The OXIDATION of 3-AMINO-
PHTHALHYDRAZIDE ("LUMINOL")
as a LECTURE DEMONSTRATION
of CHEMILUMINESCENCE*
ERNEST H. HUNTRESS, LESTER N. STANLEY,
Attention is directed to the most striking instance of
chemiluminescence whuh appears to have b m observed
and to the extraordinary advantages which the reaction
possesses mer the conventional cases. A simple eco-
nomical two-stage synthesis of 3-aminophthulhydrazide is
described which readily gives excellent yields of +re
product. Specifcc directions are given for demonstration
of its brilliant chemiluminescence i n dilute aqueous al-
kaline oxidizing agents.
T HE STUDY of chemical reactions in which visible
light is evolved a t ordinary or low temperatures
has largely been left to physics, physical .chemis-
try, or biochemistry. Nevertheless, countless in-
stances of such development of visible radiation are on
record and many (if not most) involve organic com-
pounds. However, in the vast majority of cases either
the amount of light evolved is not impressive, or the
reaction is brought about only under inconvenient con-
ditions, or the substances involved are expensive or
difficultly accessible. Partly for these reasons general
attention to the matter has been but occasional and the
most impressive and convenient illustration of the
phenomenon appears to have been largely overlooked.
Most chemiluminescent reactions are oxidations.
Extensive rCsumCs of hundreds of inbvidual cases are
already available.' Of the many instances of organic
compounds of known structure given in these refer-
ences there appears to be a small group which gives a
maximum of visible light. This group includes the
following: (1) various reactions of Grignard com-
pounds, such as the Wedekindereaction between chloro-
piain and phenylmagnesium bromide in ether solu-
tion, the reaction of Grignard compounds with other
nitro compound^,^ or the oxidation of various Grignard
* Contribution No. 101 from the Research Laboratory of Or-
ganic Chemistry, Massachusetts Institute of Technology, Cam-
bridge, Massachusetts.
(a) Truurz, Z. wiss. Phot., 2 , 217-23 (1904); (b) Z. phyn'k.
Chcm., 53, 1-111 (1905); (c) HARVEY, 3. Biol. Chem., 31, 311-6
,---.,.
(1417)
( a ) WEDEKIND, Ber. deut. physik. Ges., 4 , 417 (1906); ( 5 )
Phyrik. Z . , 7 , 805 (1906); (6) Z . nuiss. Phot., 5, 29 (1907); ( d )
HECZKO, Chem.-Ztg.. 35, 199 (1911).
(a) GILMAN, MCGLUMPW, AND FOTEERGILL, RCC.trav. chim.,
49, 526-31 (1930); ( b ) 49, 726-8 (1930).
1 2
AND ALMON S. PARKER
compound^,^ notably (1) pchlorophenylmagnesium
bromide; (2) the Trautz-Schorigin6 oxidation of pyro-
gallol-formalin solutions in potassium carbonate by
means of 30% hydrogen peroxide (perbydrol); (3)
the oxidation%y means of hydrogen peroxide or halo-
gens of alcoholic alkaline solutions of lophine (2,4,5-tri-
phenylimidawle) or amarine (4,5-dihydro-2,4,5-tri-
phenylimidazole). It will readily be noted that even
these reactions involve certain difficulties. For ex-
ample, some of the materials are unpleasant to handle
(e. g., chloropicrin is a lachrymator, Grignard reagents
are unstable and involve the use of ether solutions, 30%
hydrogen peroxide is dangerous to handle, unstable,
expensive, and not always readily accessible, amarine
and lophine not very readily preparable, and their
oxidation in strong alcoholic alkali is not well suited
to convenient demonstration).
In contrast to these difficulties the production of
chemiluminescent effects by means of the oxidation of
3-aminophthalhydrazideposse& great advantage. The
reaction is carried out in exceedingly dilute aqueous
alkaline solution, it requires but the very mildest and
most easily accessible oxidants, it utilizes ordinary
pharmaceutical (3%) hydrogen peroxide, and the
essential reagent itself is naw readily obtainable.
Furthermore, the brilliance of tht effects which can be
produced far surpasses any of the other recorded cases.
Dufford, Calvert, and Nightingale" state: "The
[chemiluminescence from the] compound p-CICaHa-
MgBr is much brighter than any other so far found.
