The Miller-Urey Experiment

J. H. John Peet BSc, MSc, PhD, CChem, FRSC
Charles Darwin recognised that a basic problem of his theory of evolution was to produce life itself. In a letter to Joseph Hooker in 1871, he wrote: … if (and oh! what a big if!) we could conceive [of] some warm little pond, with all sorts of ammonia and phosphoric salts, light, heat, electricity etc. present, that a protein compound was chemically formed ready to undergo still more complex changes, at the present day such matter would be instantly absorbed, which would not have been the case before living creatures were found. 1953 was a landmark year for scientists researching an evolutionary explanation for the appearance of life.  Stanley Miller reported that he had conducted an experiment which replicated the primeval conditions on Earth and had produced the chemicals that were essential for life to begin. Extravagant claims were made by some, even that he had synthesised life itself! Over fifty years have passed and we can make a sober and scientific assessment of that experiment and others like it.  Let’s consider firstly what he did and got. 

Figure 1. Miller's Apparatus

He assembled a closed system (figure 1) into which he pumped a mixture of gases (methane, ammonia and hydrogen).  There was a flask of boiling water in order to add water vapour to the mixture and the gases were circulated around the apparatus. The gaseous mixture was subjected to a high voltage electrical discharge and then passed through a condenser to cool it down before going through a “trap” cooled in ice to collect any liquid products. Unchanged material was cycled through the apparatus repeatedly to maximise the yield.

This was a good chemical experiment but was it relevant to the objective?  Let’s look closely at the detail.

Irrelevant atmosphere

Firstly, consider the gaseous mixture.  This was supposed to replicate the primeval atmosphere on the Earth.  You will notice that there is an absence of oxygen and nitrogen which are the main elemental constituents of our present environment.  The problem recognised by Miller and his colleagues was that oxygen would destroy any organic material in the experiment and certainly in the period of time they allocated to the early period on the planet.  For example, when we die, we decay.  A part of that process (in addition to bacterial action) is the oxidation of the organic materials in the body, generating carbon dioxide and water.

Consequently, evolutionary scientists have proposed that the early Earth had no elemental oxygen. It would, in fact, be a “reducing atmosphere”, the opposite of the modern oxidising one.  (They go on to hypothesise that this would gradually change as primitive life produced oxygen through processes such as photosynthesis).  However, the evidence for this reducing atmosphere is very tenuous. Increasingly we are finding from geological and palaeontological research that an oxygen-based atmosphere must have existed from the earliest times.
But, we can ask whether the atmosphere proposed by Miller was likely to be stable.  Abelson reports that the ammonia in the atmosphere would have decomposed within 30,000 years: it is inherently unstable, decomposing into nitrogen and hydrogen. Also, much of it would dissolve out of the atmosphere due to its great solubility in water. Methane would only have lasted for about 1% of the time required for the appearance of life by this process, according to Shimzu. Brinkman has shown that even the water vapour would have been broken down due to the sun’s radiation.  The trouble is that we think of these gases as stable – indeed they are relative to our lifetime, but not on the evolutionary timescales.  And hydrogen?  We know that hydrogen does not exist as an element on this planet: it escapes into space very rapidly due to its low density.
Various other alternative atmospheres have been proposed, but these either don’t generate the materials required or are faced with similar problems to those mentioned for Miller’s work.

Irrelevant conditions

So, the atmosphere used was irrelevant. In fact, the experimental conditions are also irrelevant. We have to ask how we could get the circulatory system necessary for the build up of the quantity of chemicals.  Where would the cooling systems have been that are needed to isolate the products and protect them from further reaction?  What was the source of energy?  Miller used electrical discharges and compared them to lightning. But the intensities required would be far greater than those experienced today.  Others have argued that the sun provides large amounts of continuous energy (which is used today in photosynthesis, for example). This, they claim, over extended periods of time could synthesise the required chemicals. But this overlooks something important. 

Basically, this argument is saying,

Raw materials  +  Energy  ⇒ Life molecules.

But this omits an important factor. In any process that leads to complexity there must be an information source. For example, in photosynthesis a complex system involving chlorophyll captures energy from the sun and uses it to build molecules from raw materials. Can you imagine shaking a flask containing the basic materials for the production of life (amino acids, sugars, nucleotides, fatty acids, etc.) and continuing to do so until life appeared?  That is essentially what we are requiring in an undirected synthesis of this type. “Shake it more vigorously and for longer” is not an encouraging command!

