Three Laws of Nature A Little Book On Thermodynamics 9780300238785 0300238789
Three Laws of Nature A Little Book On Thermodynamics 9780300238785 0300238789
Three Laws of Nature A Little Book On Thermodynamics 9780300238785 0300238789
R. Stephen Berry
Published with assistance from the foundation established in memory of Philip Hamilton McMillan of the
Class of 1894, Yale College.
This book may not be reproduced, in whole or in part, including illustrations, in any form (beyond that
copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the
public press), without written permission from the publishers.
Yale University Press books may be purchased in quantity for educational, business, or promotional use.
For information, please e-mail sales.press@yale.edu (U.S. office) or sales@yaleup.co.uk (U.K. office).
Set in Janson Roman type by Integrated Publishing Solutions. Printed in the United States of America.
10 9 8 7 6 5 4 3 2 1
Contents
Preface
ONE
What Is Thermodynamics?
The First Law
TWO
Why We Can’t Go Back in Time
The Second and Third Laws
THREE
How Did Classical Thermodynamics Come to Exist?
FOUR
How Do We Use (and Might We Use) Thermodynamics?
FIVE
How Has Thermodynamics Evolved?
SIX
How Can We Go Beyond the Traditional Scope of Thermodynamics?
SEVEN
What Can Thermodynamics Teach Us About Science More Generally?
Index
Preface
This book has a very specific purpose: to use the science of thermodynamics as a
paradigm to show what science is, what science does and how we use it, how
science comes to exist, and how science can evolve as we try to reach and
address more and more challenging questions about the natural world. It is
intended primarily for people with little or no background in science, apart,
perhaps, from some exposure in school or college to “science for everyone”
courses.
The approach, quite frankly, came from three stimuli. One was a course I
gave for non-scientist undergraduates at the University of Chicago that evolved
over years. The second was an adult education course that grew out of that
undergraduate course—but was at a somewhat lower level, primarily in terms of
the mathematics. The third, first in time, which had a very crucial influence on
that evolution, was a pair of essays published in 1959 by the British scientist and
novelist C. P. Snow, titled “The Two Cultures,” based on lectures he had given at
Cambridge University. Snow said, in essence, that those who belong to the
culture of scientists know vastly more about the culture of humanities than the
reverse, and that this is a serious problem for our society. He argued,
metaphorically, that if we were to achieve a proper balance, the non-scientists
would know as much about the second law of thermodynamics as the scientists
know about Shakespeare. In a later essay, actually published in 1964 with the
first two, he retreated and replaced the second law of thermodynamics with
contemporary biology.
I might have disagreed with his change even then, but as biology has
advanced it is very clear that, in our time, he was right in the first place. Why?
To understand biology today, one must learn many, many facts, but one needs
very few facts to understand thermodynamics. It is a subject rooted in a few
concepts, far more than on systematic inferences from vast amounts of
information, as contemporary biology is now.
This book is written for people who are interested in what science is and does,
and how it comes to be, but who have little or no prior substantive knowledge of
any science. In writing for this audience, there was a clear choice to be made
regarding the order of the chapters. Would the ideas be clearer if the history
came first, or if the first thing one read was the substance of the science as it is
today? I have chosen the latter. Thus I first show, in Chapters 1 and 2, what the
science of thermodynamics is as we now use it almost constantly, and then, in
Chapter 3, address the question of how we arrived where we are.
The first two chapters develop an overview of what the traditional science of
thermodynamics is, of the concepts that embody it, and how it uses these
concepts. These chapters are built around the three laws of thermodynamics,
namely the conservation of energy, the inevitability and direction of change with
time, and the existence of a limiting absolute zero of temperature.
The third chapter presents the history of the science. There we see how the
concepts, the variables, and the tools that we call the laws of thermodynamics
evolved. This is important because some of the concepts of thermodynamics are
so ubiquitous and pervasive in our thinking and language that they seem obvious
and trivially simple to us, where, in reality, they evolved only with great
difficulty, and remain in many ways amazingly subtle and not so simple at all.
That history has been, and even still is, a tortuous but fascinating path, especially
when we understand where and what thermodynamics is today. And it offers a
rich insight into how a science evolves, often having to resolve competing,
conflicting concepts.
The fourth chapter describes the ways we apply thermodynamics, especially
how we use information from this science to improve the performances of many
things we do in everyday life. The fifth chapter describes the way the science of
thermodynamics has continued to evolve since its pioneers established its
foundations. This is, in a sense, a contemporary counterpart to Chapter 3.
Despite the elegant and in many ways the most natural presentation of the
subject, in a formal, postulational and essentially closed format, the subject has
continued to undergo evolutionary changes, expanding and strengthening as new
discoveries require that we extend our concepts. A crucial advance discussed in
this chapter is the connection between the macroscopic approach of traditional
thermodynamics and the microscopic description of the world, based on its
elementary atomic building blocks. This link comes via the application of
statistics, specifically through what is called “statistical mechanics.”
The sixth chapter examines the open challenges that we face if we wish to
build upon the thermodynamics we now have to provide us with tools to go
beyond, to other, still more difficult and usually more complex problems of how
nature functions. It addresses questions of whether and when there are limits to
what a particular science can tell us, and whether inferences we make using the
tools and concepts of a science can sometimes have limits of validity. This
chapter reveals how this science continues to offer new challenges and
opportunities, as we recognize new questions to ask.
The seventh and final chapter addresses the question of how thermodynamics,
one particular example of a kind of science, can show us what a science is and
does, and hence why we do science at all. This chapter is a kind of overview, in
which we look at how thermodynamics can serve as a paradigm for all kinds of
science, in that it reveals what scientific knowledge is, how it differs from
knowledge in other areas of human experience, how scientific knowledge, as
illustrated by thermodynamics, has become deeply integrated into how we live,
and how we can use scientific knowledge to help guide human existence.
A comment directly relevant to the intent and content of this book:
“Thermodynamics is the science most likely to be true.” This is a paraphrase of a
comment made by Albert Einstein; he was putting thermodynamics in the larger
context of all of science. Einstein certainly recognized that every science is
always subject to change, to reinterpretation, even to deep revision; he himself
was responsible for such deep changes in what had been accepted as scientific
laws. His comment reflects his belief that thermodynamics contains within itself
aspects—concepts and relations—that are unlikely to require deep revision,
unlike the way quantum theory revised mechanics and relativity revised our
concepts of space and gravitation. Was he right?
I would like to acknowledge the very helpful comments from Alexandra Oleson
and Margaret Azarian, both experienced and skilled at editing.
Three Laws of Nature
ONE
What Is Thermodynamics?
The First Law
ΔE = E(2) – E(1).
The important point we must now recognize is that this change in energy doesn’t
depend in any way on how we bring the system from state 1 to state 2. We might
be able to make the change by heating or cooling alone, or by doing or extracting
work with no exchange of heat with an outside source, or by some combination
of the two. No matter how we take the system from state 1 to state 2, the energy
change is the same. This means that however we exchange heat or perform or
extract work, their difference must be the same fixed amount, the difference
between E(1) and E(2). We can write this as an equation, either for measurable
changes of energy, as
ΔE = Q - W
δE = δQ - δW.
These equations say that however we put in or take out heat, and however the
system does work or has work done on it, the difference between the heat taken
in and the work done is precisely equal to the change in energy of the system
from its initial to its final state. It is these equations that are our terse statement
of the first law of thermodynamics. These two equations tell us that however
much work we do or extract to take our system, say our engine, from its initial
state 1 to its final state 2, there must be an exchange of heat corresponding
precisely to that amount of work so that the change of energy is just that
difference between the two states. Said another way, these equations tell us that
there is no way to create or destroy energy; we can only change its form and
where it is.
One very simple example illustrates this principle. The energy of a 20-pound
weight is higher if it is on top of a 50-foot pole than if it is on the ground at the
base of the pole. But the difference between the energies of that weight does not
depend in any way on how we brought it to the top of the pole. We could have
hoisted it up directly from the ground, or dropped it from a helicopter hovering
at an altitude of 100 feet; the difference in the energy of the ground-level state
and the state 50 feet higher in no way depends on how the states were produced.
This also tells us something more about how real machines operate. Think of
an engine such as the gasoline engine that drives a typical automobile. Such an
engine operates in cycles. Strictly, each cylinder goes through a cycle of taking
in fuel and air, igniting and burning the fuel, using the heat to expand the gas-air
mixture, which pushes the moveable piston of the cylinder. The moving piston
connects via a connecting rod to a crankshaft and the expansion in the cylinder,
pushing the piston, in turn pushes the crankshaft to make it rotate. That
crankshaft’s rotation eventually drives the wheels of the car. The important part
of that complex transformation for us at this point is simply that each cylinder of
the engine goes regularly through a cycle, returning to its initial state, ready to
take in more fuel and air. In the process of going through the cycle, the burning
of the gasoline in air releases energy stored in chemical bonds and turns it into
heat. The engine turns that heat into work by raising the pressure of the burned
gasoline-air mix enough to push the piston through the cylinder. At the end of
each cycle of each cylinder, the piston returns to its original position, having
converted chemical energy to heat, and then, transformed that heat into work.