Careful pyrometric measurements show that it is
brighter even than luciferm, except possibly for the
bright specks in the luciferin solution and therefore
probably the brightest case of chemiluminescence on
record." We find, however, that the brilliance of light
which is produced under optimum conditions from
the oxidation of "lnminol" renders that from this
Grignard reagent dim by comparison.
' (a) EVANSAND DUPPORD,3. Am. C h m . Soc., 45, 278-85
(1923); (b) DUPFORD, CALVERT, AND NIGR~NGALE, ibid., 45,
2058-72 (1923); (c) D~FFORD, NIGHTINGALE, AND CALVERT,
ibid., 47,95-102 (1925); ( d ) EVANSAND DIEPENHORST, ibid., 48,
715-23 (1926).
(a) TRAuTz, Z. Elcktrochem., 10, 593-6 (1904); (b) TRAWTZ
AND SCHORIGIN, Z. tom. Phot., 3, 121-30 (1905-6).
"(a) RADZISZEWSKI, Ber., 10, 70-5 (1877); (b) BHATNAGAR
AND MATHUR. Z. physik. Chem., 159, 454-8 (1932).
 3-Aminophthalhydrazide (I), also referred to in the (Cf. Refs. 7, 8) and begins directly with 3-nitrophthalic
literature as 5-amino-1.4-dihydroxyphthalazine,*was acid.'3 It avoids the excess of expensive hydrazine re-
first reported by Schmitz' as the result of a long syn- quired by early investigators and uses only that amount
thesis starting with triethyl hemimellitate. Later actually needed for the heterocyclic ring. By sub-
stituting sulfide as the reducing agent it avoids the
0 ONa 0 difficulties attendant upon the separation of an am-
photeric reduction product from a stannous solution.
The new process is readily adaptable to the preparation
of large amounts of material, if desired, and gives ex-
cellent yields of pure product. The yield of 3-nitro-
phthalhydrazide is quantitative, its reduction to
aminophthalhydrazide is nearly so.
The PreNration of 3-Nitrophthu1hydrazide.-Bogert
and Boroschek,14who first reported this compound, and
Curtius and Sempers obtained it by heating the mono also Radulescu and AlexalZbevaporated the liquid ob-
ethyl ester of 3-nitrophthalic acid for ten hours with tained by adding to one mol of 3-nitrophthalic acid in
excess hot hydrazine hydrate. In 1928 Albrechtg alcohol one mol of hydrazine (in the form of hydrazine
drew attention to the chemiluminescence of the ma- hydrate solution), and heated the residue a t 250"
terial and carried out some spectral measurements. for a long time to form the cyclic hydrazide. We prefer
However, the actual method of preparation of his ma- the following method which avoids the use of hydrazine
terial was not given in the above reference but only hydrate, is more rapid, and gives quantitative yields.
in the dissertation10 itself. Starting from 3-nitro- Solid hydrazine sulfate (130 g. or 1 mol) and crys-
phthalic acid it passed successively through the diam- tallized sodium acetate (272 g. or 2 mols) are dis-
monium salt, 3-nitrophthalimide, 3-aminophthalimide, solved in 400 ml. of hot water and the clear solution
and finally 3-aminophthalhydrazide. No yields are added to solid 3-nitrophthalic acid (211 g. or 1 mol)
recorded in any of these steps and our attempts to use contained in an 8" porcelain evaporating dish. The
the method gave very unsatisfactory results. In 1929 latter is placed upon a tripod and evaporated as rapidly
Harvey," obtaining his material from Albrecht, ex- as possible over a free flame with constant stirring with
amined the chemiluminescence observed during elec- a flat porcelain spatula to avoid bumping or decom-
trolysis. During the progress of our present work a position. This operation requires about 1.5 hours and
statement1% appeared noting that 3-nitrophthalhy- is best conducted a t a hood in order to facilitate evapo-
drazide could be reduced to the amino compound by ration and to remove the acetic acid vapors. The
means of hydrogen sulfide, hut no details were given residual dry solid is removed from.the dish, ground to
nor yields stated. Apart from the above references no a fine powder, placed in a wide-mouthed flask or beaker,
further mention of 3-aminophthalhydrazide could be and heated for a t least three hours a t a temperature
found in the chemical literature. of 160 * lo0 in a suitable oil bath. During this opera-
tion the solid should frequently be stirred and if there is
THE PREPARATION OF 3-AMINOPHTHALHYDRAZIDE'
any tendency toward caking in the earlier part of the
baking it may be necessary to remove and powder the
We report here a method of preparation which avoids material before continuing. It is advfsable to measure
all the weaknesses of preceding procedures. The the temperature in the heated powder rather than
reaction involves first the preparation of 3-nitrophthal- externally and care must be taken not to overheat the
hydrazide and its subsequent reduction to 3-amino- mass. The escape of steam may also be facilitated by
phthalhydrazide by means of ammonium sulfide. The passing a gentle stream of air through the flask.
