31 July 1959, Volume 130, Number 3370

CURRENT PROBLEMS IN RESEARCH

Organic Compound Syntheson the Primitive Eart

Several questions about the origin of life h;been answered, but much remains to be studi

Stanley L. Miller and Harold C. L

Since the demonstration by Pasteurthat life does not arise spontaneously at

the present time, the problem of theorigin of life has been one of determin-ing how the first forms of life arose,

from which all the present species haveevolved. This problem has received con-

siderable attention in recent years, butthere is disagreement on many points.We shall discuss the present status ofthe problem, mainly with respect to theearly chemical history of, and the syn-

thesis of organic compounds on, theprimitive earth.Many of our modern ideas on the

origin of life stem from Oparin (1), whoargued that the spontaneous generation

of the first living organism might reason-

ably have taken place if large quantitiesof organic compounds had been present

in the oceans of the primitive earth.Oparin further proposed that the atmos-

phere was reducing in character and thatorganic compounds might be synthesizedunder these conditions. This hypothesisimplied that the first organisms were

heterotrophic-that is, that they ob-tained their basic constituents from theenvironment instead of synthesizing themfrom carbon dioxide and water. Horo-witz (2) discussed this point further andoutlined how a simple heterotrophic or-

ganism could develop the ability to

synthesize various cell constituents andthereby evolve into autotrophic organ-

isms.In spite of the argument by Oparin,

31 JULY 1959

numerous attempts wei

thesize organic compo

oxidizing conditions no'

earth (3). Various sour(ing on carbon dioxideto give reduced carboncept when contaminatin,were present. The one

was the synthesis of fformaldehyde in very EH2CO molecules per iuse of 40-million-elecions from a 60-inch cyclthe simplest organic cor

deed synthesized, the yiethat this experiment c;preted to mean that itbeen possible to synthespounds nonbiologicallying conditions were prepThis experiment is imi

induced a reexaminathypothesis of the i

phere (5).

The Primitive Atmosph

Our discussion is basstion that conditions (earth were favorable ftof the organic compou

up life as we know it.sets of conditions undcompounds could haveAll these conditions are

ducing. However, befoi

SCIENCE

of conditions for the primitive earth, one

must show that reactions known to takeplace will not rapidly change the atmos-

phere to another type. The proposed set

of conditions must also be consistent with

IS the known laws for the escape of hy-drogen.

ft Cosmic dust clouds, from which the

earth is believed to have been formed,contain a great excess of hydrogen. The

ave planets Jupiter, Saturn, Uranus, andNeptune are known to have atmospheres

led. of methane and ammonia. There has not

been sufficient time for hydrogen to es-

cape from these planets, because of theirIrey lower temperatures and higher gravita-

tional fields. It is reasonable to expectthat the earth and the other minorplanets also started out with reducing

re made to syn- atmospheres and that these atmospheresiunds under the became oxidizing, due to the escape of

w present on the hydrogen.ces of energy act- The meteorites are the closest approxi-and water failed mation we have to the solid material

l compounds ex- from which the earth was formed. Theyg reducing agents are observed to be highly reduced-the

exception to this iron mostly as metallic iron with some

formic acid and ferrous sulfide, the carbon as elemental

small yield (10-7 carbon or iron carbide, and the phos-ion pair) by the phorus as phosphides.tron-volt helium The atmosphere under these reducinglotron (4). While conditions would contain some hydro-npounds were in- gen, methane, nitrogen, and ammonia;elds were so small smaller amounts of carbon dioxide and

an best be inter- carbon monoxide; and possibly small

would not have amounts of other substances such as

size organic com- higher hydrocarbons, hydrogen sulfide,as long as oxidiz- and phosphine. These substances were

sent on the earth. probably not present in equilibriumportant in that it concentrations, but compounds which

:on of Oparin's are thermodynamically very unstable

reducing atmos- in this highly reducing atmosphere-such as oxygen, oxides of nitrogen, andoxides of sulfur-could not have beenpresent in more than a few parts per

iere million. This is true of compounds which

are unstable in the present oxidizing at-

ed on the assump- mosphere of the earth, such as hydrogen,on the primitive ozone, methane, and nitrous oxide.or the production The over-all chemical change has beeninds which make the oxidation of the reducing atmosphereThere are many to the present oxidizing atmosphere. This

Ler which organic

e been produced.e more or less re-

re accepting a set

Dr. Miller is a member of the staff of the de-partment of biochemistry, College of Physiciansand Surgeons, Columbia University, New York.Dr. Urey is on the staff of the University of Cali-fornia, La Jolla, Calif.

245

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

is caused by the loss of hydrogen, whichresults in the production of nitrogen,nitrate, sulfate, free oxygen, and ferriciron. As is discussed below, many com-plex organic compounds would havebeen formed during the course of thisover-all change, thereby presenting a fa-vorable environment for the formationof life. Whether the surface carbonof the present earth was all part of theinitial atmosphere or whether it has beenescaping from the earth's interior in asomewhat reduced condition is not im-portant to the over-all picture.