In an ideal engine, with no friction or loss of heat to the surroundings, the
engine would return to its initial state and as much as possible of the chemical
energy that became heat would have become work, turning the wheels of the car.
(“As much as possible” is a very important qualifier here, which will be a central
topic of a later discussion.) Such an engine, first envisioned by the French
engineer Sadi Carnot, has the remarkable property that at any stage, it can move
equally easily in either direction, forward or backward. We call such an ideal
engine or process a reversible engine or reversible process. One extremely
important characteristic of reversible processes is that they can only operate
infinitely slowly. At any instant, the state of a reversible engine is
indistinguishable from—or equivalent to—a state of equilibrium, because it can
go with equal ease along any direction. Such reversible processes are ideal and
unreachable limits corresponding to the (unreachable) limits of the best possible
performance that any imaginable engine could do. However, we can describe
their properties and behavior precisely because they are the limits we can
conceive by extrapolating the behavior of real systems. They would operate
infinitely slowly, so obviously are only useful as conceptual limits, and in no
way as anything practical.
Of course real engines are never those ideal, perfect machines. They always
have some friction, and there is always some leakage of heat through the metal
walls of the engine to the outside. That the engine feels hot when it has been
running testifies to the loss of heat to the environment; that we can hear the
engine running is an indication that there is friction somewhere that is turning
some, maybe very little, of that energy that was heat into sound waves, generated
by friction. But we try to minimize those losses by using lubricating oil, for
example, to keep the friction as low as we can. We make real engines that are as
much like those ideal engines as we can, but of course require the real engines to
operate at real rates, not infinitely slowly, to deliver the work we want from
them.
One very important, inherent characteristic of what we call energy is the
remarkably broad scope of what it encompasses, as we already have begun to
see. Every activity, every process, even every thing there is has associated with it
an energy or energy change. The concept of energy includes, in its own way,
some aspect of virtually everything in human experience. In the next chapter, we
shall examine how this concept evolved and grew, as human experience and
understanding of nature grew. At this stage, we need only recognize that energy
is a property we can identify, often measure, and detect in an amazingly broad
variety of forms. Light, sound, electricity and magnetism, gravitation, any
motion whether constant or changing, the “glue” that holds atomic nuclei
together, even mass itself, are all examples of the manifestation of energy. And
now, there are even forms of energy that we do not yet understand, what we call
“dark matter” and “dark energy.” That the human mind has recognized the very
existence of such a universal manifestation of the natural world is itself a source
of amazement and wonder.
A Closer Look at Some Basic Concepts for the Next Big Step
Summary
We have seen how the first law of thermodynamics can be expressed in words:
energy can never be created or destroyed but only changed in form and location.
We have also seen how this law can be expressed in a very simple equation,
ΔE = Q - W
which tells us that ΔE, the change in energy of any system, is precisely the heat
Q that the system takes in, minus the work W that the system does on something
outside itself. It says that the change of energy of the system, from an initial state
to some final state, depends only on the difference of the energies of those two
states and not on the pathway taken to go from one to the other. It says that if a
system traverses a closed loop, that brings it back to its initial state, that the
system has undergone no change in its energy. It is a law we believe is valid
everywhere in the universe, a law that describes every form that energy can take.
When we realize all the forms that we know energy can have—heat, work,
electromagnetic waves, gravitation, mass—this becomes a remarkable, even
awesome statement.
TWO
Why We Can’t Go Back in Time
The Second and Third Laws
The second law of thermodynamics is very different from the first law. The first
law is about something that doesn’t change, specifically the total amount of
energy, or even the amount of energy in some specific part of the universe,
closed off from the rest of the world. The second law is about things that can’t
happen, which can be interpreted as distinguishing things that can happen from
those that can’t—which is a way of showing what the direction of time must be.
The first explicit formulations of the second law were three verbal statements
of the impossibility of certain kinds of processes. In 1850, the German physicist
and mathematician Rudolf Clausius put it this way: no engine, operating
cyclically, can move heat from a colder to a hotter body without doing work.
Refrigerators can’t operate spontaneously. Then, the next year, the British
scientist William Thomson, Lord Kelvin, stated it differently: no engine,
operating cyclically, can extract heat from a reservoir at a constant temperature
and convert that heat entirely into work. The third of these statements, by
Constantin Carathéodory in 1909, is, in a sense, more general and less
specifically related to work than its predecessors; it says that every equilibrium
state of a closed system has some arbitrarily close neighboring states that cannot
be reached by any spontaneous process (or by the reversible limit of a
spontaneous process) involving no transfer of heat. “You can’t get there from
here.” This, one can see, encompasses the two earlier statements and carries with
it the implication that one may be able to go from state A to state B by a
spontaneous or limiting reversible process, yet be unable to go from B back to A
by such a process. This is a very fundamental way to express the unidirectional
nature of the evolution of the universe in time. For example, if state A consists of
a warm object X connected to a cooler object Y by some means that can conduct
heat, then X will cool and Y will warm until X and Y are at the same
temperature—and that is our state B. We well recognize that the system
consisting of X, Y, and a connector between them will go spontaneously from
the state A in which the temperatures of X and Y are different to the state B, in
which X and Y are in thermal equilibrium, but that we will never see the system
go from thermal equilibrium to two different temperatures.
At this point, we must introduce a new, fundamental concept that becomes a
most important variable now, for our second law of thermodynamics. This
variable is very different from energy, and the natural law associated with it,
likewise, is very different from the first law. This new, fundamental variable is
entropy. Of all the fundamental concepts in thermodynamics, entropy is surely
the one least familiar to most people. The concept of entropy arose as the
concepts of heat and temperature evolved. More specifically, entropy emerged
out of an attempt to understand how much heat must be exchanged to change the
temperature of an object or a system by a chosen amount. What is the actual heat
exchanged, per degree of temperature change in the system? Then the natural
next question is, “What is the absolute minimum amount of heat, for example
the number of calories, that must be exchanged to change the system’s
temperature by, say, 1 degree Celsius?” The most efficient engine imaginable
would exchange only the minimum amount of heat at each step of its operation,
with none wasted to friction or loss through the system’s walls to its
environment.
Now, in contrast to those earlier statements, the most widely used expression
of the second law is in terms of entropy. Entropy, expressed as a measure of the
amount of heat actually exchanged per degree of temperature change, means that
it is very closely related to the property of any substance called its heat capacity,
or, if it is expressed as a property of some specified, basic unit of mass of a
material, as specific heat. Precisely defined, the heat capacity of a substance is
the amount of heat it must absorb to increase its temperature T by 1°.
Correspondingly, the specific heat of that substance is the amount of heat
required to raise the temperature of one unit of mass by 1°. Typically, both are
expressed in units of calories per degree C or K, where K refers to the Kelvin
temperature scale, named for Lord Kelvin. The units of measure in this scale,
which are called kelvins, not degrees, are the same size as those of the Celsius
scale, but zero is at the lowest physically meaningful temperature, a presumably
unattainable level at which point absolutely no more energy can be extracted
from the system; at 0 K, there is zero energy. This point, which is often referred
to as absolute zero, is just below -273° C; hence, 0° C is essentially 273 K. (We
shall return to the Kelvin scale when we consider the third law of
thermodynamics.) At this point, we need to look at heat capacities and specific
heats, in order to probe this approach to the concept of entropy.
The heat capacity of anything obviously depends on its size; a quart of water
has twice the heat capacity of a pint of water. To attribute the temperature
response of a substance without having to refer to the amount of matter, we use
the heat capacity per unit of mass of that substance, which is the specific heat of
the substance. Thus the specific heat of water is 1 calorie per degree C, per gram
or cubic centimeter of water. Specific heat is an intensive quantity (which does
not depend on the size of the object); heat capacity is its corresponding extensive
quantity, proportional to the mass of the object. Hence the heat capacity of 1
gram of water is the same as its specific heat, but the heat capacity of 5 grams of
water is 5 calories per degree C. In the next chapter, we shall discuss the role
played by specific heats and heat capacities in the evolution of the concepts of
thermodynamics. Typical specific heats range from about 1.2 calories per degree
C per gram for helium to much lower values, such as approximately 0.5 for ethyl
alcohol, only 0.1 for iron, and 0.03 for gold and lead. Metals require
significantly smaller amounts of heat to raise their temperatures.
One subtlety we can mention here is the dependence of the specific heat and
heat capacity on the constraints of the particular process one chooses to heat or
cool. In particular, there is a difference between heating a system at a constant
pressure and heating it while fixing its volume. If the pressure is constant, the
volume of the system can change as heat is added or withdrawn, which means
the process involves some work, so we must make a distinction between the heat
capacities and specific heats under these two conditions.
