute Water was inaected into
the pump 19 (sho~n in Fi~ 7) at a rate in the range 60 to 95 litres¦
hour.
For the purposes of comparison~ the =ercury cell was operated
under th~ same conditions of brine flow and current, with a mean
. . . .
mercury Mow ra~e o~ 47 litr~sjminute~ but ~ithout water injection.
The measuremenes uere carriedo~e~ seven periods of 5-8 days eaph.
The observed meQn rates o~ ~ormation o~ thick mercury at
di~rerent position~ along the cell~ with and without water
injectio~, are sho~n belowD
Me~n rate of formation of
. thick mercury ~m¦day)
-- Dl~ta~ce along
the cell ~feet) uith waterwithout water
i~jectioninjection
.
: 2012~5 OoO10 0~035
0~034 00355
375 o.38? . 0~450
~ The uater content o~ the amalga~ and the particle 3ize of
-~ the ~ater was ~e~sured by a}lo~ing water particle~ to rise to
~` 25 the surrace o~ the amalga~ and ~easuri~g the s~ount of ~ater
collecting st the surface at Ysrious time intervals. When there
~as no ~ster injection, the mercury rced to the cell contained
0~02 to Ool ppm by ueight Or w~ter whose particle size was le8S
~' .
`` i6)67~49
than 12 micron Sto~e~ diameter; when water was injected, the
mercury feed contained 0.2 to o~6 ppm by wei~ht of water whose
particle size was less than 12 micron Stokes diameter. These
measurements confirmed th~t injection Or w~ter into the pu~p had
produced an increased amount Or dispersion of water in the amal~am
feed to the cell (there was already some water present in this
amalgam).
' ~
An emulsifier (as shown at A in Fig 7) was incorporated in
the mercury feed to the cell. The mercury cell was operated under
the s~me condition~ of brine flow and current as in Example 19 with
a mesn mercury flow rate of 62 litre~/minute~ over ~i~e periods Gf
5 to 9 days each~ but using the emulsirier instead of water
injection into the pump. The emulsirier~ which was of the type
described hereinbefore was manufactured by Silver~on Machines
Limited, Eneland.
For purposes of cQmparison~ the mercury cell was operated
under the~sa~e conditions of brine flow and current, with a mean
mercury flow rate of 10 litres/minute~ over five periods of 5 to
9 days each, but without the emulsifier.
. ~he obser~ed mean rates o~ formation of thick mercury at
different positions along the cell, with and without the emulsifier,
are sho~n below: .
Distance along Mean rate o~ formation of
the cell (feet) thick mercury (mm/day)
with emulsifier without emulsifier
6.3 0 0.015
12.5 0.009 0~065
1~,8 0.011 0.072
: 25 0.085 0 193
31 0~1~5 0.204
i7~4~
Measurements as de~cribed in Example 1 showed that the
emulsifier produced sn increased amount of dispersion of water in
the mercury reed to the cell. Thus, when the emulsirier was not
in use, the mercury ~eed to the cell contained 0.1 to 0~4 ppm by
wei~ht of wster who~e particle size wa~ less than 12 micron
Stokes diameter; when the emul~ifier ~as used, the mercury feed
contained 1 to 4 ppm by weight Or water whose particl~ size was
less eh 12 micron Stokes di~meter.
'
:
j .
',
~ '
~ .
'
, : :
, : .
~ .
-13-
