Fundamentals of Qnovo Adaptive

Charging in Lithium Ion Batteries

This document contains material protected by patents owned by Qnovo.

For a listing, please visit http://www.qnovo.com/patents
The purpose of this white paper is to provide a technically inclined reader
insight into the challenges faced by lithium-ion batteries in mobile and
consumer devices. The fundamentals of Qnovo’s adaptive charging
technology are introduced, and its benefits in enabling improved battery
features such as very fast charging without battery degradation are
discussed.

Introduction to key terms

A lithium-ion battery is a complex device frustration with loss of battery capacity,

that stores electrical energy in the form of but the culprit is usually rapid aging due
an electrochemical reaction. A number of to poor cycle life.
basic parameters and terms are often
utilized to define this energy storage “Charge rate” or C-rate is a measure of
process: charging speed. It is the amount of
electrical current applied to the battery
“Battery capacity” is a measure of the divided by the battery capacity. For
battery’s ability to hold electrical charge. example, a typical 5-Watt AC wall adapter
The more charge it can hold, the longer supplies approximately 1 A into the
the battery life and the longer a consumer battery. Assuming a 3 Ah cell, the C-rate
can use their mobile device. Battery is 1A/3Ah, or 0.33C. Faster charging
capacity is typically measured in Amp- equates to higher C-rates, typically 0.7C
hours (Ah) or alternatively may be or greater. The higher the current, the
expressed in its ability to store electrical faster it charges. As every consumer can
energy, measured as Watt-hours (Wh) attest to, batteries typically charge too
slowly. This is due to the fact that charge
“Cycle life”, also known as battery rates using conventional charging
lifespan, is a measure of how many times methods are intentionally kept low to
a battery can be charged and discharged. preserve cycle life.
It is the primary factor behind why
consumers notice after a few months of To increase battery capacity AND
use that their device can no longer last increase cycle life AND shorten the
through the day. As the battery is charged
charge times simultaneously
and discharged, it loses its ability to hold
charge. Battery capacity declines with without compromise, the charging
usage; this loss is referred to as capacity process must be optimized for
fade. When battery capacity fades to 80% maximum efficiency. This is
relative to its new capacity, it is precisely what Qnovo’s adaptive
considered depleted and should be charging technology and software
replaced. Consumers often articulate their
products do.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 2
 The basic structure and functionality of a lithium-ion
battery

In its simplest structure, a rechargeable lithium-ion battery consists of two

electrodes opposite each other with a gap that is filled with lithium ions
suspended in a liquid or gel electrolyte solution. The positive electrode
during charging is called the cathode, and the opposite electrode is called
the anode. The physical transfer of lithium ions from one electrode to the
other is the basis of storing electrical energy. The illustration in Figure 1
shows the basic physical structure of a lithium-ion battery.

Figure 1: Illustration of the basic structure and operation of a lithium-ion

battery. During charging, lithium ions move from the cathode to the anode,
and are physically intercalated within the material grains of the anode.
Graphite (carbon) is nearly universal for anodes. Cathode materials are usually
made of a complex metal oxide, with Lithium Cobalt Oxide (LiCoO2) commonly
used for mobile and consumer devices.

The electrode layers are mechanically separated with a porous polymeric

material called the “separator” – its role is to allow the ions to freely move
yet prevent the two electrodes from electrically shorting and creating a
safety hazard. Figure 2 shows an electron microscope cross-section of the
alternating layers of the battery: anode / separator / cathode.

Volumetric energy density is a key metric representing the evolution of

lithium-ion batteries since the first commercial introduction by Sony in 1991.
It is a measure of the stored energy in Watt-hours divided by the battery’s
physical volume in liters, designated as Wh/l. This fundamental property
drives the ability to increase stored energy without increasing the battery’s

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 3
 physical volume. Historically, it has increased at about 5% annually,
primarily the result of denser packing of the materials described in Figures
1 and 2. A state-of-the-art polymer lithium-ion battery delivers an energy
density near or greater than 600 Wh/l. A 3 Ah smartphone battery,
equivalent to 11.4 Wh, occupies a tiny volume of 19 cc, which equates to
slightly more volume than a tablespoon.