The entropy change, when a substance absorbs or releases energy as heat, is
symbolized as ΔS, where S is the symbol generally used for entropy, and Δ, as
before, is the symbol meaning “change of.” As we discussed, we use Δ to
indicate a significant or observable change, while the lowercase δ indicates a
very tiny or infinitesimal change—of anything. Then when a system undergoes a
change of state by absorbing a tiny bit of heat, δQ, its entropy change δS must be
equal to or greater than δQ/T. Formally, we say δS ≥ δQ/T. The entropy change
must be equal to or greater than the amount of heat exchanged, per degree of
temperature of the system. Thus, for a given exchange of heat, the entropy
change is typically greater at low temperatures than at higher ones. An input of a
little heat at low temperatures makes a bigger difference in the number of newly
accessible states than at higher temperatures, and it is just that number of states
the system can reach that determines the entropy.
That is one quantitative way to state the second law of thermodynamics, in
terms of an inequality, rather than an equation of equality, such as the statement
of the first law. But that doesn’t yet tell us what entropy is, in a useful, precise
way. That only tells us that it is a property that is somehow bounded on one side,
that its changes can’t be less than some lower limit related to the capacity of a
system to absorb or release energy as heat. And that it depends inversely on the
temperature at which that heat exchange occurs and hence somehow on the
number of ways the system can exist. But from this inequality, we can get a first
sense of what entropy tells us. If a substance absorbs an amount δQ of heat and
all of that heat goes into raising the temperature of the substance according to
that substance’s heat capacity, then the process would have an entropy change δS
precisely equal to δQ/T. If some of the heat were to escape by leaking to the
surroundings, then the entropy change would be greater than δQ/T. Hence we
can already see that entropy somehow reflects or measures something wasteful
or unintended or disordered.
A simple example illustrates how temperature influences entropy. For a gas
with a constant heat capacity, the entropy increases with the logarithm of the
absolute temperature T. Hence the entropy doubles with each tenfold increase in
the absolute temperature; it doubles when the temperature goes from 1 K to 10
K. But it only doubles again when the temperature rises from 100 K to 1000 K!
The lower the temperature, the more a given rise in temperature (in degrees)
affects and thereby increases the system’s entropy. Hot systems are inherently
more disordered and thereby have higher entropy than cold systems, so a small
increase in the temperature of a hot and already disordered system doesn’t make
it much more disordered.
Thus far, we have used the word “entropy,” and begun to describe a little of
its properties, but we have not yet said what entropy is. To do that, we will begin
to bridge between the macroscopic world of our everyday observations and the
microscopic world of atoms and molecules. At our macroscopic level, we
describe a state of a system in terms of the familiar macroscopic properties,
typically temperature, pressure, density, and mass or volume. In fact, any
equilibrium state of a macroscopic system can be specified uniquely by
assigning specific values to only a subset of these variables; that is enough to
determine all the other state variables. That is because the state variables of a
substance are linked in a way expressed by an equation, the equation of state,
unique for each substance. These equations can be simple for simplified models,
as we shall see, or they can be very complex if they are put to use to get precise
information about real systems, such as the steam in a steam turbine.
In the following discussion, we shall designate a state specified by those state
variables as a macrostate. The number of those variables that are independent—
that is, can change independently—is fixed by what is perhaps the simplest
fundamental equation in all of science, the phase rule, introduced by a great
American scientist, long a professor at Yale, J. Willard Gibbs, between 1875 and
1878. This rule relates three things. The first, the one we want to know, is the
number of degrees of freedom, f, which is the number of things one can change
and still keep the same kind of system. This is directly and simply determined by
the second quantity, the number of substances or components c in the system,
and the third, the number of phases p, such as liquid, gas, or solid, of a specific
structure that coexist in equilibrium. These are linked by the relation
f = c - p + 2.
Not even a multiplication! It’s just addition and subtraction. We will later go
much deeper into this profound equation. For now, we need only say that every
phase of any macroscopic amount of a substance has its own relation, its
equation of state, that fixes what values all the other variables have if we specify
the values of just any two of them.
Typically, there is a different equation of state for the solid, the liquid, and the
gaseous form of any chosen substance. (Some solid materials may have any of
several forms, depending on the conditions of temperature and pressure. Each
form has its own equation of state.) The simplest of all these equations is
probably the equation of state for the ideal gas, a sort of limiting case to describe
the behavior of gases in a very convenient way that is often a very good
approximation to reality. This equation relates the pressure p (not to be confused
with the number of phases p), the volume V, and the temperature T based on the
absolute Kelvin scale, where the lowest point is known as absolute zero. The
relation involves one natural constant, usually called “the gas constant” and
symbolized as R, and one measure of the amount of material, effectively the
number of atoms or molecules constituting the gas, counted not one at a time or
even by the thousands, but by a much larger number, the number of moles.
A mole is approximately 6 × 1023 things, and is essentially an expression of a
conversion from one mass scale to another. In the atomic mass scale, a hydrogen
atom has a mass of 1—that is, 1 atomic mass unit. Collecting 6 × 1023 hydrogen
atoms, we would have 1 gram. Hence we can think of that number, 6 × 1023, as
the conversion factor between atomic mass units and grams, just as 2.2 is the
number of pounds in a kilogram. The number 6 × 1023 is known as Avogadro’s
number, and is symbolized as NA. The number of moles of the gas is represented
as n. To get a feeling for the amount of material in one mole, think of water,
H2O, which has a molecular weight of 18 atomic mass units, so one mole of
water weighs 18 grams. But 18 grams of water occupies 18 cubic centimeters, or
about 1.1 cubic inches. Hence 18 cubic centimeters of water contains 6 × 1023
molecules, one mole of water.
So now we can write the ideal gas equation of state:
pV = nRT.
Most other equations of state are far more complicated. The equation of state for
steam, used regularly by engineers concerned with such things as steam turbines,
occupies many pages. But every state of every substance has its own equation of
state. From these equations, we can infer the values of any macroscopic variable
we wish from knowledge of any two and, if the variable we want is extensive, of
the amount of material. We shall return to this subject in later chapters and
derive this equation. (If we wanted to count the particles individually, instead of
a mole at a time, we could write the equation as pV = NkT, where we use N to
indicate the number of atoms or molecules and k is known as the Boltzmann
constant, so the gas constant R = NAk, Avogadro’s number of little Boltzmann
constants.)
At this point we can begin to explore what entropy is. We easily recognize
that every macroscopic state of any system, the kind of state specified by the
variables of the appropriate equation of state, can be realized in a vast number of
ways in terms of the locations and velocities of the component atoms or
molecules. All those individual conditions in which the positions and velocities
of all the component particles are specified are called microstates. Microstates of
gases and liquids change constantly, as the component atoms or molecules move
and collide with one another and with the walls of their container. In fact, when
we observe a macrostate, whatever process we use invariably requires enough
time that the system passes through many microstates, so we really observe a
time average of those microstates when we identify a macrostate. The molecules
of oxygen and nitrogen in the air in our room are of course in constant—and
constantly changing—motion, hence constantly changing their microstate.
However, we think of the air in a room as being still, at a constant temperature
and pressure, and thus in a nonchanging macrostate.
In a microstate, any two identical particles can be exchanged, one for the
other. We would call the two states, with and without that exchange, two
microstates. Nevertheless we would identify both as representative of the same
macroscopic state. The entropy of a macroscopic state or macrostate is the
measure of the number of different microstates that we identify as corresponding
to that macrostate. Put more crudely, we can say that entropy is a measure of
how many ways all the component atoms could be, that we would say are all the
same for us. For example, the microstates of the air molecules in a room are all
the possible positions and velocities of all those molecules that are consistent
with the current temperature and pressure of the room. Obviously the numbers of
those microstates are so vast that, instead of using the actual number of
microstates accessible to a system, we use the logarithm of that number.
Logarithms are very convenient for any situation in which we must deal with
very large numbers. The logarithm of a number is the exponent, the power to
which some reference number must be raised to produce the given number. The
reference number is called the base. The three most common bases are 10, 2, and
the “natural number” e, approximately 2.718. Logarithms to the base 10 are
written “log,” so 2 is log(100) because 102 = 100. Logarithms to the base e are
usually written as ln, meaning “natural logarithm.” Logarithms to the base 2 are
often written log2( ) so log2(8) = 3 because 23 = 8. We shall discuss this in more
detail in the context of entropy in the next chapter.