Escape of Hydrogen

We have learned in recent years thatthe temperature of the high atmosphereis 2000'K or more, and there is no rea-son to suppose that the same tempera-ture was not present in the past. Onemight expect that a reducing atmospherewould be cooler than an oxidizing atmos-phere because methane and ammoniacan emit infrared radiation while thediatomic molecules, nitrogen and oxy-gen, cannot. Curtis and Goody (6) haveshown that carbon dioxide is ineffectivein emitting infrared radiation in the highatmosphere. This is due to the low effi-ciency of energy transfer from the trans-lational and rotational to the vibrationaldegrees of freedom, and it seems likelythat this would apply to methane aswell.The loss of hydrogen from the earth

is now believed to be limited by the dif-fusion of H2 to the high atmosphere,since almost all the water is frozen outbefore it reaches the high atmosphere.Urey (7) has discussed this problem andfinds that the loss is entirely due to theseeffects and not to the Jeans escapeformula.The present rate of escape is 107

atoms of hydrogen per square centimeterper second, and it is proportional to theconcentration of molecular hydrogen inthe atmosphere, which is now 10-6 atmat the earth's surface. This rate wouldresult in escape of hydrogen equivalentto 20 g of water per square centimeterin the last 4.5 x 100 years. This rate isnot sufficient to account for the oxygenin the atmosphere (230 g/cm2).

In addition, we must account for theoxidation of the carbon, ammonia, andferrous iron to their present states ofoxidation. The oxidation of the 3000 gof surface carbon per square centimeterpresent on the earth from the 0 to the+4 valence state (that is, from C or

246

H2C0 to CO2) would require the lossof 1000 g of hydrogen per square centi-meter. At the present rate of escape thiswould require 2.5 x 1012 years. In orderfor this escape to be accomplished in2.5 x 109 years (that is, between 4.5 x 109and 2.0 x 109 years ago), a pressure ofhydrogen at the surface of the earth of0.7 x 10-3 atm would have been required.In order for the nitrogen, sulfur, andiron also to be oxidized, even larger lossesand a higher pressure of hydrogen wouldhave been needed. We use a figure of1.5 x 10-3 atm for the hydrogen pressurein the primitive atmosphere.

These calculations are greatly over-simplified, since methane and other vola-tile hydrogen compounds would be de-composed in the high atmosphere andtherefore a higher concentration of hy-drogen might exist in the high atmos-phere than is indicated by surface par-tial pressures. However, the results ofthe calculation would be qualitativelythe same for hydrogen pressures differ-ent from the chosen value by an orderof magnitude.

Equilibria of Carbon Compounds

The partial pressure of CO2 in the at-mosphere is kept low by two buffer sys-tems. The first system, which is rapid, isthe absorption in the sea to form HCO3-and H2C03. The second, which is slow,is the reaction of carbon dioxide withsilicates; for example

CaSiO8+C02-+CaCOs + SiO2 Km. = 10,

The partial pressure of CO2 at sea level(3.3 x 10-4 atm) is somewhat higher thanthe equilibrium pressure (10-8 atm), butvery much lower than would be the casewithout the formation of limestones(CaCO3).The equilibrium constant at 250C in

the presence of liquid water for the re-action

C02 +4H2-> CH4+ 2H20

is 8 x 1022. Assuming that equilibriumwas attained, and using partial pressuresPC02 = 10-8 atm and PH2 = 1.5 X 10-8atm, we find that the pressure of CH,would be 4 x 103 atm. In order to havea reasonable pressure of CH4, the par-tial pressure of C02 would have to beless than 10-8 atm, and limestones wouldnot form.

Complete thermodynamic equilibriumcould not exist in a reducing atmosphere

because of the dependence of the equi-librium proportions of compounds onpressure and hence on altitude. It ismore likely that the steady-state concen-trations of CO2 and CH4 would be deter-mined not by the equilibrium at sea levelbut rather by the equilibrium at higheraltitude, where the ultraviolet lightwould provide the activation energy tobring about rapid equilibrium. Underthese conditions water would be a gas,and the equilibrium constant would be1020, so

K2s = 10' = PCH4P2H20/PCO2P'H2= XCEI4X2H2O/XCO2X'H2 P

where the X's are the mole fractions andP is the total pressure. If the surface par-tial pressures were PCH4 = 1, PC02 = 3.3 x10-4 (the present value), and PH2 = 1.5 x10-3, the X's would be equal to these par-tial pressures. We use XH20 = 10-6, whichis the present value for H20 above thetropopause. Equilibrium will be estab-lished under these conditions where P =2.5 x 10-9 atm-the present atmosphericpressure at about 180 km. It is reason-able to assume that equilibrium was es-tablished at some high altitude; there-fore, carbon dioxide and hydrogen couldboth have been present at small partialpressures and methane could have beenpresent at a moderate partial pressure ina reducing atmosphere where the pres-sure of hydrogen was 1.5 x 10-3 atm.Carbon monoxide should not have

been an important constituent of the at-mosphere, as can be seen from the fol-lowing reaction

C02+H2 ->CO+H20(1) K1 = 3.2 x 10-'Pco/Pco2 = 3.2 X 10P-4p2

Using PH2 = 1.5 x 10-3, we have the ratioPa0/PC02 =5 x 10-7, which is independ-ent of pressure. Furthermore, carbonmonoxide is a relatively reactive com-pound, and should any significant quan-tities appear in the atmosphere, it wouldreact rather rapidly to give organiccompounds, carbon dioxide and hydro-gen, and formate.Rubey (8) and Abelson (9) have

argued that the surface carbon and nitro-gen have come from the outgassing ofthe interior of the earth instead of fromthe remaining gases of the cosmic dustcloud from which the earth was formed.The carbon from the outgassing of theearth is a mixture of C02, CO, and OH4,and hydrogen may be present. While out-gassing may have been a significant proc-ess on the primitive earth, this does not