Figure 2: An electron microscope photograph of the alternating layers

of a battery. The layers are typically 10 to 50 micrometers each. The
granular nature of the layers is visible. The capacity of the battery is
determined by the available material amount of anode and cathode.
Battery manufacturers coat the electrode materials over large surface
areas to increase the battery capacity to usable levels. Courtesy:
TechInsights.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 4
 The role of charging in a lithium-ion battery

Charging is an integral part of the battery’s practical utility to end-users, and

also plays a significant role in the battery’s performance. In particular,
charging is a major culprit in the loss of capacity and cycle life.

Battery charging remains a relic of the early days of the lead-acid battery
from the 19th century. It uses a simple approach named CCCV, an acronym
for “constant-current, constant-voltage,” as illustrated in Figure 3. Charging
begins by applying a constant DC current to the battery, what is called the
“CC phase.” During this phase, the battery’s voltage at its terminals rises
until it hits a maximum allowed voltage – primarily for safety reasons. At
this point, the charging electronics switch to applying a constant voltage to
the battery’s terminals. The charging current gradually decays until the
battery is fully charged, at which point the charging current becomes
negligible.

Figure 3: CCCV charging consists of a sequence of two steps: (i)

constant current, during which the battery voltage rises, and (ii)
constant voltage, after the battery reaches a predetermined
maximum voltage.

CCCV charging does not take into consideration any of the degradation
mechanisms that are inherent to the battery’s chemistry – as a result, it
amplifies any minutely present damage within the battery and accelerates
capacity loss (Figure 4). It also lacks any means to compensate for varying
operating conditions such as temperature or age. In a simple analogy, it is
akin to unleashing a barrage of cars onto a highway with no consideration
for traffic management; the result is a traffic nightmare with an elevated risk
for accidents.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 5
 One such damage mechanism is lithium metal plating. The anode has a
limited acceptance rate of lithium ions during charging. Uncontrolled CCCV
charging, especially during fast charging, results in excessive accumulation
of lithium ions at the boundary of the anode. The result is the formation of
unsafe lithium metal: lithium ions will favor bonding together instead of
making their way inside the anode material.

Figure 4: A lithium-ion battery loses some of its capacity through

use. When it reaches 80% of its original capacity as a new battery, it
is deemed depleted and unusable. CCCV charging is responsible for
greatly accelerating this capacity loss. For a consumer device that is
charged daily, 500 cycles corresponds to only one and half year of
operation, typical of most modern smartphones.

There are several other degradation mechanisms that may take place during
a battery’s operation. In addition to lithium plating, a thin protective layer on
the surface of the anode, called the SEI layer, may grow uncontrollably thus
consuming precious lithium ions. The grains of the electrodes, especially the
anode, may also pulverize under the repeated mechanical stress of
intercalated lithium ions. The integrity of the electrolyte – the solution that
facilitates the physical transport of the ions – is also another factor to take
into account. By all measures, historical charging methodologies including
CCCV failed to consider these important degradation mechanisms. The result
has been a premature loss of battery capacity, worsening cycle life
performance and unnecessary compromises by the battery manufacturers.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 6
 And to make matters worse, the demand for higher energy density and
faster charging are accentuating these problems – as witnessed by the
constant battery complaints from users of mobile devices.

You can’t fix what you can’t measure

Qnovo’s adaptive charging technology is predicated on real-time in situ

diagnosis of the degradation mechanisms that occur within the battery as a
result of its operation and due to inherent manufacturing variations. Qnovo
leverages the existing large body of scientific knowledge related to battery
degradation and cleverly adapts it into compact computational algorithms
that operate in real-time on commonly available processors within modern
consumer devices.

One critical diagnostic measurement that Qnovo derives with each charging
cycle for every battery is an effective diffusion time of the lithium ions as
they make their journey from the cathode to the anode. This diffusion time
governs the effectiveness of the battery materials and design in the storage
reaction. Naturally, one can readily gather that this measurement depends
on several factors including the manufacturing process, temperature, battery
age and health.