Defined precisely, the entropy of any macrostate is the natural logarithm, the
logarithm to the base e, of the number of microstates that we identify with that
macrostate. Of course the microstates are constantly changing on the time scale
of atomic motions, yet we observe only the one constant macrostate when we
observe a system in thermodynamic equilibrium. Virtually every macrostate we
encounter is some kind of time average over many, many microstates through
which our observed system passes during our observation. The oxygen and
nitrogen molecules in a room move constantly and rapidly—but randomly—
while we sit in what we consider still air at a constant temperature and pressure.
Logarithms are pure numbers, not units of any property. However, strictly for
practical use, we express the entropy in physical units; usually in units of energy
per degree of temperature, the same units that express heat capacity. These may
be either energy per mole per degree, in which case we multiply the natural
logarithm by R, or energy per molecule (or atom, if the system is composed of
independent atoms), in which case we use the much smaller constant factor, the
Boltzmann constant, symbolized k, the gas constant divided by Avogadro’s
number, R/N. It is in these units that the entropy was introduced implicitly in the
statement of the second law of thermodynamics.
Now we can express the second law in other ways. One is this: All systems in
nature evolve by going from states of lower entropy to states of higher entropy—
that is, from macrostates of fewer microstates to macrostates of more possible
microstates—unless some work is done in the process of changing the state.
Another way to say this is that systems evolve naturally from states of lower
probability to states of higher probability. A gas expands when its volume
increases because there are more ways the molecules of that gas can be in
macroscopic equilibrium in the larger volume than in the smaller. To go to a state
with fewer microstates, we must do work; we must put in energy.
Another statement of the second law, less general than the above, is this:
Systems in contact that are initially at different temperatures but are allowed to
exchange energy as heat will eventually come to some common temperature,
determined only by the initial temperatures and heat capacities of each of the
systems. The composite system relaxes to a common equilibrium by going to the
state of highest entropy available to it. That is the state in which the systems
share a common temperature. All states in which the systems have different
temperatures are at lower entropy.
Here we have crossed a bit from the traditionally macroscopic approach to
thermodynamics to the statistically based microscopic approach. While
thermodynamics developed originally as a macroscopic view of heat, work, and
energy, the introduction of an approach based on statistical mechanics, the
statistical analysis of the behavior of the mechanics of a complex system,
provided a whole new level of insight and a perspective on the relationship
between the traditional approach and our new understanding of the atomistic
nature of matter. Statistical mechanics is our first bridge between the macro-level
and the micro-level for describing natural phenomena. And it is in the second
law that the statistical, microscopic approach gives us the greatest new insight.
The world simply moves from the less probable state where it is now to the next,
more probable state, the new state which has more available microstates, more
ways of being—that is, has a higher entropy—than the present state which it is
now leaving. If we want some subsystem, some machine for example, to do
otherwise, we must provide energy and do work to make that happen.
Historically, the way entropy was first introduced was, in fact, based on the
amount of heat exchanged, per degree of temperature. Specifically, the entropy
change is expressed as a limit, based on the ideal, reversible process. If there are
just two components, one at a higher temperature T1 and the other at a lower
temperature T2, then the heat lost from the high-temperature component, Q1, and
that gained by the low-temperature component, Q2, satisfy the relation that
Q1/T1 + Q2/T2 = 0 in that ideal limit. Obviously in a natural flow of heat, Q1
must be negative and Q2 positive. The ratio of heat exchanged to the temperature
at which that heat is gained or lost is the lower bound on the entropy change in
this purely thermal process. If the process is real and not an ideal, reversible one,
then the total entropy change must be greater than the sum of the ratios of heat
exchanged per degree and the sum of heats per degree of temperature must be
greater than zero. This macroscopic approach does not offer a way to evaluate
the actual entropy change; it only sets a lower bound to its value. Another way to
look at that relation is to rearrange the equation to the form Q1/Q2 = -T1/T2. That
is, the ratio of heat changes is given by the ratio of the initial temperatures; since
T1 is the higher temperature, the heat lost from that component is greater than
the heat gained by the colder component, when the two components come to
equilibrium at a common temperature between T1 and T2.
In the following chapter we shall investigate the nature of entropy in more
detail, as we explore how the concept developed, first at the macroscopic level,
and then in terms of behavior at the microscopic level. The essence of the latter,
which perhaps gives us the deepest and fullest understanding of the concept, is
the intuitively plausible idea that if a system can evolve spontaneously from a
macrostate with a smaller number of microstates to one with a larger number of
microstates, it will do just that. We can think of that statement of spontaneous
behavior, obvious as it seems when expressed in those terms, as one statement of
the second law of thermodynamics.
Figure 1. Walter Nernst. (Courtesy of the Smithsonian Institution Libraries)
The Newcomen engine operated in a simple way (Fig. 2). A wood fire supplied
heat to convert water to steam in a large cylindrical boiler equipped with a piston
or in an external boiler from which steam flowed into the cylinder with its
moveable piston. The steam pushed the piston, which, in turn, moved the arm of
a pump, which would lift water, typically out of a flooded mine. The hot cylinder
would be cooled with water, enough to condense the steam, and the piston would
return to its position near the bottom of the cylinder. Then the whole cycle would
start again, with heating the cylinder, and repeat itself as long as its work was
needed. Each stroke of the piston required heating and then cooling the entire
cylinder. The first installed Newcomen engine drove pumps that drew water
from coal mines on the estate of Lord Dudley, and soon these engines became
the most widely used means to pump water from mines in England, Scotland,
Sweden, and central Europe. Occasionally, others tried to introduce different
kinds of heat engines, but at the time none could compete with Newcomen’s.
One of his, called “Fairbottom Bobs,” is preserved at the Henry Ford Museum in
Dearborn, Michigan (Fig. 3). A functioning model of the Newcomen engine still
operates at the Black Country Museum in Birmingham, England.
Figure 2. Schematic diagram of a Newcomen engine. Steam generated in the cylinder pushes the piston up,
which in turn moves the lever arm at the top. Then water cools the steam and cylinder so the piston moves
back down, pulling the lever arm down. (Courtesy of Wikimedia)
Figure 3. The Newcomen engine “Fairbottom Bobs,” a real, working example of this early steam engine.
(Courtesy of Wikimedia)
Here we introduce a brief aside, on what constitutes a phase. There is one and
only one gas phase. Likewise, there is only one ordinary liquid phase for most
substances. Each structural form of a solid, however, is a phase of its own. Many
solids change their form if they are subjected to high pressures. We identify each
of those forms as a unique phase. Furthermore, some substances, especially
those whose elements have shapes far from spherical, can exhibit both the
ordinary liquid form with those elements oriented and distributed randomly (but
dense, of course) and at least one other form, under suitable conditions, in which
those nonspherical elements maintain some ordered orientation. We identify two
or more different liquid phases for such substances. Hence phase is a
characterization of the kind of spatial relation of the component elements of a
system.
The next major advance in steam engines came half a century after
Newcomen, and was a major breakthrough. James Watt (1736–1813) was a
young instrument maker and scientist at Glasgow University, where Joseph
Black was a professor. Watt speculated at least as early as 1760 about creating a
better engine, but his breakthrough came in 1763, when he was repairing a
miniature model of a Newcomen engine. He recognized that the model’s
cylinder cooled much more than the cylinders of full-sized counterparts, when
cooling water condensed the steam. This led him to investigate just how much
steam was actually generated on each cycle, compared with the amount of steam
needed just to fill the volume of the expanded cylinder. Watt found that the
actual amount of heat used, and hence of steam generated, for each cycle was
several times what would generate just enough steam to fill the cylinder. From
this, he inferred, correctly, that the excess steam was simply reheating the
cylinder. He then realized that trying to improve the performance of the
Newcomen engine posed a dilemma: How could one simultaneously keep the
cylinder hot, so that it would not require all that reheating on each cycle, and still
achieve an effective vacuum by condensing the steam to liquid water cold
enough to have a low vapor pressure? At that point, his genius led him to the
idea of having a second cylinder, or chamber of some sort, that would be cool
enough to condense the steam, but that would only become accessible to the
steam after the piston had been pushed to maximize the volume of the cylinder.
That second chamber can be merely a small volume that can be opened to allow
steam to enter from the main cylinder, in which that steam can be cooled and
condensed to the much smaller volume of liquid water. This was the genesis of
the external condenser, which essentially revolutionized steam engines and
stimulated a whole new level of interest in improving their performance. The
small external condenser could be kept cool, while the main cylinder could
remain hot. The key to its use is simply the valve that stays closed except when
it’s time to condense the steam and reduce the volume of the main chamber.
(And, of course, the condenser has a drain to remove the water that condenses
there.)