SCIENCE, VOL. 130

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

mean that the atmosphere was neces-sarily composed of CO2 and CO. Thethermodynamic considerations discussedabove would still apply. The carbondioxide would dissolve in the ocean toform bicarbonate, and CaCO3 would bedeposited, and the CO would be un-stable, as is demonstrated above.Many writers quote "authorities" in re-

gard to these questions without under-standing what is fact and what is opinion.The thermodynamic properties of C,CO, CO2, CH4, N2, NH3, 02, H20, andother similar substances are all wellknown, and the equilibrium mixtures canbe calculated for any given compositionwithout question. The only point opento argument is whether equilibrium wasapproximated or whether a nonequi-librium mixture was present. A mixtureof hydrogen and carbon monoxide orhydrogen and carbon dioxide is very un-stable at 250C, but does not explode orreact detectably in years. But wouldsuch mixtures remain in an atmospherefor millions of years subject to energeticradiation in the higlh-atmosphere? Webelieve the answer is "No." These mix-tures would react even without suchradiation in geologic times. Hydrogenand oxygen will remain together at lowtemperatures for long times without de-tectable reaction by ordinary methods.The use of radioactive tracers showsthat a reaction is proceeding at ordinarytemperatures nonetheless.The buffer systems of the ocean and

the calcium silicate-calcium carbonateequilibrium were of sufficient capacityto keep the partial pressure of the car-bon dioxide in the atmosphere at a lowvalue; hence, the principal species of car-bon in the atmosphere would have beenmethane, even though the fraction of sur-face carbon in the oxidation state of car-bon dioxide was continuously increasing.This would have been true until the pres-sure of H2 fell below about 10-6 atm. Itis likely that shortly after this, signifi-cant quantities of molecular oxygenwould have appeared in the atmosphere.

Equilibria of Nitrogen Compounds

The equilibrium concentrations of am-monia can be discussed by consideringthe reaction

1 3Nz +H2+16-NHs K,,= 7.6x1022 2

Using PH2 = 1.5 x 10-3, we have PNH3/PN2 = 0-04.Ammonia is very soluble in water and

31 JULY 1959

Table 1. Present sources of energy aver-aged over the earth.

Source Energy(cal cm7" yrl)

Total radiation from sun 260,000Ultraviolet lightX<2500A 570A < 2000A 85X < 1500 A 3.5*

Electric discharges 4 tCosmic rays 0.0015Radioactivity(to 1.0 km depth) 0.8t

Volcanoes 0. 13§* Includes the 1.9 cal cm-2 yr-1 from the Lymana at 1216 A (39). t Includes 0.9 cal cm yr-' fromlightning and about 3 cal cm-2 yr-1 due to coronadischarges from pointed objects (40). t The value,4 x 109 years ago, was 2.8 cal cm-' yr-1 (41).§ Calculated on the assumption of an emission oflava of 1 km8 (Cp = 0.25 cal/g, P = 3.0 g/cm5)per year at 1000C.

therefore would displace the above re-action toward the right, giving

2N2 + 3H2+ H+ -> NH4+2 2(NH4+)/PN21/1Pa2'1 = 8.0 x lO(H+)

which is valid for pH's less than 9. AtpH = 8 and PH2 = 1.5 x 10-3, we have

(NH4+) /PN21' = 47

which shows that most of the ammoniawould have been in the ocean instead ofin the atmosphere. The ammonia in theocean would have been largely decom-posed when the pressure of hydrogen fellbelow 10-5 atm, assuming that the pH ofthe ocean was 8, its present value. Ahigher -pH would have made the am-monia less stable; the converse is true fora lower pH.

All the oxides of nitrogen would havebeen unstable and therefore rare. Hydro-gen sulfide would have been present inthe atmosphere only as a trace constitu-ent because it would have precipitated asferrous and other sulfides. Sulfur wouldhave been reduced to hydrogen sulfideby the reaction

H2+S-H2S K=6x 104

It is evident that the calculations donot have a quantitative validity becauseof many uncertainties with respect totemperature, the processes by whichequilibrium could be approached, the-atmospheric level at which such proc-esses would be effective, and the partialpressure of hydrogen required to providethe necessary rate of escape. In view ofthese uncertainties, further calculationsare unprofitable at the present time.However, we can conclude from this dis-

cussion that a reducing atmosphere con-taining low partial pressures of hydrogenand ammonia and a moderate pressureof methane and nitrogen constitutes areasonable atmosphere for the primitiveearth. That this was the case is notproved by our arguments, but we main-tain that atmospheres containing largequantities of carbon monoxide and car-bon dioxide are not stable and cannotaccount for the loss of hydrogen fromthe earth.