Low yield

So, what about the results of Miller’s experiment?  He obtained a “soup” that contained around 9 amino acids, 2% of the simplest, glycine and alanine, and traces of 7 others. (A number of other organic compounds were produced in small quantities but they have no significance in the origin of life scenario and could even hinder further progress by reacting with the amino acids).  Amino acids have the general formula:


Where COOH is an acidic group, NH2 is the amino group and R represents a variety of organic groups that can be inserted. These amino acids (20 different ones occur in most living organisms) can be joined through their acidic and amino groups to give proteins. These in turn are fundamental to the structure of living organisms (muscles, skin, hair, etc.) and to their chemical activity (through enzymes). Chemically, this group of chemicals in living organisms are the simplest to produce. Attempts to produce other materials of this sort have been less successful.

You can imagine, therefore, the excitement with which Miller’s work was received. But, even as he acknowledges now, it proved to have limited relevance to the problem.  The yields obtained under these conditions were very small.  This is not surprising if we consider the physical chemistry of the reaction.

Let’s consider glycine, the simplest amino acid (R is a hydrogen atom). According to DE Hull, the synthesis of glycine can be represented by the equation:

2 CH4 + NH3 + 2 H2O   H2N.CH2.COOH + 5 H2

You can write the equilibrium constant expression for this:

                        K =         p(gly).p(H2)5

We can calculate the value of the constant from thermodynamic information and it is

                        K =      2 x 10-40

This would give (at proposed primeval pressures) a concentration of 10-27 molecule in 10,000 litres)!  Not a good yield.  More complex amino acids would give lower yields still. The only way to shift the equilibrium in favour of an increased yield would be to remove the products as they are formed. As Miller found, this still gives a very low yield, but without it the products are destroyed in the recycling process.
This means that the probability of the amino acid molecules coming into contact and forming a protein is negligible: too few in too large a volume of water. Of course, it is not only an equilibrium problem but a kinetic one: the time taken to find another molecule would be too great to produce the materials needed.
Miller’s experiment did produce the amino acids, but only by continuously circulating the reaction mixture and isolating products as they were formed. The quantities were still tiny and not in the same proportions as found in nature. 
One of the causes of the low yield has been identified by Peltzer who worked with Miller. As the amino acids were formed they reacted with reducing sugars in the Maillard reaction, forming a brown tar around Miller's apparatus. Ultimately, Miller was producing large compounds called mellanoids, with amino acids as an intermediate product.

Wrong forms of amino acid

But there is a more fundamental problem with this scenario which can easily be overlooked. Amino acids, like all chemicals, are three-dimensional structures.  The arrangement of the central carbon atom is tetrahedral (figure 2).  In the diagram you will see two versions of this.  Unless you are used to studying these sorts of arrangements, you will think they are the same; it would seem that you could just rotate one to get the other.  This is not, in fact, the case.  We compare them to our hands: right-handed and left-handed. A left-handed glove will not fit on a right hand, for example.

Figure 2. Right and left-handed molecules

Does this matter?  The answer is a very loud “Yes!”.  In nature, we only have left-handed (levo) amino acids.  (Glycine mentioned above is an exception; it does not have two forms – make a model and you will see why!).  Miller’s experiment gives a mixture of both forms but nature requires the levo form only. Again, does it matter?  Functional proteins cannot contain more than traces of right handed (dextro) amino acids. Right-handed forms (dextro) can have very different, even fatal, effects in some circumstances. 
It is not a simple process to separate them and there is no natural system that can do so.  In fact, L-amino acids have a tendency with age to undergo a chemical inversion to the D-form. This is called racemization. (This again gives a headache to the evolutionists: if amino acids could have been synthesised in a pure L-form, within a short time they would have racemized to give a 50:50 mixture of the two forms!). This racemization occurs in nature and can cause severe problems. For example, teeth and eye proteins racemize with age and so affect their health; Alzheimer’s disease also may be caused by racemization of a protein.

This structural distinction is a property that occurs widely in organic chemistry. For example, from non-protein substances we can observe the effect.  Limonene occurs in these two forms: one gives the smell of lemons and the other of oranges!  More seriously, the drug thalidomide was produced to aid pregnant mothers in order to combat “morning sickness”.  It was very effective but sadly serious deformities occurred in many babies.  The reason was that the commercial drug was sold in a mixture of the handed forms.  
A similar problem arises with naturally occurring sugars: they are found in the dextro­form, not the levo one as in amino acids!