Watt, who was deeply concerned with making steam engines as economically
efficient as possible, also recognized that the pressure of the steam on the piston
did not have to remain at its initial high level. If the pressure was high enough
initially, it would be possible to shut off the supply of steam and the steam in the
cylinder would push the piston to the further limit of its stroke, with the pressure
dropping all the way but still remaining high enough to keep the piston moving
and expanding the chamber. At that extreme point of expansion, Watt realized
that the still-hot steam would be at a pressure still barely above atmospheric, so
that a valve could open to allow the steam that caused the excess pressure to
condense and then flow back to the boiler. Only then would another valve open,
this one to the external condenser, into which the remaining steam would flow,
emptying the cylinder so the piston would return to its original position. Watt
recognized that by letting the steam expand and cool almost to the surrounding
temperature, one could obtain the maximum work that the steam could provide.
It would be a way to obtain as much of the heat energy in the steam as
practically possible (Fig. 6).
This combination of simple steps emerged from Watt’s remarkable insight
into how the real Newcomen engine operated, and, more important, what kinds
of changes would make a steam engine much more efficient than the original
model. Four related ideas—keeping the cylinder hot, using an amount of steam
as close as possible to the minimum required to drive the piston through its
cycle, letting it cool as it expanded, and then condensing the remaining steam
externally, outside the cylinder—enabled Watt to make an altogether far more
efficient and more economical engine than its predecessors.
Figure 6. The steam engine of James Watt. Instead of cooling the entire large cylinder on each cycle, only
the small chamber, C, receives cooling water, but C opens to the main cylinder B at the end of the
expansion of chamber B, so the steam in B condenses in C, B remains hot and new water can be introduced
into B and converted into steam, driving the piston P. (Courtesy of Wikimedia)
Another very important contribution emerged from the firm of Boulton and
Watt, which made and oversaw the operations of those efficient steam engines.
This was the graphical means to measure the amount of work done by the engine
on each cycle, which became known as the “indicator diagram.” This diagram,
still constructed in its original and traditional form, represents each branch of the
cycle as a curve in a graph whose axes are pressure (on the vertical axis) and
volume (on the horizontal axis). John Southern, working with Boulton and Watt,
invented a simple device that literally had the steam engine draw its own
indicator diagram. The device consisted of a rolling table linked to the shaft
holding the piston, so that the table moved back and forth as the piston moved.
This motion revealed the changes in volume of the cylinder. The table held a
sheet of paper. Connected to the cylinder was a pressure-sensitive arm that held a
pencil, which moved across the paper, perpendicular to the back-and-forth
motion of the table, with its tip on the paper, so that the pencil inscribed a curve
on the paper as it and the table moved. Thus, with each cycle, the table moved
and the pencil drew a closed curve of pressure as a function of piston position—
that is, of the volume of the chamber. We shall address this in the next chapter,
where we see that the area enclosed by the closed curve measures the work done
by the engine on each cycle. Boulton and Watt incorporated these devices into
their engines and thus were able to assess the performance of each one. The
devices were held in locked boxes that only the employees of Boulton and Watt
could open, keeping the nature of these devices secret, until someone in
Australia opened one and revealed what they were and what they measured.
While Watt did make remarkable advances in the performance of steam
engines, an even better steam engine than his appeared and actually became the
dominant form of engine in the Cornish mines and elsewhere. Advances by
Richard Trevithick, Jonathan Hornblower, and Arthur Woolf specifically used
steam at as high a temperature and pressure as possible to start pushing the
piston and then continued the expansion until the steam had nearly cooled.
Consistent with the later findings of Sadi Carnot, these engines proved to be
even more efficient than those made by Boulton and Watt and were widely
adopted.
What Is Heat?
During the years when the steam engine was being developed, the question arose
regularly of just what heat is. Some of the properties of heat emerged from
observations and measurements. The amount of heat required to convert a given
amount of water to steam, for example, was the same as the amount one could
recover by converting that amount of steam back to water at its original
temperature. Melting a given sample required exactly as much heat as would be
delivered when that sample again turned solid. The concept that heat is
conserved emerged from such observations (but, as we shall see, that concept
was inadequate and hence incorrect). Furthermore, people began to recognize
that heat could be transferred in at least three different ways. It can go from one
place to another much like light, via radiation, for example warming by the sun.
It can also be transferred by the flow of a warm substance such as hot air or hot
water, which we call “convection.” Or it may be transferred by conduction, as it
passes through a material that itself does not appear to change or move, as it
does through metals, for example, but not through straw or wood. Yet these
properties were not sufficient to tell what heat is. They only tell us something
about what it does. The essential nature of heat remained a mystery,
notwithstanding the manifold ways it was used for practical purposes.
The recognition that motion, and what we now call kinetic energy, can be
converted to heat was the crucial observation from the famous cannon-boring
experiment of 1798 by Count Rumford, born Benjamin Thompson in America
(Fig. 7). He showed that the friction from boring out the metal from a cylinder
could generate enough heat to boil the water in which the borer and cylinder
were immersed. His finding laid the groundwork for the subsequent realization
that mechanical work can be converted to heat and then, with heat engines, that
the conversion can go in the opposite direction. Thus people came to realize that
heat itself need not be conserved—but something more fundamental and general
is indeed conserved, namely energy, as expressed in the first law of
thermodynamics.
Two concepts, mutually exclusive and thus competitive, evolved regarding
the nature of heat. One identified heat as the degree of motion of the (still
conjectured) particles that composed materials, including solids, liquids, and
gases. The other identified heat as a material substance, often as a mutually
repelling matter that surrounds every material particle. The former, the kinetic
model, actually emerged first, but the material or “caloric” concept overtook the
kinetic model, especially in the late eighteenth century. Two eminent and
distinguished French scientists, Antoine Lavoisier and Pierre-Simon Laplace,
published Mémoir sur la Chaleur in 1783, in which they presented both theories,
without suggesting that one was superior to the other, but observing that both
were consistent with the conservation of heat and with the different modes by
which it can move. It was Lavoisier who introduced the term “caloric” in 1789,
suggesting that perhaps he personally favored the fluid model. However, based
on his friction experiments, Rumford himself rejected the idea that heat was a
fluid, a concept then considered plausible by many people, obviously including
Lavoisier. (Lavoisier was executed during the French Revolution; his widow, a
brilliant woman herself, later became the wife of Rumford.)
Figure 7. Benjamin Thompson, Count Rumford; engraving by T. Müller. (Courtesy of the Smithsonian
Institution Libraries)
According to the concepts of that time growing from the work of Joseph
Black, the content of heat in any body of material is a simple function of the heat
capacity of the material, or of the specific heat and the amount of material, since
“specific” here refers to the heat capacity of a specific amount of the material,
such as one gram. There was a belief for a time that the specific heat of any
given substance is constant, but that was proven wrong. Heat capacities and
specific heats definitely depend on conditions of temperature and pressure, for
example. But every substance has its own specific heat, which we can think of as
a property we can represent by a curve in a graph with temperature as the
horizontal axis and the value of the specific heat as the quantity on the vertical
axis. But we would need one such curve for every pressure, so we should think
of the specific heat as represented by a surface in a space in which one axis in
the bottom plane is temperature and the other is pressure, and the vertical axis is
the specific heat. But constant or variable as the heat capacity might be, it was a
property that could be interpreted in terms of either of the concepts of what heat
is, motion or caloric fluid. Heat capacity could be interpreted in terms of the
intensity of atomic motion or of the amount of a physical fluid. This property,
important as it is for knowing how real materials behave and for explaining the
conservation of heat, gives us no new insight into the nature of heat. It is simply
useful as a property characterizing each substance, in terms of how it contains
heat and responds to its input or outflow. Nonetheless, it will turn out to provide
insight into how complex the particles are that make up each substance.
An important property of heat capacities, particularly of gases, is the way
they depend on the conditions under which heat is exchanged. Specifically, we
can add heat to a given volume of gas, holding the volume of the gas constant,
or, alternatively, we can hold the pressure of the gas constant. If we hold the
pressure constant, then the gas expands to a larger volume; if we hold the
volume constant, the pressure exerted by the gas increases, as Amontons
showed. We now understand why this difference exists. If the pressure is held
constant, the gas, by expanding, moves the walls of its container, doing work in
that process as it increases its temperature. If the volume is kept constant, the
system increases its temperature without doing any work. We shall explore this
more quantitatively shortly, but at this stage, we merely need to recognize that
the extra heat required to raise the gas temperature at constant pressure is
responsible for doing the work of expansion, an amount not needed for constant-
volume heating. As a consequence, we distinguish heat capacities at constant
volume from those at constant pressure. (These are even slightly different for
solids and liquids, because they do change their volumes and densities a bit with
changes in temperature—but much less than gases do.)