Synthesis of Organic Compounds

At the present time the direct or indi-rect source of free energy for all livingorganisms is the sunlight utilized by pho-tosynthetic organisms. But before theevolution of photosynthesis other sourcesof free energy must have been used. It isof interest to consider the sources of suchfree energy as well as the origin of theappropriate chemical compounds con-taining excess free energy which suppliedthe energy for chemical evolution priorto the existence of what should be calledliving organisms, and before the evolu-tion of photosynthesis.Table 1 gives a summary of the sources

of energy in the terrestrial surface 're-gions. It is evident that sunlight is theprincipal source of energy, but only asmall fraction of this is in the wave-lengths below 2000 A which can be ab-sorbed by CH4, H20, NH3, C02, andso on. If more complex molecules areformed, the absorption can move to the2500-A region or to longer wavelengthswhere a substantial amount of energy isavailable. With the appearance of por-phyrins and other pigments, absorptionin the visible spectrum becomes possible.

Although it is probable, it is not cer-tain that the large amount of energyfrom ultraviolet light would have madethe principal contribution to the synthe-sis of organic compounds. Most of thephotochemical reactions at these lowwavelengths would have taken place inthe upper atmosphere. The compoundsso formed would have absorbed at longerwavelengths and therefore might havebeen decomposed by this ultraviolet lightbefore reaching the oceans. The ques-tion is whether the rate of decomposi-tion in the atmosphere was greater orless than the rate of transport to theoceans.

Next in importance as a source ofenergy are electric discharges, such aslightning and corona discharges frompointed objects, which occur closer to

247

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

the earth's surface and hence would haveeffected more efficient transfer to theoceans.

Cosmic-ray energy is negligible atpresent, and there is no reason to assumeit was greater in the past. The radioac-tive disintegration of uranium, thorium,and potassium was more important 4.5 x109 years ago than it is now, but still theenergy was largely expended on the in-terior of solid materials such as the rocks,and only a very small fraction of thetotal energy was available in the oceansand atmosphere. Volcanic energy is notonly small but its availability is very lim-ited. A continuous source of energy isneeded. It contributes little to the evo-lutionary process to have a lava flow inone part of the earth at one time and tohave another flow on the opposite sideof the earth years later. For a brief timeheat is available at the surface of thelava, but the surface cools and heat flowsslowly from the interior for years, mak-ing the surface slightly warm. Only avery small contribution to the evolution-ary process could be contributed by theseenergy sources.

Electric Discharges

While ultraviolet light is a greatersource of energy than electric discharges,the greatest progress in the synthesis oforganic compounds under primitive con-ditions has been made with electric dis-charges. The apparatus used by Millerin these experiments was a closed systemof glass, except for tungsten electrodes.The water is boiled in a 500-ml flaskwhich mixes the water vapor and gasesin a 5-lit. flask where the spark is located.The products of the discharge are con-densed and flow through a U-tubeback into the 500-ml flask. The first re-port (10) showed that when methane,ammonia, water, and hydrogen were sub-jected to a high-frequency spark for aweek, milligram quantities of glycine,alanine, and a-amino-n-butyric acid wereproduced.A more complete analysis (11, 12) of

the products gave the results shown inTable 2. The compounds in the tableaccount for 15 percent of the carbonadded as methane, with the yield of gly-cine alone being 2.1 percent. Indirectevidence indicated that polyhydroxylcompounds (possibly sugars) were syn-thesized. These compounds were prob-ably formed from condensations of theformaldehyde that was produced by the

electric discharge. The alanine was dem-onstrated to be racemic, as would be ex-pected in a system which contained noasymmetric reagents. It was shown thatthe syntheses were not due to bacterialcontamination. The addition of ferrousammonium sulfate did not change theresults, and the substitution of N2 forthe NH3 changed only the relative yieldsof the compounds produced.

This experiment has been repeatedand confirmed by Abelson (13), byPavlovskaya and Passynsky (14), andby Heyns, Walter, and Meyer (15).Abelson worked with various mixtures ofH2,, CH4, CO, C02, NH3, N2, H20, and02. As long as the conditions were re-ducing conditions-that is, as long aseither H2, CH4, CO, or NH3 was pres-ent in excess-amino acids were synthe-sized. The products were the same andthe yields as large in many of these mix-tures as they were with methane, am-monia, and water. If the conditions wereoxidizing, no amino acids were synthe-sized. These experiments have confirmedthe hypothesis that reducing atmospheresare required for the formation of organiccompounds in appreciable quantites.However, several of these mixtures ofgases are highly unstable. Hence the syn-thesis of amino acids in these mixturesdoes not imply that such atmosphereswere present on the primitive earth.

Heyns, Walter, and Meyer also per-formed experiments with different mix-tures of gases, with results similar toAbelson's. These workers also used GH4,NH3, H20, and H2S. They obtained am-monium thiocyanate, thiourea, and thio-acetamide as well as compounds formedwhen H2S was absent.The mechanism of synthesis of the

amino acids is of interest if we are toextrapolate the results in these simplesystems to the primitive earth. Two al-ternative proposals were made for thesynthesis of the amino and hydroxy acidsin the spark discharge system. (i) Alde-hydes and hydrogen cyanide are syn-thesized in the gas phase by the spark.These aldehydes and the hydrogen cya-nide react in the aqueous phase of thesystem to give amino and hydroxy ni-triles, which are hydrolyzed to aminoand hydroxy acids. This mechanism isessentially a Strecker synthesis. (ii) Theamino and hydroxy acids are synthesizedin the gas phase from the ions and radi-cals that are produced in the electricdischarge.