So, we see in this first stage experiment that we have irrelevant conditions, a wrong atmosphere, low yields of chemicals in wrong proportions and a serious structural problem. If other compounds necessary to life were present (indeed other compounds at all), we would also have the problem of competitive reactions effectively lowering the yields even further.

The problem of building a protein

We can see that the process of chemical evolution has failed at the first hurdle.  But, in order to get a complete picture, let’s assume the problem can be solved (and no-one has done that yet!).  We now need the amino acids to join together (polymerise) to form proteins.

Figure 3. Primary, secondary, tertiary and quaternary protein structure

Here again we have a string of problems.  Let’s start with the basic chemical one.  To link the small molecules together, we need to remove water molecules between adjacent amino acid molecules.  In the case of two amino acid units, it looks like this:

            HOOC -  CHR1 – NH2     +     HOOC – CHR2 - NH2

        HOOC - CH R1 – NH- OC - CHR2 - NH2   +  H2O

This is an equilibrium reaction, which does not occur spontaneously, and the yield of protein depends on removing the water.  But, the scenario pictured by evolutionary scientists is one that occurs in a pool of water!  Not a promising start!

Since it is an equilibrium system, we can apply equilibrium calculations to it.  Consider a protein of just 100 amino acids (rather a small one in terms of naturally occurring materials),
                        K =     [protein]  =  10-36

If all the atmospheric nitrogen was used to produce the maximum amount of protein, the concentration of protein would be about 10-106  And that is for just one protein – we need hundreds of different ones!

Miller and his colleague Orgel, summed up the position themselves:  “Another way of examining this problem is by asking whether there are places on the earth today where we could drop, say, 10 grams of a mixture of amino acids and obtain a significant yield of polypeptides … We cannot think of a single such place.”  (Polypeptides are small proteins).
To form these proteins so quickly in the cell, we need accelerators, called enzymes, to enable the reactions to occur rapidly (before the cell dies through lack of a protein!). These enzymes enable reactions to occur in milliseconds. Without them, the reactions can take millions, even trillions, of years. The problem is that enzymes are proteins themselves, and they need enzymes to  form themselves!
Consider a cell containing just 124 proteins. Professor Morowitz has calculated that the chance of all these forming without information input is 1 in 10100,000,000. One of the smallest known genomes is that of Mycoplasma genitalium which manufactures about 600 proteins, so what are the chances of that happening without intelligent input? Humans have about 100,000 proteins.

But the problems are only just beginning!

Another big hurdle lies in the structure of the protein molecule.  We have seen that it has to be formed by the joining together of these twenty amino acids.  For example, the sequence might begin something like this:

            Lys – Ala – His – Gly – Lys –Lys – Val – Leu – Gly – Ala -

where the three letters are shorthand for specific amino acids.  “Gly” stands for glycine, the simplest amino acid.  This chain then twists into a helix. The sequence is called the primary structure and the helix is the secondary structure.  Other than the fact that the helical structure can twist in one of two directions (“clockwise” or “anticlockwise”) and it only takes one of these forms in nature, there is no real problem in this second step.

The helix then folds over on itself to give a more complex structure (tertiary structure).  This can be imagined most easily by thinking of a floppy spring. If it is released, it will fold over on itself. With the protein chain, there are estimated to be some 100 million different ways it can fold.  BUT, only one of these is biologically active.  How does it achieve the correct conformation?

The correct tertiary structure for each protein is, in turn, dependent on the primary structure: if the amino acid sequence is changed, the structure will fold incorrectly and lose some or all of its activity. An example of this is in haemoglobin. This is a large molecule with protein side chains. It occurs in our red blood cells and transports oxygen around the body. In one example of the effect of a change in the amino acid sequence, just one change can convert the cell from the very efficient structure we have to a very fragile cell which results in sickle cell anaemia.  A person suffering from this deficiency will die young unless they get regular blood transfusions.

A super-computer (“Blue Gene”) is being constructed in order for it to work out what is the best conformation of the protein chain in such structures.  When it is complete, it will take a year to do all the calculations.  The cell does this in less than a second!