Another related aspect of the behavior of gases is the way they behave when
they expand under different conditions. Recognizing that there are such
differences was the initial step toward the discovery of the ideal gas law, the
simplest of all equations of state. If a gas expands against a force, such as the
pressure of the atmosphere or a piston linked to a driveshaft, and there is no heat
source to maintain its temperature, that gas cools as it expands. It loses heat by
doing work against that resisting force. However if a gas expands into a vacuum,
its temperature remains essentially unchanged. John Dalton (1766–1844, Fig. 8),
who is best remembered for his exposition of the atomic theory of matter,
showed that all gases expand equally when they undergo the same increase in
temperature. This finding, ironically, reinforced the belief by Dalton and many
contemporaries that heat is a pervasive fluid, known as “caloric,” which
surrounds every atom, and that it is the increase of caloric with temperature that
is responsible for the uniform expansion. This property of expansion was made
more precise by Dalton’s French contemporary Joseph-Louis Gay-Lussac
(1778–1850, Fig. 9), who studied and compared the expansion of oxygen,
nitrogen, hydrogen, carbon dioxide, and air. He showed that when the
temperature of a specific volume of each of these gases increases from what we
now call 0° Celsius, the freezing point of water, to 100° Celsius, the boiling
point, they all undergo the same increase in volume. This discovery led to the
relation that the change in pressure p is directly proportional to the change in
temperature, or Δp = constant × ΔT. This relation is now known in many parts of
the world as Gay-Lussac’s law, but because of a historical anomaly it is called
Charles’s law in the English-speaking world. Here we have the first precise
relation that will be incorporated into the ideal gas law.
Figure 8. John Dalton. (Courtesy of the Library of Congress)
Figure 10. Sadi Carnot as a young man, about eighteen years old. (Courtesy of Wikimedia)
This remarkably simple expression tells us that the best possible conversion of
heat into work, using the cycle devised by Carnot, depends only on the two
temperatures between which the system operates. Moreover if we think of the
low temperature as something close to room temperature and the high
temperature as the one we can control, this expression tells us to make the high
temperature TH as high as possible if we want to make the efficiency as high as
possible. This was, of course, the basis for Trevithick’s efficient steam engines.
One important characteristic to recognize about Carnot’s ideal engine is that
in order to have no heat losses to friction, the engine must operate very, very
slowly—infinitely slowly, to be precise. The engine must operate so slowly that
at any instant, it is impossible to tell whether it is moving forward or backward.
We call such an engine reversible. Yes, it is unrealistic to think of using an ideal,
reversible engine to carry out any real process, yet the concept serves as a limit,
however unattainable, to real processes. It enabled Carnot to find that ultimate
limit of efficiency for his engine. But it was more: introducing the idea of a
reversible engine as the hypothetical limit of real engines was one of the novel
concepts that Carnot introduced to revolutionize the way we use
thermodynamics.
Then Carnot took one more major step, often overlooked in discussions of his
contributions. He began with the realization that it is impossible to create a
perpetual motion machine. With this as a basic premise, he showed that all ideal
machines must have that same limit of efficiency. The reasoning is remarkably
straightforward. Imagine an ideal engine of any kind operating between the same
two temperatures as the Carnot engine, but suppose it has an efficiency greater
than that of the Carnot engine. Then it can run forward, driving the Carnot
engine backward, so that it delivers heat QH to the high-temperature reservoir,
having extracted it from the cold reservoir. Then that heat can be used to drive
the more efficient engine, which in turn pushes the Carnot engine to raise
another QH to the hot reservoir. This means that the two engines could go on
forever, together making a perpetual motion machine, taking heat from the cold
reservoir at TL and raising it to the hot reservoir at TH. Carnot recognized that
this was completely inconsistent with all our experience and therefore must be
impossible. Hence the other ideal engine can be no more efficient than the
Carnot cycle. The same argument tells us that the other ideal engine cannot be
less efficient than Carnot’s, because if it were, we could drive it backward with
the Carnot engine and, in that way, raise heat from a low temperature to a high
temperature without limit. Hence all ideal heat engines must have the same
limiting efficiency, given by equation (1) above. This reasoning, with its
amazing generality, is the culmination of Carnot’s contribution and is effectively
the origin of thermodynamics as a precise science.
Despite its deep innovation and key role in shaping the subject, Carnot’s work
was virtually unnoticed for a decade. Its first real recognition came in 1834,
when Émile Clapeyron published a paper utilizing Carnot’s concepts,
particularly by representing Carnot’s hypothetical engine via a form of its
indicator diagram, the plot of the cycle of the engine as a function of the pressure
and volume for each of the four steps.
Some insight into the capacity to do work came with the identification by G.
W. Leibnitz (Fig. 13) of the product of mass times the square of velocity, mv2, as
what he and others called vis viva, his measure of force. But not long after, both
G. G. de Coriolis and J. V. Poncelet showed that it is mv2/2 that is the relevant
quantity. Meanwhile, others, Descartes and the “Cartesians” (in opposition to the
“Leibnitzians”), argued that mass times velocity, mv, should be the measure of
capability to do work and that, in effect, this quantity is conserved. At that time,
the first half of the nineteenth century, it seemed impossible that both could be
valid and useful concepts. Today, we recognize mv2/2 as the kinetic energy of an
object and mv as its momentum, both of them valid, useful characteristics of a
body’s motion, and have forgotten altogether that at one time people believed
only one or the other could be a valid “measure of motion.”
Figure 13. Gottfried Wilhelm Leibnitz. (Courtesy of the Herzog Anton Ulrich-Museum)
William Thomson, who became Lord Kelvin of Largs in 1892 (Fig. 15), and
his older brother James played very important roles in the next steps in
developing the science of energy. Working first in Glasgow, a center at least as
important for the field as Manchester, they focused first on the efficiency of
engines of various kinds. Both steam and heated air were the driving materials in
the heat engines of that time, which by midcentury had replaced water-powered
pumps. William devised a temperature scale, taking the freezing point and
boiling point of water as the fixed points, with regular steps or degrees between
them. His scale was essentially identical to the one developed by Anders Celsius
in 1742, a century earlier. Thomson introduced the term “thermodynamic” in a
crucial paper of 1849, in which he analyzed the work of Sadi Carnot—having
finally found and read Carnot’s nearly forgotten book. He connected that
analysis to the experiments by Clapeyron and by Victor Regnault (Fig. 16),
measuring the ratio of the work done to compress a gas to the heat released by
that compression. Specifically, these experiments were meant to determine a
quantity fundamental to quantification of Carnot’s analysis, the maximum
amount of work that could be done by a temperature change of 1°.
Figure 15. William Thomson, Baron Kelvin. (Courtesy of the Smithsonian Institution Libraries)
In the context of the steam engine, we first encountered the indicator diagram, a
graph showing pressure as a function of the volume for a typical steam engine.
This is a closed loop, exemplified by the indicator diagram for the idealized
Carnot engine that we saw in Figure 11. There are many kinds of real engines,
with many different cycle pathways; in reality, no operating engine actually
follows a Carnot cycle. Traditional automobile engines, burning gasoline and air
when their mixture is ignited by a spark, follow what is called an “Otto cycle.”
Diesel engines follow a different cycle called a “Diesel cycle.” These differ
simply in the particular pathway each follows. The important point for us here is
to recognize something that Boulton and Watt and their engineer and later
partner John Southern recognized, and that still remains an important diagnostic
application of thermodynamics: that the area within the closed figure of the
indicator diagram is the actual work done in a single cycle.
We see this because (a) the work done to produce a small change of volume
dV against a pressure p is just the product of these two, pdV; (b) if we follow
one curve of the indicator diagram as the arrows indicate, say the highest branch
of the loop, by adding all the small amounts of work that are the result of going
from the smallest to the largest volume on that branch of the loop—that is,
adding up all the small pdV values—we would obtain the total work done when
the system moves along that curve.
We call that process of adding all the tiny increments “integration,” and we
symbolize it as ∫ pdV, which we call the integral. Each increment, each product,
is just the area under the curve for that tiny interval dV. Hence the integral from
beginning to end of that top branch is the area under that branch! That area is
just the work done by the system as it expands along that top curve, as the
volume increases. We can go on to the next branch, and see what work is done
there. Then, when we come to the long, lower branch, a compression step, the
volume diminishes with each dV so work is done on the system. The process of
integration along each of the lower branches involves volume changes that
reduce the size of the chamber, so there, the dV values are negative, work is
done on the system, and hence the areas under the lower branches are negative.
Thus, when we add the integrals, the areas beneath all four branches, those from
the top two are positive, but those from the bottom two are negative, so the sum
of all four is just the difference between the two sets, the area within the closed
loop. This is precisely why finding the area of the loop of the indicator diagram
is a straightforward way of determining the work done on each cycle by an
engine, any cyclic engine.
We can construct the diagram for the ideal performance of any cyclic engine
we can devise. Then we can let the machine draw its own indicator diagram, as
Boulton and Watt did; we can compare it to the ideal diagram for the same
ranges of volume and pressure. The comparison of the areas of the two loops
immediately tells us how near the actual performance comes to that of the ideal
engine.