It was shown that most, if not all, ofthe amino acids were synthesized accord-

ing to the first hypothesis, since the rateof production of aldehydes and hydrogencyanide by the spark and the rate ofhydrolysis of the amino nitriles were suf-ficient to account for the total yield ofamino acids (12).

This mechanism accounts for the factthat most of the amino acids werea-amino acids, the ones which occur inproteins. The 13-alanine was formed notby this mechanism but probably by theaddition of ammonia to acrylonitrile (oracrylamide or acrylic acid), followed byhydrolysis to 13-alanine.The experiments on the mechanism

of the electric discharge synthesis ofamino acids indicate that a special setof conditions or type of electric dis-charge is not required to obtain aminoacids. Any process or combination ofprocesses that yielded both aldehydesand hydrogen cyanide would have con-tributed to the amount of a-amino acidsin the oceans of the primitive earth.Therefore, whether the aldehydes andhydrogen cyanide came from ultravioletlight or from electric discharges is nota fundamental question, since both proc-esses would have contributed to thea-amino acid content. It may be thatelectric discharges were the principalsource of hydrogen cyanide and thatultraviolet light was the principal sourceof aldehydes, and that the two processescomplemented each other.

Ultraviolet Light

It is clear from Table 1 that the great-est source of energy would be ultravioletlight. The effective wavelengths wouldbe CH4 < 1450 A, H20 < 1850 A,NH3 < 2250 A, CO < 1545 A, C02 <1690 A, N2 <1100 A, and H2< 900 A.It is more difficult to work with ultra-violet light than with electric dischargesbecause of the small wavelengths in-volved.The action of the 1849-A Hg line on

a mixture of methane, ammonia, water,and hydrogen produced only a verysmall yield of amino acids (16). OnlyNH3 and H20 absorb at this wavelength,but apparently the radical reactionsformed active carbon intermediates. Thelimiting factor seemed to be the syn-thesis of hydrogen cyanide. Groth (17)found that no amino acids were pro-duced by the 1849-A line of mercurywith a mixture of methane, ammonia,and water, but that amines and aminoacids were formed when the 1470-A and

SCIENCE, VOL. 130248

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

1295-A lines of xenon were used. The1849-A line produced amines and aminoacids with a mixture of ethane, ammo-nia, and water. The mechanism of thissynthesis was not determined. Terenin(18) has also obtained amino acids bythe action of the xenon lines on methane,ammonia, and water.We can expect that a considerable

amount of ultraviolet light of wave-lengths greater than 2000 A would beabsorbed in the oceans, even thoughthere would be considerable absorptionof this radiation by the small quantitiesof organic compounds in the atmos-phere. Only a few experiments havebeen performed which simulate theseconditions.

In a most promising experiment,Ellenbogen (19) used a suspension offerrous sulfide in aqueous ammoniumchloride through which methane wasbubbled. The action of ultraviolet lightfrom a mercury lamp gave small quan-tities of a substance with peptide fre-quencies in the infrared. Paper chroma-tography of a hydrolyzate of thissubstance gave a number of spots withNinhydrin, of which phenylalanine,methionine, and valine were tentativelyidentified.Bahadur (20) has reported the syn-

thesis of serine, aspartic acid, asparagine,and several other amino acids by theaction of sunlight on paraformaldehydesolutions containing ferric chloride andnitrate or ammonia. Pavlovskaya andPassynsky (21) have also synthesized anumber of amino acids by the action ofultraviolet light on a 2.5-percent solu-tion of formaldehyde containing ammo-nium chloride or nitrate. These highconcentrations of formaldehyde wouldnot have occurred on the primitive earth.It would be interesting to see if similarresults could be obtained with 10-4M or10-5M formaldehyde. This type of ex-periment deserves further investigation.

Radioactivity and Cosmic Rays

Because of the small amount of energyavailable, it is highly unlikely that high-energy radiation could have been veryimportant in the synthesis of organiccompounds on the primitive earth. How-ever, a good deal of work has been donein which this type of energy has beenused, and some of it has been interpretedas bearing on the problem of the originof life.

Dose and Rajewsky (22) produced

Table 2. Yields from sparking a mixtureof CH4, NHs, HO, and H2; 710 mg ofcarbon was added as CH,.

Compound

GlycineGlycolic acidSarcosineAlanineLactic acidN-Methylalaninea-Amino-n-butyric acida-Aminoisobutyric acida-Hydroxybutyric acid

0-AlanineSuccinic acidAspartic acidGlutamic acidIminodiacetic acidIminoacetic-propionic acid

Formic acidAcetic acidPropionic acidUreaN-Methyl urea

Yield[moles (x 10')]

63.56.5.

34.31.

1.5.0.15.