Figure 5: The diffusion time of lithium ions is one of several key

diagnostic measures Qnovo’s algorithms perform in real- time during
charging. It depends on several design and operating factors,
including energy density. Higher values of diffusion time strongly
correlate with increased battery degradation.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 7
 Electrochemical Impedance Spectroscopy (EIS) is a diagnostic tool widely
used by chemists in battery laboratories around the world. It provides
valuable insight into a continuum of chemical processes within the battery
and has been a workhorse in battery development. EIS measurement
equipment is expensive and bulky with a test cycle typically taking several
hours, thus making it impractical for use in consumer devices. Qnovo’s
algorithms are able to reconstruct through appropriate signal processing
techniques sufficient portions of the EIS spectrum to gain equally valuable
insight into the chemical processes, but perform such functions in real-time
and in situ within a consumer device. Figure 6 depicts a reconstructed EIS
graph of a battery in a mobile smartphone using Qnovo’s algorithms.

Figure 6: Qnovo algorithms reconstruct in situ the EIS graph of a

mobile smartphone battery. The semi-circular curves display the real
and imaginary components of the battery impedance over a range of
frequencies, in this particular case, spanning from 10 Hz to 10 kHz,
representing a number of processes including ion kinetics and growth
of the SEI layer.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 8
 Such diagnostic measurements performed in situ provide real-time data into
the health of the battery and the chemical processes that govern the
battery’s behavior. Naturally, these data vary with operating conditions,
e.g., temperature and usage, as well as design and manufacturing origin.
We routinely observe stark differences and signatures that we uniquely
attribute to particular battery manufacturers and designs. Qnovo has
accumulated a large set of test data covering lithium-ion batteries from
major battery manufacturers including LG Chem, Sony Energy, Samsung
SDI, ATL and others. This set of diagnostic tools enables Qnovo’s algorithms
to distinguish unique characteristics of batteries and shift the burden of
correcting inherent battery problems from expensive manufacturing
processes to much less costly readily-available computational power.

The need for adaptive systems: Closing the feedback loop

In principle, an optimal charging profile (inherently different from CCCV)

may be obtained by solving the set of non-linear equations that describe the
chemical processes within the battery. However, a numerical solution is
highly impractical. First, it completely ignores the uniqueness of each
battery. There is sufficient variability in the manufacturing of batteries, their
material properties and operating conditions that it is essential to have an
adaptive system to monitor these differences and close the loop – in other
words, there is not a universal static mathematical solution. Secondly,
solving these equations is computationally intensive (taking hours on a high-
end computer) and cannot be practically implemented in a consumer device.
Finally, owing to the exponential nature of the chemical processes, small
deviations from the optimal charging point are rapidly amplified and become
detrimental. This places unique requirements on the nature of the adaptive
algorithms and corresponding feedback systems; and yet such systems need
also to perform with as little power consumption as possible.

Figure 7 illustrates Qnovo’s adaptive charging process, where the feedback

loop is closed by dynamically modulating the charging current. Traditional
charging methodologies including CCCV rely on applying static DC currents.
In contrast, Qnovo’s algorithms use a broader spectrum of frequencies both
to excite the battery response and to modify the charging current. In
particular, Qnovo’s most advanced Adaptive Pulse Modulation™ algorithms
use a sequence of short charging pulses separated by measurement periods,
and then adaptively modulate these charging pulses in relation to the
received diagnostic results. It is very important to distinguish this approach
from past “pulse charging” approaches. Pulsing is not the key characteristic
of Qnovo’s charging; it is rather the closed loop behavior and modulation of

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 9
 the charging signal that make our algorithms effective. Pulsing is only one
practical way to modulate the charging signal.

Figure 7: An illustration of the basic elements of Qnovo’s

algorithms. The diagnostics require sampling the current,
voltage and temperature of the battery. No modification of the
battery is required. A model of the battery performs the
requisite analysis of the diagnostics data; modulating the
amplitude and timing of the pulses in the charging signal closes
the feedback loop.

Benefits of using adaptive charging

The improvement in battery performance from adaptive charging can be

quite significant. Figure 8 shows an example of a typical polymer lithium-
ion battery used in a mobile smartphone. The cell is rated at a capacity of
2.5 Ah with a state-of-the-art energy density of 560 Wh/l manufactured by
one of the leading battery suppliers. To deliver the minimum necessary cycle
life of 500 cycles with traditional CCCV charging, the battery can only be
charged at a maximum charge rate of 0.5C – corresponding to a maximum
charging current of 1.25 A, and a total charge time of 3 to 4 hours.