method avoids difficultly accessible starting materials When the heating is finished, remove the solid,
powder, and extract twice with 350 ml. of hot water to
* In the interest of assigning to this compound a simpler and remove sodium sulfate. The residual solid is then
more euphonious name we prefer to csll the material "luminol."
This associates it with the idea of luminescence and denotes its dried a t 105' to constant weight: yield, 206 g. 3-nitro-
enolic character. We do not regard as important the possible phthalhydrazide (99.5% theoretical). The product so
objection that the name might be confused with the hypnotic
"luminal." obtained is free from sulfate, melts 297' to 300' u.c.,
1 (a) SCHMITZ,Dissertation, Heidelberg. 1902; (6) Cf. and is sufficiently pure for reduction without further
Cun'crus AND Scmrrz, J. prakt. Chem., 91,46.97 (1915). treatment.f
CURTIUS AND SEMPER, Ber., 46,1162-71 (1913).
ALBREWIT, Z.physik. Chenz.. 136, 321-30 (1928). Reduction of 3-Nitrophthulhydrazide to 3-Amino-
' 0 ALBREWIT.Dissert~tion.Friedrich-Wilhelms Univ.. Berlin, -
1928. '3 "Organic syntheses," John Wiley & Sons, Iuc.. New York
"HARVEY,J . Phys. Chem.. 33, 145G9 (1929). City, 1927, Vol. 7, pp. 70-2.
1% (a) RnoaEscu am ALEXA.2. phy{k. C k m . , B8, 382-94 I4 BOGERT AND BOROSMEK, J. Am. Chem. SOC..23,750 (1901).
(1930); (b)Bull. soc. chim. RomSnia, 12, 140-63 (1930). $NOTE: Although BOGERT AND BonoscHEKU state that 3-
t Experimenters who may not care to prepare this substance nitrophthalhydrazide melts with decomposition at about 320"
we have consistently observed values close to 300' u s . FLADU-
for themselves can now obtain it from the Synthetic Organic
Chemicals Department of the Eastman Kodak Company, Roch- ~ s s c uAND ALEXAI~ give m. p. 310-311' but do not state
ester, N. Y. whether corrected or not.

LIQUIDLIGHT
This photograph was taken in an otherwise absolutely dark
room by means of the light given off during the oxidation of
"lnminol" according to the method described in the accom-
panying article.
In some runs all the sulfur remains in the excess sulfide
solution and a yield of very pure "luminol," as high as
64.5%, has been obtained directly a t this point.
The filtrate from this primary precipitate is then
acidified with a slight excess of glacial acetic acid and
the resultant precipitate of mixed sulfur and "luminol"
is filtered with suction, washed with water, and dried.
In order to free the aminophthalhydrazide from sulfur,
the crude mixture is stirred up with that amount of 5%
aqueous sodium hydroxide solution which would just
correspond to the assumption that the dry solid cou-
tained no sulfur. After stirring and very slight warm-
ing the solution is filtered from undissolved sulfur,
cooled to 0°, and stirred, and scratched. Presently
precipitation of the mono-sodium salt of 3-amino-
phthalhydrazide (II or Ila)+begins and increases for
some time. Finally, the solid is filtered with suction,
pressed as dry as possible on the filter, then washed
sparingly with dry alcohol or ether. It may then be
dried in the air, if it is to be preserved as the sodium salt.
If, however, it is desired to reconvert the substance to
"luminol" the original solid is not washed with alcohol
but is redissolved in water, and reprecipitated by adding
a slight excess of glacial acetic acid. The voluminous
flocculent precipitate is again filtered with suction,
washed free from sodium acetate with water, and dried.
