Browse The Gjenvick-Gjønvik Archives Home Page

Scientific Coffee Research Ongoing At M.I.T. - 1922

Scientific Coffee Research Ongoing At M.I.T. - 1922

Here Is the Coffee-Research Laboratory at the M. I. T.
All the Official N.C.R.A. Tests Are Made in These Scientific Quarters. Prof. Prescott Is Seen Leaning Over the Testing Table in The Center Foreground and Prof. Emerson At the Extreme Left Examining One of The Coffee-Drinking Rabbits.

Brewing Method More Important Than Quality of Bean - Study of Pure Caffein vs. Caffein in Coffee Now Going On at the Massachusetts Institute of Technology

The investigation work authorized by your committee has been continuously maintained during the past 12 months, especially along the lines which it was believed by the committee and those in charge of the work would give us the facts of greatest interest and importance regarding the chemistry of coffee and the changes that it undergoes in the treatments to which it is subjected in the transformation from the green bean to the infusion which brings comfort and cheer to millions of breakfast tables.

As a result of our observations on methods of preparation of coffee, we felt that it was advisable to attack this subject from a chemical as well as a physiological standpoint, and this work has been carried out, repeated, and confirmed.

In brief, we have found that the quality of beverage coffee is more influenced by a certain few factors in its brewing than it is by the differences in flavor, aroma, and taste attributed to coffee from different climates, soils, or altitudes when these factors are neglected.

In other words, we have found that a very fair beverage coffee can be brewed from a mediocre bean when these points are followed, and neglect of them will spoil the finest bean that otherwise might appeal to any individual taste.

Most significant of these factors are the effects of temperature of the water used in brewing, the time of contact between the ground coffee and the water, and the composition of the utensil in which the coffee is brewed.

Of secondary significance are the fineness of the grind and the potable character of the water, although it is found that waters of great hardness and alkalinity do affect adversely the quality of beverage coffee made with them.

The Brewing Conclusions

Assuming that a perfect or satisfactory coffee beverage is one pleasing to the senses of taste and smell as well as mildly stimulating and wholesome, we have made many experiments, and have sought unprejudiced opinion in many quarters.

As a result of two years' constant study, we have reached certain conclusions, which may be formulated tersely as follows:

  1. Coffee should be brewed with water at temperatures between 185° F. and 200° F. Above this temperature decompositions yielding bitter flavors take place, and mask to some degree the characteristic and delicate true coffee flavor.
  2. The time of contact of the ground coffee and the water should be brief. We have found the best results when this time period did not exceed two to two and a half minutes. With long contacts woody flavors are extracted from the bean, and bitter principles gradually pass into the solution.
  3. The coffee infusion should not come into contact with metals but should be brewed and kept in glazed or vitrified containers, such as porcelain, glass, or agateware. Even brief contacts with some metals yield pronounced bitter. astringent, or metallic flavors, due undoubtedly to the actual solution of and combination with minute amounts of the metal.
  4. Eighty to 85% of the total caffeine is extracted in a contact of two minutes when all the grains of coffee are thoroughly wetted. As coffee is an oily seed, it is necessary to have the water penetrate the superficial film of oil in order to emulsify it and to gain access to the other ingredients. Obviously, the small grains of a fairly fine grind will yield more dissolved material in a unit of time than if the coffee is coarsely ground. In the latter case more coffee must be used to secure the same strength of infusion.
  5. Freshly roasted and freshly ground coffee yields a beverage far superior to that made from ground coffee which has been exposed to the air. Our results seem to indicate that under these conditions there is a loss of the volatile ethereal flavors, and possibly an oxidation of the oils bringing about an incipient rancidity. These facts have been repeatedly observed, throughout the course of the work.

“Gross Chemistry” of Coffee

It is, however, with the further developments of this work that our present interest lies—and we have especially directed the work of the past five or six months to studies on what we might call the gross chemistry of coffee.

That is, we are attempting to isolate and identify those great groups of organic substances which may be regarded as probably characteristic of all coffees, rather than to attempt to identify and correlate the minute quantities of substances which may be found in any given sample of coffee as a result of the effect of oil, altitude, condition of ripeness, or other condition.

The substances in which interest now centers include:

  1. The oils, fats, and waxes.
  2. The free acids, if any.
  3. The so-called tannins.
  4. Astringent or bitter substances.
  5. The glucosides, sugars, fiber, and other constituents.
  6. The nature of the aroma or flavor-producing components.
  7. Further studies on physiological action.