The capacity fade curve in Figure 8 highlights the challenge of fast

charging. Increasing the charging current by 3X to 3.75 A (also known as

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 10
 1.5C) using conventional CCCV creates so much damage within the battery
that the cycle life degrades from the nominal 500 cycles at low charge rate
to a measly 100 cycles – in other words, this battery is rendered useless and
unsafe within months. Replacing the CCCV charging protocol with Qnovo’s
adaptive charging algorithms has a dramatic effect on the battery
performance. Cycle life climbs to 800 cycles, capacity loss is greatly reduced
and the battery charges 3X faster. Qnovo adaptive charging gives this
battery exceptional performance combining high energy density, long
battery life, long battery lifespan (cycle life), and naturally, very fast
charging – all at the same time. This is what we mean by no compromise
charging.

Figure 8: The battery’s capacity fade curve illustrating its lifetime

ability to charge and discharge is greatly impacted by the charge rate
and type of charging algorithms. CCCV charging at a fast charge rate of
1.5C results in such great damage to the cell that its cycle life is
reduced to 100 cycles, rendering it useless for consumers. In contrast,
charging the cell under otherwise identical conditions but using Qnovo’s
adaptive charging algorithms at 1.5C extends the battery cycle life to
800 cycles.

A serious adverse condition of fast charging that is seldom discussed in the

media is battery swelling. A polymer lithium-ion battery grows in thickness
– swells – with use and time. The extent of this growth depends on several
factors, one of which is speed of charge. Device manufacturers typically
allow for additional space to accommodate battery thickness growth. But in

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 11
 an age where consumers are favoring thin and sleek devices with embedded
batteries (not just smartphones but thin tablets and smart watches), battery
swelling is now playing an important and limiting factor in design. Now in
addition to capacity, cycle life and charge rate considerations, a design
engineer must take into account cell swelling. Figure 9 compares battery
swelling as a percentage of its initial thickness between conventional CCCV
and Qnovo’s adaptive charging.

Figure 9: A polymer lithium-ion battery swells with use. The swelling

is a function of cycles as well as speed of charging. Device
manufacturers allow for swelling in their designs, typically up to 10%
of the battery’s original thickness. Fast charging using CCCV results in
excessive swelling such that the battery exceeds its maximum
thickness very rapidly. Qnovo’s adaptive charging is very effective in
controlling battery swelling. The battery in this particular case is rated
at 2.4 Ah and was charged at a rate of 1C (equivalent to 2.4 A).

Safety first

Battery safety is of paramount importance. While every lithium-ion battery

comes with associated protection circuitry to prevent unsafe conditions such
as over-voltage and over-current, our philosophy is that this protection
should activate only as a last resort. Intelligent algorithms and software
should complement such protection circuitry, and should be active at all
times during the operation of the battery, closely monitor its health and
make the appropriate decisions to ensure that unsafe conditions are always
avoided.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 12
 Qnovo’s algorithms incorporate several features designed to raise the safety
level of lithium-ion batteries. For example, our predictive analytics tools can
project forward the state-of-health of the battery and enable an estimation
of the progression of the inherent degradation mechanisms. Should our
algorithms identify a cell that may be suspect, precautionary measures are
taken well in advance to avoid the possible occurrence of a catastrophic
failure.

Summary

Qnovo adaptive charging adopts an innovative and unique

approach to battery charging, utilizing software algorithms and
existing computational power in mobile devices to significantly
improve the performance of lithium-ion batteries. Qnovo
technology does not require any modifications to the battery, and
has been shown in extensive testing to work consistently across
the spectrum of Li-ion suppliers and chemistries. Adaptive
charging is complementary to ongoing battery chemistry and
manufacturing improvements. Qnovo enables mobile device
manufacturers to deliver a no-compromise mobile experience that
alleviates the everyday pain consumers feel from battery
performance limitations.

© 2015 Qnovo Incorporated. All rights reserved.

v2.4, 12/18/2015 13