An example is the indicator diagram for an ideal Otto cycle (Fig. 19). This
cycle is actually more complex than the Carnot cycle, because it is a “four-
stroke” cycle. In the Carnot cycle, one complete cycle has the piston moving out
and back, in what we call two “strokes.” In contrast, in the Otto cycle, the piston
moves back and forth twice in each full cycle, doing useful work only in one
expansion step. That is the step represented by the highest curve in the diagram,
the step in which the gasoline burns in air, forcing the piston out as the gases in
the cylinder expand. At the end of that stroke, the gases, still at fairly high
pressure, escape through a now open valve, so the second stroke consists of a
reduction of pressure while the volume remains constant. Then the piston
returns, at constant pressure, to its position corresponding to the minimum
volume of the cylinder (to the left in the diagram), driving the rest of the burned
gases out. Next, the exhaust valve closes and the intake valve opens, fresh air
enters the cylinder and the piston moves out, still at constant pressure. These two
steps correspond to the back-and-forth horizontal line in the diagram. Then the
piston again moves back to compress the air in the cylinder, preparing for the
introduction of gasoline. That step is the one represented by the curve rising
from right to left at the lower side of the loop in the diagram. At the end of that
step, gasoline enters the cylinder, either by simple flow through a now open
valve or through a jet. Then the spark goes off, igniting the gasoline-air mixture.
The introduction of the gasoline and the firing of the spark correspond to the
vertical upward line in the diagram, the step that occurs so suddenly in the ideal
engine that the volume does not change. That brings the engine back to the
“power stroke” from which the automobile derives its motion.
Figure 19. Indicator diagram for the ideal Otto cycle that drives the gasoline-powered automobile. Unlike
the earlier diagram for the Carnot cycle, this diagram uses pressure and volume for the axes, rather than
temperature and volume. On each full cycle, the system moves twice between the lowest and highest
volume; in the upper loop, the cycle generates work equal to the area within that loop; the lower “back-and-
forth” line is merely emptying and refilling the cylinder. (Courtesy of Barbara Schoeberl, Animated Earth
LLC)
Figure 20. James Clerk Maxwell. (Courtesy of the Smithsonian Institution Libraries)
where the kinetic energy K.E.(v) is mv2/2 and e is the “natural number,”
approximately 2.718. (As we saw previously, the quantity in the exponent is
called a “natural logarithm”; were we to write the expression as a power of 10,
we would multiply that natural logarithm by the power to which 10 must be
raised to get the value of e, approximately 0.435.) But the number of ways the
speed v could be directed (in other words, the number of ways the velocity could
be aimed) is proportional to the area of a sphere with radius v, so the probability
of finding v must also be directly proportional to v2. Without deriving the full
expression, we can write the probability distribution so that the total probability,
summed over all directions and magnitudes of the speed, is exactly 1:
Figure 21. The shapes of the Maxwell-Boltzmann distribution of particle speeds at three temperatures. The
peak moves to higher temperature and the area spreads to higher speeds, but the area under the curves is the
same for all temperatures because the amount of matter is the same for all three of them. (Courtesy of
Barbara Schoeberl, Animated Earth LLC)
We can be more precise about the deviations from the most probable by using
the concept of standard deviation from the mean, especially for the most
common and widely encountered form of distribution called a “Gaussian
distribution,” a symmetric, bell-shaped curve. A band within that curve that is
one standard deviation wide from the most probable (and mean) value of that
distribution either above or below that most probable value includes about 34%
of the total population of the distribution, so about 68% of the total population
lies within less than one standard deviation above or below the most probable
value. Put another way, only about 1/3 of all the values of the distribution fall
more than one standard deviation away from the mean, half of that third in the
low tail and half in the high. And the deviation to which the above paragraph
refers is specifically labeled as the standard deviation.
Maxwell arrived at the expression for the distribution of speeds by using
plausible and intuitively acceptable ideas about the randomness of the particle
velocities and their collisions. He did not find it by any set of logical steps or
mathematical derivations. Carrying out that derivation mathematically was the
first major contribution to this subject made by Ludwig Boltzmann. In 1872,
Boltzmann started with the ideas that (a) the particles of a gas must be moving
randomly because their container does not move, (b) their speeds, more strictly
their velocities (which indicate both speed and direction), are randomly
distributed at every instant, and (c) the particles collide with one another and,
thereby, change speeds and directions with every collision, and the collisions
happen very often. Using these completely plausible assumptions, he was able to
derive, from a combination of physics, mathematics, and statistics, a distribution
of velocities (and hence of speeds) for a gas in a state of equilibrium at a fixed
temperature. And that distribution was exactly the one that Maxwell had
produced intuitively! That distribution, the one given in the equation above and
shown schematically in the accompanying diagram, is now known as the
Maxwell-Boltzmann distribution, and is one of the fundamental links between
the macroscopic approach of traditional thermodynamics and the microscopic
approach based on the mechanics describing the behavior of matter at the atomic
scale.
Boltzmann went further in linking thermodynamics to other aspects of
science. He explained the basis of the connection, found experimentally by his
mentor Josef Stefan, between the electromagnetic radiation emitted by a warm
body and the temperature of that body; this relation, that the total rate of
emission of electromagnetic energy is directly proportional to the fourth power
of the body’s temperature, is known as the Stefan-Boltzmann law. Moreover,
Boltzmann showed that radiation exerted a pressure, just as a gas does. As we
have seen, a warm object gives off radiation of very long wavelength, radio
waves; a still warmer object loses energy by emitting microwaves; we know that
very hot objects, such as flames and filaments in incandescent lamps, emit
visible light. Still hotter objects, such as our sun, emit visible, ultraviolet, and
even X-ray radiation.
One more very important contribution Boltzmann made to thermodynamics
was a further application of statistics. He showed that if any collection of atoms
with any arbitrary initial distribution of velocities is left in an isolated
environment to evolve naturally, it will reach the Maxwell-Boltzmann
equilibrium distribution. He also introduced a function he labeled “H” that
measured how far that evolution to equilibrium has gone; specifically, he proved
that function achieves its most negative value when the system comes to
equilibrium. After Boltzmann first introduced this concept, it was called the H-
function. The initial negative sign, however, was later reversed, so that the proof,
applied to the redefined statistical H-function now called entropy, shows that
function evolves to a maximum. And the proof was a microscopically based,
statistically based demonstration that a system’s entropy takes on a maximum
value when it is in equilibrium. The H-function was precisely the micro-based
equivalent and counterpart of the macro-based entropy, as developed by
Clausius, but with the opposite sign.
The next major contribution to the linking of thermodynamics and mechanics
came from J. Willard Gibbs. Largely unrecognized in the United States until
European scientists read his works and realized their import, Gibbs established
the field of statistical thermodynamics on a new, deeper foundation. His
approach to applying statistics was based on the then novel concept of
ensembles, hypothetical large collections of replicas of the system one is trying
to describe. For example, if one wants to understand the behavior of a system in
equilibrium at a constant temperature, presumably by its contact with a
thermostat, then one imagines a very large collection of such systems, all with
the same macroscopic constraints of constant temperature and presumably
volume and pressure, but each system with its own microscopic assignments of
molecular or atomic properties that would be consistent with those macroscopic
constraints. Thus the individual systems are all different at the microscopic level,
but all identical at the level of the macroscopic, thermodynamic variables. Then
the properties one infers for the real system are the averages of those properties
over all the systems of the ensemble. Constant-temperature ensembles, also
called “canonical ensembles” or “Gibbsian ensembles,” are not the only useful
ones; others, such as constant-energy ensembles, called “microcanonical
ensembles,” are used to describe the behavior of isolated systems, for example.
The energies of the systems in the constant-temperature ensemble are not all the
same; they are distributed in a manner analogous to the Maxwell-Boltzmann
distribution of velocities in a single constant-temperature system. Other kinds of
ensembles also come into play when one introduces other constraints, such as
constant volume or constant pressure.
A very specific product of Gibbs’s work was the famous phase rule, often
called the Gibbs phase rule, which we discussed in Chapter 2. We shall return to
this later, when we investigate the boundaries of applicability of
thermodynamics. It was Gibbs who introduced the term “statistical mechanics”
and the concepts of chemical potential (free energy per unit of mass) and of
phase space, the six-dimensional “space” of the three spatial position coordinates
and the three momentum coordinates.
We have already seen that there are some situations in which general properties
of macroscopic systems, well described by thermodynamics, lose their
applicability when we try to apply those ideas to small systems. One example we
have seen is the possible violation of the third law, for collections of atoms or
molecules so few in number that those particles might all be brought to their
lowest quantum state.