It is not illogical to suppose that the roasting process brings about such deep-seated changes that a roasted coffee may differ materially from a green coffee in some of these components. This we find to be the case.

Oils, Fats, and Wares

The oils, fats, and waxes have been prepared from both green and roasted coffee, with a good degree of purity, and the work of identification, or classifying of these bodies into the groups to which they belong is well under way.

Here we are considering not only the character of the oily and waxy substances as they occur in the green bean, but also the changes they undergo during the roasting process and the decomposition products which may result from this heat treatment.

We have found that in both green and roasted coffee there occurs a bland vegetable oil, similar in odor, color, and taste to olive or peanut oil and probably belonging in the same group chemically. The various tests and quantitative studies necessary for actual identification are going on.

Study of these bodies indicates that they may have an important bearing on the flavor and aroma of coffee, that they are of importance in connection with the keeping quality, development of rancidity, stale or “off” tastes and odors, and also that they may have some physiological significance, especially for those who are supersensitive to coffee.

It is not uncommonly reported by those who substitute one or two grains of pure caffeine for their morning coffee that in the latter case the reason is due to the oily substances.

A second type of substance of this group found in both green and roasted coffee is a hard wax, resinous or varnish-like in character, colorless, and soluble in water or gastric juice, but existing in a coffee extract possibly held in solution by the other oil constituents.

This occurs in greater quantity in boiled than in unboiled coffee, which seems to indicate that during the process of boiling some chemical action takes place resulting in an increased formation of bodies of this class.

This waxy substance attaches itself to glass or other utensils and apparatus from which it can be removed only by alcohol or by saponification. To what extent this is prejudicial to digestion is at present a matter of conjecture only. These substances are now in the hands of an expert organic chemist for identification.

Occurrence of Acids

The use of the ordinary methods of acid determination by use of indicators is impossible in coffee extracts, because of the interference due to coloring matters.

The green coffee substance is neutral to litmus; that is, no acids in free condition are present in significant amounts. A coffee which had undergone fermentation would naturally show traces of lactic or similar acids, but normal coffees seem to be free from these.

During the process of roasting, hydrolysis of some of the components may give rise to weak acids. While these cannot be measured by the usual methods in fresh extracts, it is probable that by resorting to the complicated methods of hydrogen-ion determination they may be detected, and then recovered in a study of the fractions of decomposition substances obtained by special Solvents. The department has recently installed a hydrogen-ion apparatus, and this investigation will be made at once.


While our program as outlined did not specify any detailed study of the proteins, it seems to me that this complex group of substances should not be entirely overlooked.

The amount of protein in coffee, as indicated by the nitrogen, is not large, and because of their colloidal nature and difficulty of separation it may be impossible to prepare pure products for identification.

But we feel that if any inkling can be gleaned as to the effect on the consumer of these native proteins or their decomposition products of roasting, or by preparation of beverage coffee, an attempt to study their physiological action on test animals may be worthwhile.

Carbohydrates and Glucosides

Carbohydrates exist in green coffee as cellulose and its derivatives, and, as starch and its derivatives, sugar glucosides (consisting of complexes of carbohydrates and tannin-like bodies) also occur.

During roasting, which is essentially a process of destructive distillation, these compounds are hydrolyzed and split into simpler bodies yielding a number of cleavage products which fail to give the notions of unaltered starch and may occur in suspended or colloidal condition too finely divided to be removed by a filter. Some of these substances, especially in boiled coffee, produce the haze or smoke which appears on standing.

The presence of these starchy or cellulose bodies is also shown in the green beans by the effects produced by roasting after treatment with solvents which do not affect or remove the starch.

The roasting produces a series of decomposition substances which have a characteristic odor like that of popcorn. The number and character of the products thus produced is dependent on the degree of roasting to which the sample is subjected. Thus two samples will be alike only when they were alike at the start and have been treated in an identical manner.

In practice, conditions are sufficiently similar to make the results of a single series of determinations fairly representative.

These carbohydrates are derived in part from the glucosides of the seed which are broken down by the roasting process into sugars and tannin like bodies. A group of what we call the tannin-like compounds has been separated, and their chemistry and classification is now being studied.

These bodies fail to give some of the most characteristic reactions of true tannins, but they no doubt represent the substances that have frequently been called the tannins of coffee.

The possible deleterious effects of these tannin-like bodies may conceivably be shown by animal experimentation, and work along this line is going on.

Animals are tested with different fractions of our extracts in the hope that we may be able to detect the substance or fraction which makes the animal susceptible to coffee.

As animal experiments require constant repetition in order to draw fair conclusions, the work must necessarily be continued until a considerable number of observations have been made.