To form these proteins so quickly in the cell, we need accelerators, called enzymes, to enable the reactions to occur rapidly (before the cell dies through a lack of protein!). These enzymes enable reactions to occur in milliseconds.  Without them, the reactions can take millions, even trillions of years (100 times the claimed age of the universe!).  The problem is that enzymes are proteins themselves – and they need enzymes to form themselves!

Consider a cell containing just 124 proteins.  Prof. Morowitz has calculated that the chance of all these forming without information input is 1 in 10100,000,000.  The smallest genome is in the Mycoplasma genitalium which manufactures about 600 proteins, so what are the chances of that happening without intelligent input?  Humans have 100,000 proteins!

Other chemicals needed for life
As we examine the other types of chemical in the cell (and they are all essential!), we find the problems tend to become greater than those we have outlined for the proteins. For example, complex carbohydrates are formed from sugar molecules. As with the amino acid to protein conversion, the formation of large carbohydrates from sugars is not spontaneous. The probability of their formation is such that there would be only 1 molecule in 1030 times the volume of the universe!  And, sugar molecules are only right-handed in nature.

Most scientists acknowledge that these are big problems and that an evolutionary approach has not offered a reasonable scientific explanation for the origin of the molecules needed for a living cell.  We have examined the work of Miller.  Obviously other scientists have been involved and have suggested alternative approaches, but these have not overcome the difficulties.

RNA World?
Proteins can act as catalysts for chemical reactions but cannot replicate without DNA. However, a slightly simpler molecule, RNA can replicate itself and sometimes can also act as a catalyst. Therefore some scientists have suggested that RNA was the first molecule of life formed.  If this could be formed, then it could possibly initiate some of the essential functions required in the cell until the modern structures could evolve.  There has been no experimental indication of the formation of either RNA or DNA in a Miller-type synthesis.
Prof. Orgel, a leading scientist in this field of research calls it “the prebiotic chemist’s nightmare”.  The RNA molecule may be simpler than DNA, but it is still complex and involves a chemical structure that does not form spontaneously. According to Dr Cairns-Smith, it requires 14 major hurdles with 10 steps in each, giving a probabilility of 1 in 10109 for their successful formation. The first “ribo-organism” would need all the cell’s metabolic functions in order to survive and the is not evidence that such a range of functions is possible for RNA.

Could clays help?
Cairns-Smith considered an alternative approach. He considered that naturally occurring clays might provide a basis for the synthesis of these chemicals. There are irregularities in the structures of clays and the process of crystallisation enables the replication of these structures. Crystals can also fracture producing smaller units of the same symmetry. We do know that clays can catalyse some chemical reactions, so, he proposed, perhaps these irregularities could be the basis on which specific organic reactions might develop, resulting in a primitive cell.  He considered that these crystal structures in the clays might be considered as “crystal genes” to direct these organic processes. Though it is an ingenious theory, it is just that.  It has not been demonstrated practically as a means to produce the molecules required for a living cell.

Various other chemicals have been used as alternatives to Miller’s mixture, but they all have the same problems: a lack of relevance to the known composition of the primeval earth, low yields of the products of interest, inadequate explanations of stereochemical specificity and the destruction of the key compounds by the prevailing conditions or by other chemical by-products.


One textbook, edited by Soper (“Biological Science 1 and 2”; 3rd edition; Cambridge University Press) summarises the situation well (p. 883): Despite the simplified account given above, the problem of the origin(s) of life remains. All that has been outlined is speculation and, despite tremendous advances in biochemistry, answers to the problem remain hypothetical. … Details of the transition from complex non-living materials to simple living organisms remain a mystery.
This conclusion is echoed by those who have spent many years researching in this field of biochemistry. Dr D E Hull wrote, The conclusion from these arguments presents the most serious obstacle, if indeed it is not fatal, to the theory of spontaneous generation.
Prof Francis Crick, who was a great believer in the accidental origin of life on Earth, said, “The origin of life appears to be almost a miracle, so many are the conditions that had to be satisfied to get it going.” Prof. Crick goes on to argue that this might be overcome in long periods of time. However, there is no justification for believing that time can overcome basic chemical laws.

Dr H P Yockey (in the Journal of Theoretical Biology, 1981, 91, 26-29) wrote, You must conclude that no valid scientific explanation of life exists at present… Since science has not the vaguest idea how life originated on earth, … it would be honest to admit this to students, the agencies funding research and the public.

J H John Peet has taught chemistry at all levels from secondary schools to degree courses.