15.4.0.40.65.51.5

233.15.13.2.01.5

amines and amino acids through the ac-tion of x-rays on various mixtures ofCH4, C02, NH8, N2, H20, and H2. Asmall yield of amino acids was obtainedthrough the action of 2 Mev electronson a mixture of CH4, NH3, and H20(23).The formation of formic acid and

formaldehyde from carbon dioxide andwater by 40 Mev helium ions was men-tioned previously. These experimentswere extended by using aqueous formicacid (24). The yield per ion pair wasonly 6 x 10-4 for formaldehyde and 0.03for oxalic acid. Higher yields of oxalicacid were obtained from Ca(HCO3)2and NH4HCO3 by Hasselstrom andHenry (25). The helium ion irradiationof aqueous acetic acid solutions gavesuccinic and tricarbolic acid along withsome malonic, malic, and citric acids(26).The irradiation of 0.1- and 0.25-per-

cent aqueous ammonium acetate by 2Mev electrons gave glycine and asparticacid (27). The yields were very small.Massive doses of gamma rays on solidammonium carbonate yielded formicacid and very small quantities of glycineand possibly some alanine (28).

The. concentrations of carbon com-pounds and the dose rates used in theseexperiments are, in all probability, verymuch larger than could be expected onthe primitive earth, and the productsand yields may depend markedly on

these factors, as well as on the effect ofradical scavengers such as HS- andFe2+. It is difficult to exclude highenergyradiations entirely, but if one is to makeany interpretations from laboratorywork, the experiments should be per-formed with much lower dose rates andconcentrations of carbon sources.

Thermal Energy

The older theories of the formation ofthe earth involved a molten earth duringits formation and early stages. Thesetheories have been largely abandoned,since the available evidence indicatesthat the solar system was formed froma cold cloud of cosmic dust. The mecha-nisms for heating the earth are thegravitational energy released during thecondensation of the dust to form theearth and the energy released from thedecay of the radioactive elements. It isnot known whether the earth was moltenat any period during its formation, butit is clear that the crust of the earthwould not have remained molten for anylength of time.

Studies on the concentration of some

elements in the crust of the earth indi-cate that the temperature was less than1500C during this lengthy fractionation,and that it was probably close to presentterrestrial temperatures (29).Fox (30) has maintained that organic

compounds were synthesized on the earthby heat. When heated to 1500C, malicacid and urea were converted to asparticacid and ureidosuccinic acid, and some

of the aspartic acid was decarboxylatedto a- and 0-alanine. The difficulty withthese experiments is the source of themalic acid and urea on the primitiveearth-a question not discussed by Fox.Fox has also synthesized peptides by thewell-known reaction (31) of heatingamino acids at 1500 to 1800C, and theyield of peptides has been increased byusing an excess of aspartic or glutamicacid (32). There is a difficulty connectedwith heating amino acids and other or-

ganic compounds to high temperatures.Geological conditions can heat aminoacids to temperatures above 10OCC over

long periods of time, but it is not likelythat this could occur over short periods.Abelson (33) has shown that alanine,one of the more stable amino acids, de-carboxylates to methylamine and carbondioxide. The mean life of alanine is .1011years at 250C but only 30 years at

1500C. Therefore, any extensive heating

31 JULY 1959

_

1

249

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

of amino acids will result in their de-struction, and the same is true for mostorganic compounds. In the light of this,and since the surface of the primitiveearth was probably cool, it is difficult tosee how the processes advocated by Foxcould have been important in the syn-thesis of organic compounds.

Surface Reactions, OrganicPhosphates, and Porphyrins

It is likely that many reactions werecatalyzed by adsorption on clay and min-eral surfaces. An example is the poly-merization of aminoacetonitrile to gly-cine peptides in the presence of acidclays, by Akabori and his co-workers(34). Formaldehyde and acetaldehydewere shown to react with polyglycine ad-sorbed on kaolinite to give serine andthreonine peptides. This field offersmany possibilities for research.

Gulick (35) has pointed out that thesynthesis of organic phosphates presentsa difficult problem because phosphateprecipitates as calcium and other phos-phates under present earth conditions,and that the scarcity of phosphate oftenlimits the growth of plants, especially inthe oceans. He proposes that the pres-ence of hypophosphites, which are moresoluble, -would account for higher con-centrations of phosphorus compoundswhen the atmosphere was reducing.Thermodynamic calculations show thatall lower oxidation states of phosphorusare unstable under the pressures of hy-drogen assumed in this article. It is pos-sible that stronger reducing agents thanhydrogen reduced the phosphate or thatsome process other than reduction solu-bilized the calcium phosphate. Thisproblem deserves careful attention.The synthesis of porphyrins is consid-

ered by many authors to be a necessarystep for the origin of life. Porphyrins arenot necessary for living processes if theorganism obtains its energy requirementsfrom fermentation of sugars or otherenergy-yielding organic reactions. Ac-cording to the heterotrophic theory ofthe origin of life, the first organismswould derive their energy requirementsfrom fermentations. The metabolism ofsulfate, iron, N2, hydrogen, and oxygenappears to require porphyrins as well asphotosynthesis. Therefore, porphyrinsprobably would have to be synthesizedbefore free energy could be derived fromthese compounds. While porphyrins mayhave been present in the environment

before life arose, this is apparently nota necessity, and porphyrins may havearisen during the evolution of primitiveorganisms.

Intermediate Stages inChemical Evolution

The major problems remaining foran understanding of the origin of life are(i) the synthesis of peptides, (ii) thesynthesis of purines and pyrimidines,(iii) a mechanism by which "high-energy" phosphate or other types ofbonds could be synthesized continuously,(iv) the synthesis of nucleotides andpolynucleotides, (v) the synthesis ofpolypeptides with catalytic activity (en-zymes), and (vi) the development ofpolynucleotides and the associated en-zymes which are capable of self-dupli-cation.