In addition to these, studies on the decomposition products are being pursued by special methods of chemical identification of organic compounds. The causes of the bitter or astringent taste in poorly brewed or over brewed coffee may he referable to these decompositions and the varying solubilities of the substances they yield.

Gases in Coffee

We have already reported that when ground roasted coffee is heated with water to a temperature of 98°-100° C. a very marked change with evolution of gas takes place. It is a matter of common knowledge that, when coffee is made in the ordinary household way and set on the back of the stove to brew, it is prone to “boil over" unless watched.

The significance of this phenomenon has possibly been overlooked, although the action itself is striking to one studying the process of coffee making.

In view of the method of roasting and the changes that coffee undergoes during roasting, the questions that at once arise are:

  1. Are the gases which are evolved due to absorbed or occluded gases having their origin in the fuel used for roasting?
  2. Are they products of a chemical decomposition completed when the coffee and water are brought up to a definite critical temperature?

We have collected the gases evolved and made analyses to determine what they are. We have found carbon dioxide in largest quantity, but also carbon monoxide, acetylene, and others of this series of hydrocarbons, and a residuum of aromatic gases of different odors, some pleasant and some otherwise.

Often these have the vegetable odor found in water filled with decaying vegetation or what the chemist calls a “vegetable odor.” The amount of these aromatic gases is very small, even from large quantities of coffee.

A comparison of the gases evolved from stale roasted coffee, roasted more than six months previous to the experiment, with a coffee roasted only a few days prior to the experiment, but both ground at the same time, shows that there is a very marked loss or diminution in the amount of carbon dioxide given off when coffee was brewed with water beginning at low temperature and gradually increasing the heat to boiling.

In the case of a dark-roasted coffee examined we collected only about 20% of CO from the stale coffee compared with a freshly roasted one. The loss of 80% of the CO, although not heretofore observed, must have been by slow evolution, and in our opinion might have an important bearing on the so-called vacuum-pack process and would explain in part at least the swelling or bulging of the containers when this style of packing is carried in storage.

What this denotes as to the change in composition of the coffee or the origin of the gas is not yet clear, for it may conceivably be due to the slow breaking down of substances rendered unstable by the roasting process, or to gases produced and occluded in the cells due to the heat of the roast.

The higher the roast, the larger is the amount of carbon dioxide given off, for similar experiments with a medium or low roast have given consistently lower results.

The relation of amount of gases evolved seems to point to the probability that the carbon dioxide is formed during the roasting process. This would also explain why freshly roasted coffee will float when thrown into cold water, due to the buoyancy of the gas,

Gases Derived from Wasting Process

A comparison of roasted coffee and green coffee of the same age shows that, while there is a large gas content in the roasted coffee, the amount in the green is practically negligible.

It is then evident that the gases are derived from the wasting process. There is no evidence yet available to determine to what extent the escaping gas may carry away other volatile flavoring substances.

Study of all these components suggests that the composite odor which we call the characteristic odor of freshly made coffee is undoubtedly due to a complex mixture or combination of all these substances, together with the odors of the oils, and other components.

An analogy is found in the chemistry of perfumes, where the characteristic odor is due to a complex of substances, some of which often are of decidedly obnoxious smell.

To what extent the character of a coffee when brewed may be due to decomposition or oxidation processes in roasting, or to saponification of the oils by heating with water, can be determined only by a study of these fractions of materials obtained in the laboratory.

For example: Coffee grounds when extracted so as to remove the oils yield on roasting a characteristic popcorn odor or the odor of roasted vegetable fiber.

This is not at all apparent in roasted whole (i. e., unextracted) coffee, but it is present, and is probably one of the essential components in the aroma of coffee.

This fact possibly explains why it is that a re-roasted or reheated ground coffee regains something of its original odor, although we know that the gain is not necessarily permanent. It is conceivable, however, that such a gain in odor may be accompanied by a loss of actual flavor forming substances.

Experiments with Animals

Testing the Effects of Caffeine on Rabbits at MIT.

Here are the bunnies at the Mass. Tech. that help Prof. Prescott in testing the effect of caffeine in coffee. The rabbits are fond of coffee and thrive on it. Note them lapping it up in the pictures on the top left and bottom right.

The experiments on animals have been continued, and, as we regard this as an essential part of the chemical as well as the physiological standpoint, we propose to continue this phase of the work until the end of the research.

The method of experiment at first adopted proved to be cumbersome and slow and might have been continued indefinitely before practical results were secured.