The product so obtained is free from sulfur and melts a t
319-320' u.c.:
The color of the final 3-aminophthalhydrazide ap-
peared to vary according to the mode of precipitation
from almost white to quite deep yellow.
Anal. Calcd. for CsH,N30z: C, 54.22, H, 3.98, N,
23.73%; Found, C, 54.20,54.31, H, 4.48,4.62, N, 24.20,
24.02; Neut. Equiv., Calcd. 177. Found 175.

phthalhydrazide ("Luminol").-The crude 3 - nitro- DEMONSTRATION OF CHEMILUMINESCENCE OF O-

phthalhydrazide (e. g., 192.5 g. or 0.93 mol) is then AMINOPHTHALHYDRAZIDE
gradually added in small increments to 1 liter of 6 N
ammonium sulfide solution.*, Vigorous spontaneous The chemiluminescence is produced by treatment
reduction occurs and the flask may r+equire considerable of a dilute aqueous alkaline solution of "luminol"
external cooling. After all the solid nitrohydrazide with (both) hydrogen peroxide and another oxidizing
has been added the resultant solution or suspension is agent. Strong radiation is not produced in the absence
kept at the boiling point for an hour while additional of any one of the four components, although "luminol"
hydrogen sulfide is passed into the mixture. During and dilute hydrogen peroxide give a very faint glow
the reduction the original nitrohydrazide finally com- even without the other .oxidizing agent. The re-
pletely dissolves; later, however, the precipitation of action occurs only in alkaline solution, and its intensity
the resultant "luminol" begins, and these two processes and duration vary with the alkalinity. Up to a cer-
may sometimes overlap in such a way that there is al- tain point the amount of light evolved increases with
ways some solid present in the flask. After the discon- increase in alkali concentration. The mode of evolu-
tinuance of the hydrogen sulfide treatment the solution tion of light appears to vary considerably with the na-
is boiled for an hour more to complete the reaction and ture of the oxidant apparently being most satisfactory
then allowed to stand until cold. The resultant yellow with very mild oxidizing agents. For demonstration
precipitate of mixed sulfur and "luminol" is filtered purposes we much prefer the use of potassium ferri-
with suction, thoroughly washed with water, and dried. -
- t Samples of the sodium salt (dried to constant weight at
* Nore: Thissolution is that available as a stock item in most 105:) were analyzed by igniting in porcelain cmcibles, and con-
laboratories. It may be prepared by passing hydrogen sulfide vertmg the ash to sodium sulfate. Calcd. for CaHsN10sNa:
gas into 200 ml. of 15 N ammonium hydroxide (sp. gr. 0.90) in a Na, 11.55%. Found: 11.77, 11.99.
bottle immersed in running water or ice water until the gas is no f The melting points reported in this paper were taken with a
longer absorbed; 200 ml. more of conc. ammonium hydroxide 360" melting-point thermometer used in a copper block of the
is then added and the solution diluted to one liter. Berl and Kuhlman type. Cf.Ber.. 60, 811-4 (1927).
cyanide as the oxidant, and the exyution of the Glass Company, No. 985, 12" diameter by 12" high),
experiment in any of the following ways. add about 14 liters of water and float in this water a
1 . Flask Method.-Provide two &liter, long-necked piece of ice of convenient size (e. g., a piece 8-10" on a
flat-bottomed flasks and arrange to mix their contents side). Have available a stout stirrer made from a t
by pouring through a large glass funnel into a 6-liter, least 10-mm. diameter glass rod. Provide also two
flat-bottomed, long-necked flask. In one of the small flat-bottomed flasks (A and B) to contain the
smaller flasks dissolve 0.2 g. of "luminol" in 20 ml. of reactants. In flask A dissolve 1.0 g. of "luminol" in
5% sodium hydroxide and dilute to 2 liters with water. 100 ml. of 5% sodium hydroxide. In flask B place a
In the other small flask dissolve 0.5 g crystals of potas- solution of 2.5 g. of potassium ferricyanide crystals in
sium ferricyanide in water, add 20 ml. of ordinary 3% 100 ml. of 3% hydrogen peroxide. After the room is
hydrogen peroxide, and dilute to 2 liters with water. completely darkened pour these two solutions together
When hoth solutions are ready for use, grasp one flask over the surface of the cake of ice in such a manner
in each hand, have the room completely darkened, and that they mix in more or less concentrated form before
then pour the contents of the two smaller flasks simul- being diluted with the surrounding water. The pur-
taneously through the funnel into the large flask. Re- pose of the ice is not to cool the mixture but merely to
action begins as soon as the liquids mix in the funnel convince the audience that the reaction can proceed a t
and continues in the large flask for many minutes. 0°C. After the reaction mixture has diffused into the
After the initial development of light has begun, swirl main body of solution, stir the jar contents vigorously
the contents of the large flask and add a small quantity and add additional solid potassium ferricyanide (crys-
of solid crystals of potassium ferricyanide. The bril- tal or powder) or alkali, or hoth as desired.