This list of problems is based on theassumption that the first living organismswere similar in chemical compositionand metabolism to the simplest livingorganisms still on the earth. That thismay not be so is obvious, but the hy-pothesis of similarity allows us to per-form experiments to test it. The surpris-ingly large yields of aliphatic, hydroxy,and amino acids-a-amino acids ratherthan the other isomers-in the electric-discharge experiments, plus the argu-ments that such syntheses would havebeen effective on the primitive earth,offer support for this hypothesis. Fur-ther support can be obtained by demon-strating mechanisms by which othertypes of biologically important com-pounds could be synthesized.

Oparin (1) does not view the firstorganism as a polynucleotide capable ofself-duplication but, rather, as a coacer-vate colloid which accumulates proteinsand other compounds from the environ-ment, grows in size, and then splits intotwo or more fragments, which repeat theprocess. The coacervate would presum-ably develop the ability to split into frag-ments which are very similar in com-position and structure, and eventually agenetic apparatus would be incorporatedwhich would make very accurate dupli-cates.

These two hypotheses for the steps inthe formation of the first living organismdiffer mainly in whether the duplicationfirst involved the relatively accurate du-plication of nucleic acids, followed bythe development of cytoplasm duplica-tion, or whether the steps occurred in

the reverse order. Other sequences couldbe enumerated, but it is far too early todiscuss profitably the exact nature of thefirst living organism.

It was probably necessary for theprimitive organisms to concentrate or-ganic and inorganic nutrients from theirenvironment. This could be accom-plished by means of a membrane or byabsorption on rocks or clays (36). Thedevelopment of optical activity in livingorganisms is another important problem.This has been discussed by many authorsand is not taken up here.

Life on Other Planets

Life as we know it-and we know ofno other variety of life than that existingon the earth-requires the presence ofwater for its chemical processes. Weknow enough about the chemistry ofother systems, such as those of silicon,ammonia, and hydrogen fluoride, torealize that no highly complex systemof chemical reactions similar to thatwhich we call "living" would be pos-sible in such media. Also, much livingmatter exists and grows actively on theearth in the absence of oxygen, so oxy-gen is not necessary for life, althoughthe contrary is often stated. Moreover,the protecting layer of ozone in theearth's atmosphere is not necessary forlife, since ultraviolet light does not pene-trate deeply into natural waters and alsobecause many carbon compounds capa-ble of absorbing the ultraviolet lightwould be present in a reducing atmos-phere.

It is possible for life to exist on theearth and to grow actively at tempera-tures ranging from 00C, or perhaps alittle lower, to about 700C. It seemslikely that if hot springs were not so tem-

porary, many plants and possibly ani-mals would evolve which could live insuch temperatures. Plants are able to

produce and accumulate substanceswhich lower the freezing point of water,and hence they can live at temperaturesbelow 00C. At much lower temperaturesthe reactions would probably be tooslow to proceed in reasonable periods oftime. At temperatures much above1200C, reaction velocities would prob-ably be so great that the nicely balancedreactions characteristic of living thingswould be impossible. In addition, it isdoubtful whether the organic polymersnecessary for living organisms would bestable much above 1200C; this is prob-

SCIENCE, VOL. 130250

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

ably true even when allowance is madefor the amazing stability of the enzymesof thermophilic bacteria and algae.Only Mars, Earth, and Venus conform

to the general requirements so far astemperatures are concerned. Mars isknown to be very cold and Venus maybe too hot. Observations of the black-body emission of radio waves from Venusindicate surface temperatures of 2900 to350PC (37). The clouds of Venus havethe polarization of water droplets.Clearing of the clouds occurs, and thisindicates that the clouds are composedof some volatile substance, for nonvola-tile dust could hardly settle out locally.However, no infrared bands of waterhave been observed. It is possible thatthis is due to a very dry, high atmos-phere, such as is characteristic of theearth, and to a cloud level that rises tovery near the tropopause, so that thereis little water vapor above the reflectinglayer.Mars is known to be very cold, with

surface temperatures of + 30'C to- 60'C during the day. The colors ofMars have been observed for many yearsby many people. The planet exhibitsseasonal changes in color-green or blu-ish in the spring and brown and reddishin the autumn. Sinton (38) has observedan absorption at 3.5 [t in the reflectedlight -of Mars. This corresponds to theC-H stretching frequency of most or-ganic compounds, but many inorganiccompounds have absorptions at thiswavelength. The changing colors of Marsand the 3.5 g absorption are the bestevidence, however poor it may be, forthe existence of life on the planet. Onething that can be stated with confidenceis that if life exists there, then liquidwater must have been present on theplanet in the past, since it is difficult to

believe that life could have evolved in itsabsence. If this was so, water must haveescaped from the planet, as very littlewater remains there now and no liquidwater has been observed. Hence, oxygenatoms must escape from the planet.This is possible if the high atmospherehas a temperature of 2000K, and thismay well be the case in view of the hightemperatures in the high atmosphere ofthe earth.