It was found advisable, therefore, to make some changes in both personnel and method, and the results to date are much more promising. One of the purposes of the ani mal experimentation is to prove whether there are substances other than caffeine which have any pharmacological action, unpleasant or otherwise, upon those who are apparently supersensitive to coffee.

Manifestly it is impossible to determine by chemical analyses solely, no matter how thoroughly or completely the work may be done, what the effects of the compounds of coffee are upon the human organism, and the nearest approach is by use of animals.

The problem to be determined is comprised in the three questions:

  1. What amount of caffeine can the animal tolerate and maintain normal conditions?
  2. What is the amount tolerated when administered in beverage coffee made in different ways?
  3. What evidence is obtainable as to the effect of specific groups of substances in coffee other than the caffeine?

Obviously the first question must be answered in order to obtain reliable data on which to proceed, or, in other words, to get a base line. A large number of animals have now been tested with this objective:

We have found that the amount of caffeine a rabbit will tolerate at one dose varies with the body weight and with the environmental factors of food, physical condition, etc., just as is the case with the human "animal.

It can tentatively be assumed on the basis of our work that the tolerance of a rabbit for caffeine is more than 1.5 grains of caffeine per pound of body weight. This amount is about the equivalent of that in an average cup of coffee.

This amount, aside from the temporary disturbance due to the large amount of liquid in the stomach, does not affect the rabbit's subsequent feeding or apparent health.

Pure Caffeine and Caffeine in Coffee

The comparison between pure caffeine derived from coffee and caffeine in coffee beverage is most important, and here many tests must be made before we are willing to draw conclusions.

Some experiments have indicated that there may be a difference in the effect of coffee beverage made at a temperature below 93° C. and a similar beverage made at a higher temperature. We also have evidence of a difference between pure caffeine and boiled coffee.

It takes man" trials to establish the actual facts and draw reliable conclusions, and until the results have been repeatedly checked and confirmed we cannot feel warranted in formulating specific statements. This work is being pushed forward vigorously and with promise of useful results.

Progress in Coffee Research at M.I.T. - 1922

Three years may seem a long time to some of us for the pursuit of research without securing conclusive and new results. That is an inevitable opinion for the average American whose trend of thought is pointed toward immediacy.

To indicate the injustice of harboring such an idea, consider the years of heart-breaking effort that passed before synthetic indigo was an accomplished fact or until our electric lights were brought to their present state of refinement.

Prof. Prescott at M. I. T. has been confronted with a real problem. Business acumen dictated that results should be quickly secured that would reimburse the coffee trade for its expenditure.

Scientific integrity demanded the stabilization of basic facts, regardless of the time involved. The course of procedure followed in the research he has directed would seem to have been that of the true scientist.

Published data relevant to coffee were collected and their correctness proved or disproved, as the case might be. From the established facts the mode of attack was planned.

New discoveries were slow in coming; so of necessity the reports emphasized the salient features of previous researches. Although risking criticism, this gave to the public necessary information backed by the authority of the Massachusetts Institute of Technology.

In the report presented before the last convention of the N. C. R. A. further weight is added to the arguments of the “fresh-roasted, fresh-ground, quick-brew, anti-boiling, anti-metal” school. The practice of packing coffee in the absence of air receives fresh support.

Chemical and physiological study of the individual constituents of coffee goes fast apace. Comparison of the caffeine in coffee with alkaloid caffeine should yield fresh evidence vindicating coffee.

Within another year we hope results will be forthcoming that will throw much light upon the behavior of coffee and indicate valuable methods for its improvement.

In the absence of scientific details we shall continue to hope that a solid foundation for coffee research has been laid. The building should be interesting and instructive and perhaps spectacular. To stop now would be to waste most of the money and effort expended.

Any change of policy toward the suppression of ascertained facts will operate in the same direction. It is to be hoped that circumstances will soon operate to permit the publication of coffee-research details.

Thus far the generalities have served, but they are not enough. The Joint Coffee Trade Publicity Committee and Prof. Prescott must know that, to have their claims accepted by scientific men and historians, the details of the work done at the M. I. T. must be published in certain recognized mediums, and that until the publication of all the facts in each experiment any advertising claims based upon them are ephemeral. Repeating what we said in May 1921, such a manner of recording the facts must of necessity lend an additional air of authority to those claims.

Samuel C. Prescott, "Scientific Coffee Research," in Tea and Coffee Trade Journal, New York: The Tea and Coffee Trade Journal Co., Vol. 42, No. 12, December 1922, p.884-888; and “Progress in Coffee Research,” p. 901h

Return to Top of Page