liance increases considerably and can be still further 3. Spray Method.-A third novel way in which the
improved by gradual addition of further amounts of reaction can be made to produce beautiful eEects in-
5% sodium hydroxide solution. The concentrations volves the mixing of sprays of the separate stock solu-
here recommended are such that the light intensity first tions. For this purpose we have found most satis-
produced is fairly small, for the increased brilliance factory a commercial device widely used in textile mill
which is produced by the further addition of oxidant humidifying systems under the name N-Type Humidi-
and of alkali is very beautiful. Enough light is pro- fiers."? Two of these vest-pocket size units are held in
duced in the experiment so that as soon as it is under suitable clamps so inclined that the resultant sprays
way the demonstrator can easily locate his materials intersect some distance above the lecture table: by
in the otherwise darkened room. Since the evolution means of rubber tube connections the humidifiers are
of light continues for some minutes it is perfectly connected to a source of compressed air.$ By means of a
feasible to allow the larger flask to circulate through suitable short length of mbber tubing each unit is fur-
the audience to demonstrate that no appreciable heat ther arranged to aspirate up the corresponding solution
effect is developed. from a stock bottle placed belowit. With a little care
In order to show the dependence of light production to be certain that the spray guns are operating a t
upon the alkalinity of the solution, swirl the luminescent equal rates, and variation of the stock solutions ac-
contents of the large flask and slowly pour in dilute (6 N ) cording to the desired effect, the resultant mist may vary
hydrochloric acid until light evolution stops. If then from a barely perceptible luminous cloud to an ex-
the solution be immediately made alkaline again with ceedingly brilliant fountain reminiscent of a pyro-
the 5% alkali the luminescent reaction recommences technical display. In using this method the alkali
and continues until either the "luminol" & the oxidant concentration must be adjusted to a point where it does
is exhausted. Another way of demonstrating this not annoy the audience.
same point is to use in flask A, not the alkaline solution 4. The Cloth Method.-The very simplest and often
of "luminol," but simply a water solution of 0.25 g. of most effectiveexperiment is merely to soak an ordinary
the "luminol" sodium salt. (See above.) Under white laboratory towel in a literpf 0.1% luminol solu-
these circumstances no significant luminescence occurs tion containing 5 to 10 cc. of 5% sodium hydroxide
on mixing until after free alkali has also been supplied. and 5 to 10 cc. of 3% hydrogen peroxide, wring out
2. The Jar Method.-For demonstration to large most of the excess liquid, and then pour directly on the
audiences the following variant of the experiment may towel 2.5% potassium ferricyanide solution. The
be employed. Provide a large pyrex jar (e. g., Corning towel then glows like a live coal and on further wring-
-.
'These experiments were included in a demonstration of iue vields liquid drops which elow like fire.
chemilumine&ence shown at the Society of Arts Popular Science
Lecture on Fehruaty 15, 1931, and also at a meeting of the
Northeastern Section of the American Chemical &kty held t These can be obtained from the Parks-Cramer Company of
at the Massachusetts Institute of Technology on April 29, 1933. Boston. Mass.
Photographs of some of the effects obtained were printed in the f NOTE: If a compressed air tap is not available it is possible
Technology Rmiew, 35, 282 (May, 1933), by whose permission to utilize a commercial tank of oxygen or nitrogen, provided that
they are here reproduced. a suitable reducing valve is attached to it far regulation.