Surely one of the most marvelous featsof 20th-century science would be thefirm proof that life exists on anotherplanet. All the projected space flightsand the high costs of such developmentswould be fully justified if they were ableto establish the existence of life oneither Mars or Venus. In that case, thethesis that life develops spontaneouslywhen the conditions are favorable wouldbe far more firmly established, and ourwhole view of the problem of the originof life would be confirmed (42).

References

1. A. I. Oparin, The Origin of Life (Macmillan,New York, 1938; Academic Press, New York,ed. 3, 1957).

2. N. H. Horowitz, Proc. Natl. Acad. Sci. U.S.31, 153 (1945).

3. E. I. Rabinowitch, Photosynthesis (Intersci-ence, New York, 1945), vol. I, p. 81.

4. W. M. Garrison, D. C. Morrison, J. G. Ham-ilton, A. A. Benson, M. Calvin, Science 114,416 (1951).

5. H. C. Urey, Proc. Nati. Acad. Sci. U.S. 38,351 (1952); The Planets (Yale Univ. Press,New Haven, Conn., 1952), p. 149.

6. A. R. Curtis and R. M. Goody, Proc. Roy.Soc. (London) 236A, 193 (1956).

7. H. C. Urey, in Handbuch der Physik, S.Fluigge, Ed. (Springer, Berlin, 1958), vol. 52.

8. W. W. Rubey, Geol. Soc. Am. Spec. PaperNo. 62 (1955), p. 631.

9. P. H. Abelson, Carnegie Inst. Wash. YearBook No. 56 (1957), p. 179.

10. S. L. Miller, Science 117, 528 (1953).11. , J. Am. Chem. Soc. 77, 2351 (1955).12. , Biochim. et Biophys. Acta 23, 480

(1957).13. P. H. Abelson, Science 124, 935 (1956); Car-

negie Inst. Wash. Year Book No. 55 (1956),p. 171.

14. T. E. Pavlovskaya and A. G. Passynsky, Re-ports of the Moscow Symposium on the Originof Life (Aug. 1957), p. 98.

15. K. Heyns, W. Walter, E. Meyer, Naturwiss-enschaften 44, 385 (1957).

16. S. L. Miller, Ann. N. Y. Acad. Sci. 69, 260(1957).

17. W. Groth, Angew. Chem. 69, 681 (1957);and H. von Weyssenhoff, Naturwiss-

enschaften 44, 510 (1957).18. A. N. Terenin, in Reports of the Moscow

Symposium on the Origin of Life (Aug. 1957),p. 97.

19. E. Ellenbogen, Abstr. Am. Chem. Soc. Meet-ing, Chicago (1958), p. 47C.

20. K. Bahadur, Nature 173, 1141 (1954); inReports of the Moscow Symposium on theOrigin of Life (Aug. 1957), p. 86; Nature182, 1668 (1958).

21. T. E. Pavlovskaya and A. G. Passynsky, In-tern. Congr. Biochem. 4th Congr. Abstr.Communs. (1958), p. 12.

22. K. Dose and B. Rajewsky, Biochim. et Bio-phys. Acta 25, 225 (1957).

23. S. L. Miller, unpublished experiments.24. W. M. Garrison et al., 1. Am. Chem. Soc. 74,

4216 (1952).25. T. Hasselstrom and M. C. Henry, Science

123, 1038 (1956).26. W. M. Garrison et al., 1. Am. Chem. Soc.

75, 2459 (1953).27. T. Hasselstrom, M. C. Henry, B. Murr, Sci-

ence 125, 350 (1957).28. R. Paschke, R. Chang, D. Young, ibid. 125,

881 (1957).29. H. C. Urey, Proc. Roy. Soc. (London) 219A,

281 (1953).30. S. W. Fox, J. E. Johnson, A. Vegotsky, Sci-

ence 124, 923 (1956); Ann. N.Y. Acad. Sci.69, 328 (1957); 1. Chem. Educ. 34, 472(1957).

31. For a review, see E. Katchalski, Advances inProtein Chem. 6, 123 (1951).

32. S. W. Fox et al., 1. Am. Chem. Soc. 80, 2694,3361 (1958); and K. Harada, Science128, 1214 (1958).

33. P. H. Abelson, Ann. N.Y. Acad. Sci. 69, 276(1957).

34. S. Akabori, in Reports of the Moscow Sym-posium on the Origin of Life (Aug. 1957),p. 117; Bul. Chem. Soc. Japan 29, 608 (1956).

35. A. Gulick, Am. Scientist 43, 479 (1955); An".N.Y. Acad. Sci. 69, 309 (1957).

36. J. D. Bernal, Proc. Phys. Soc. (London) 62A,537 (1949); ibid. 62B, 597 (1949); The Physi-cal Basis of Life (Routledge and Kegan Paul,London, 1951).

37. G. H. Mayer, T. P. McCullough, R. M.Sloanaker, Astrophys. 1. 127, 1 (1958).

38. W. Sinton, ibid. 126, 231 (1957).39. W. A. Rense, Phys. Rev. 91, 299 (1953).40. B. Schonland, Atmospheric Electricity (Me-

thuen, London, 1953), pp. 42, 63.41. E. Bullard, in The Earth as a Planet, G. P.

Kuiper, Ed. (Univ. of Chicago Press, Chi-cago, 1954), p. 110.

42. One of the authors (S. M.) is supported by agrant from the National Science Foundation.

31 JULY 1959 251

on

June

18,

200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from