Another is the Gibbs phase rule; recall that the number of degrees of freedom
f depends on the number of components c or independent substances, and on the
number of phases p in equilibrium together: f = c - p + 2, so that water and ice,
two phases of a single substance, can have only one degree of freedom. Hence at
a pressure of 1 atmosphere, there is only one temperature, 0° C, at which these
two phases can be in equilibrium. The “phase diagram” for water illustrates the
regions of temperature and pressure in which solid, liquid, and gaseous water
can exist as stable forms (Fig. 22). Phases can coexist at the boundaries between
regions. Thus, if we change the pressure, the temperature at which water and ice
can coexist changes. This rule describes the behavior of coexisting phases of
macroscopic systems, and seems to be a universally true condition—for macro
systems.
Figure 22. The phase diagram for water. In the white region, ice, solid water, is stable. In the striped region,
liquid water is stable. In the dotted region, the stable form is water vapor. Ice and liquid water can coexist in
equilibrium under conditions along the boundary separating the white and striped regions. Liquid water and
vapor can coexist in equilibrium along the curve separating the striped and dotted regions. The three forms
or phases can coexist together under only one temperature-pressure pair, labeled as the “triple point.”
(Courtesy of Barbara Schoeberl, Animated Earth LLC)
If we look at the behavior of small clusters, however, say 10, 20, or 50 atoms,
we find, both in experiments and computer simulations, that solid and liquid
forms of these tiny beasts may coexist in observable amounts within observable
ranges of temperature and pressure! Their coexistence is not restricted to a single
temperature for any given pressure. This reveals that small clusters of atoms and
molecules simply do not obey the Gibbs phase rule. But if we look carefully at
the fundamental thermodynamic foundations of equilibrium, this apparent
anomaly becomes quite easy to understand, and is not actually a violation of
fundamental thermodynamics.
If two forms of a substance, A and B, are in equilibrium, as we have seen, the
ratio of their amounts [A] and [B] is the temperature-dependent equilibrium
constant Keq = [A]/[B] = exp(ΔG/kT) = exp(NΔμ/kT). We want to use that final
form because we’ll examine what this expression tells us both when N has a
value corresponding to a macroscopic system and when N corresponds to just a
few atoms or molecules. Suppose that the free energy difference between two
phases, per atom, Δμ/kT, is very tiny, say 10-10, so that kT, effectively the mean
thermal energy per particle, is 1010 times larger than that difference in free
energies per atom or per molecule. And suppose we are dealing with a small but
macroscopic system, only 1020 particles, less than a thousandth of a mole,
roughly a grain of sand. In this case, the equilibrium constant, the ratio of the
two forms of the substance, is either exp(1010) or exp(10-10), either an
enormously large or infinitesimally small number, depending on which of the
two phases is favored by having the lower free energy. This tells us that even just
this very tiny bit away from the conditions of exact equality of the free energies
and chemical potentials of the two forms, the minority species would be present
in such infinitesimal amounts that it would be totally undetectable.
It is precisely this character that makes phase changes of ordinary substances
sharp, effectively like sudden, abrupt changes with temperature at any fixed
pressure. In principle, phase transitions are smooth and continuous, but for all
practical purposes we see them as sharp changes, so that we can treat them as if
they were discontinuous—for macroscopic systems. This is why we say water
has a sharp, specific melting temperature and a sharp, specific boiling
temperature at any chosen pressure.
Now let us apply the same kind of analysis to small clusters. Suppose our
system consists of only 10 atoms. That is, suppose N = 10, not 1020. In that case,
Keq can easily be close to 1 over a range of temperatures and pressures. For
example, if Δμ/kT is ±0.1, meaning the free energy per atom of the favored form
is 10 times lower than that of the unfavored, then Keq = exp(±0.1 × 10) =
exp(±1) or ca. 2.3 or 1/2.3. This means that it would be easy to observe the less-
favored form in equilibrium with the more favored when the free energies per
atom of the two forms differ by a factor of 10—which is a clear violation of the
Gibbs phase rule. Hence this apparent anomaly is entirely consistent with
fundamental thermodynamics. We see that the Gibbs phase rule is a rule about
macroscopic matter, not about atomic-size systems. We can go further, if we
know about how large the fraction of a minority species would be detectable
experimentally. We can find how large the size a cluster of atoms would be that
would just allow us to detect, say, a liquid at a temperature below its formal
freezing point, for example 0° C for water, and also what the lower limit of
temperature would be of the band of coexisting liquid and solid. We learn, for
example, that our experiments could reveal solid and liquid clusters of argon
atoms coexisting over a few degrees Celsius, if the clusters consist of 75 or
fewer atoms. But if the clusters were as large as 100 atoms, their coexistence
range of temperature would be too narrow for us to detect anything but a sharp
transition with the tools we have today.
Bacon, Francis, 49
Baseball, 7
Basic science to applied science, and vice versa, 48–49, 143–144
Bérard, J. E., 69
Bernoulli, Johann, 140
Biot, J. B., 69
Black, Joseph, 53–55, 64
Black body, 106
Bohr, Niels, 125
Bohr model, 125–126
Boltzmann, Ludwig, 88, 96, 116, 120, 121–122, 153
Boltzmann constant, 39, 42, 96, 97
Bose, S. N., 126
Bose-Einstein condensate, 46, 127
Bosons, 126
Boulton and Watt steam engines, 59–60, 109
Boyle, Robert, 49
Brown, Robert, 153
Brownian motion, 125, 153, 159–160
Galilei, Galileo, 50
Gas constant, 38, 39, 42, 81, 96
Gas lamps, 104
Gasoline engine, 20, 109, 111–113
Gas phase, 37–38, 50, 55, 56, 65–66, 132
Gaussian distribution, 119
Gay-Lussac, Joseph-Louis, 50, 66, 68
Gay-Lussac’s law, 50, 67
Generation of electricity, 101
Generator, 101
Gibbs, J. Willard, 37, 88, 89, 114, 122, 123
Gibbs free energy, 97, 98
Gibbs phase rule, 37, 123, 131, 133, 134–135, 148, 149
Gluons, 154
Gravity, 9
Hachette, J. N. P., 69
Halley, Edmund, 51
Heat, 7, 9, 10, 16, 20, 23, 50, 55, 61–69
Heat capacity, 32–34, 35, 36, 53, 64, 74, 81
Heisenberg, Werner, 126
Helmholtz, Hermann von, 83, 86
H-function, 122
History of thermodynamics, 48, 49
Hornblower, Jonathan, 61
Hydrogen atoms, 27, 38, 96, 125
Laplace, Pierre-Simon, 63
Lasers, 48, 125, 144
Latent heat, 53, 55
Lavoisier, Antoine, 63–64
Laws of motion, 6
Leakage, 21
Leibnitz, G. W., 78, 79
Leibnitzians, 78
Light-emitting diode (LED), 104–105
Liquid phase, 37–38, 40, 55, 56, 132, 135, 148, 149, 150–151
Local thermal equilibrium, 137
Logarithm, 41, 95
Lubricating oil, 21
Lucretius, 24, 159
Macroscopic systems, 2, 36, 91, 92, 93, 94, 96, 130, 131, 132, 135
Macroscopic variables, 6, 7, 17, 36, 39, 147–148
Macrostates, 6, 37, 40, 41, 42, 45, 94–95, 96
Mass, 5
Maximum power, 141
Maxwell, James Clerk, 73, 115–116, 119, 120, 153
Maxwell-Boltzmann distribution, 116–119, 121–122, 123
Maxwell distribution, 116
Mayer, J. R., 80, 81, 82
Methane, 103, 104
Microcanonical ensembles, 123, 128, 129
Microscopic variables, 6–7, 16–17, 145, 147
Microstates, 6, 40–41, 42, 43, 45, 91, 93, 94–95, 96
Microwaves, 121
Mixtures, 5
Mole, 38, 92
Momentum, 6
Motive power of heat, 70
Nanometer, 106
Natural gas, 13, 103, 104
Negation, 155
Nernst, Walter, 45
Newcomen, Thomas, 51
Newtonian mechanics, 89, 90, 93, 114, 126, 155
Nonequilibrium systems, 141
Nonextensive thermodynamics, 28
Nuclear energy, 102
Perpetual motion, 75
Phase changes, 134
Phase rule, 37, 123, 131, 133, 134–135, 148, 149
Phases, 37
Phase space, 123
Phlogiston, 158–159
Photoelectric effect, 125
Piston, 10, 11, 20, 25, 51–52, 58–59, 60, 111–113, 138
Planck, Max, 107–108, 124
Poncelet, J. V., 78
Pontryagin, Lev, 140
Position, 7
Potential energy, 86
Power, 71
Power stroke, 113
Pressure, 5, 6, 8, 50
Probability, 147
Probability distribution, 118
Process variables, 12, 16
Variables, 6
Velocity, 6, 12
Very small systems, 130
Vibrational degrees of freedom, 27
Volume (V), 5, 9, 15, 18
Waste, 11
Watt, James, 56
Wind, 101
Woolf, Arthur, 61
Work (W), 9–11, 12, 16, 17, 18–19, 20, 25, 29, 65, 71, 73, 100, 110, 112