Direct Electron Transfer (3rd Gen Biosensors)

Direct Electron Transfer (3rd Gen Biosensors)

Direct electron transfer (DET) is the idea behind "third-generation" glucose sensors: the enzyme hands the electrons it takes from glucose straight to the electrode, with no oxygen and no added chemical go-between [1]. As of October 2026, it is still mostly a lab technology. A 2026 review in the Journal of Diabetes Science and Technology calls fungal FAD-dependent glucose dehydrogenase, used with a mediator, the preferred enzyme for commercial blood glucose sensors, and it treats DET as a research priority [1].

Three generations, one question: where do the electrons go?

Every enzyme-based strip starts the same way. An enzyme oxidizes glucose, which means it pulls electrons off the glucose molecule. The three generations differ only in what happens to those electrons next [1].

• First generation: oxygen takes the electrons and becomes hydrogen peroxide. The device measures glucose by oxidizing that peroxide at the electrode [1].
• Second generation: a mediator, a small synthetic molecule, ferries the electrons from the enzyme to the electrode. The current from re-oxidizing the mediator tracks glucose [1].
• Third generation: the enzyme passes electrons to the electrode itself [1].

Strips built on fungal FAD-GDH belong to the second group. FAD-GDH is short for flavin adenine dinucleotide-dependent glucose dehydrogenase, an enzyme from molds, and the review calls it the standard for mediator-based disposable blood glucose sensors because it ignores oxygen and is selective for glucose [1]. Our article on electron mediator technologies covers the molecules those strips rely on.

Why oxygen and mediators are worth removing

Glucose oxidase (GOx), the classic strip enzyme, reacts with oxygen by nature. In a mediator-based design, oxygen and the mediator compete for the same electrons, and the review notes this can reduce accuracy, particularly at low glucose levels [1]. A DET sensor has no mediator for oxygen to compete with.

The second payoff is a lower working voltage. In a 2022 lab study, a miniaturized DET sensor ran at 0.15 volts, below the roughly 0.2 volts at which acetaminophen reacts at an electrode [2]. With 20 micrograms per milliliter of acetaminophen in the test solution, the signal stayed at 99.8% of its baseline. Lowering the oxygen level had a negligible effect too, which the authors contrasted with glucose oxidase sensors [2].

Why common strip enzymes can't do this

Here's what surprised us: the enzymes the review ties to blood glucose strips can't hand electrons to an electrode at all. Glucose oxidase keeps its FAD cofactor (the part that actually carries electrons) buried deep inside the protein, and the 2026 review states plainly that it is not a DET-type enzyme [1]. Fungal FAD-GDH doesn't show DET either [1]. A 2017 review says the same of the commercially available PQQ-dependent glucose dehydrogenase, another strip enzyme family: its cofactor is too deeply buried for direct electron transfer [3].

That hasn't stopped papers from reporting DET with glucose oxidase. The same 2017 review, in the Journal of the Royal Society Interface, warned that signals credited to DET can come from inactivated enzyme that has released its cofactor, and it proposed control experiments to tell real DET from that artifact [3]. So when a paper claims "direct electron transfer," it is fair to ask whether those controls were run.

The enzymes that can do it

Two enzyme families come with their own wiring:

• Bacterial FAD-GDH. The best-studied version, from the bacterium Burkholderia cepacia, has a dedicated electron-transfer subunit carrying three heme groups, iron-containing molecules that pass electrons along. Electrons travel from the FAD through an iron-sulfur cluster, then through the hemes to the electrode [1].
• Cellobiose dehydrogenase (CDH). This enzyme from wood-rotting fungi has its own heme domain that works as a built-in mediator. Its weakness is glucose itself: in its natural form, CDH works on glucose at less than 0.001% of the efficiency it shows with cellobiose, the sugar it evolved for [1].

Neither is a ready-made glucose sensor. Researchers have swapped single amino acids to sharpen glucose selectivity, and they have fused heme domains onto enzymes that lack them [1]. The review's central open goal is a DET-type glucose dehydrogenase that matches or beats glucose oxidase's efficiency [1]. For how GOx and GDH compare on strips sold now, see Enzyme Chemistry: GOx vs. GDH.

Has anyone built a DET test strip?

Yes, in a lab. In 2013, a team at Tokyo University of Agriculture and Technology made a disposable DET strip by placing an engineered bacterial FAD-GDH on a carbon screen-printed electrode [4]. The engineering was the point. The natural enzyme is highly active toward maltose, another sugar, and the strip built with it gave what the authors called "dangerously flawed results" when maltose was present. A version with two amino acid changes was virtually nonreactive to maltose, and its strip tracked glucose without being thrown off [4].

The 2026 review's summary table marks bacterial FAD-GDH as having had blood glucose applications "in the past," but it names no product and doesn't say whether those sensors used DET [1]. In the sources we read, we did not find a test strip on the US market sold as a DET design, so we don't name one here.

Where DET is heading: continuous monitors

Most recent DET work targets continuous glucose monitors (CGMs), not strips. The 2026 review states that current electrochemical CGM sensors are all based on glucose oxidase, and it calls bacterial FAD-GDH a premier candidate for third-generation CGMs [1].

Two 2022 lab studies show where that research stands. The miniaturized sensor described above, built on gold sputtered onto a thin flexible film, held a steady response through 11 days of continuous operation; ARKRAY, Inc. funded that work [2]. A second study used open circuit potential, a readout that measures voltage instead of current and keeps the enzyme's workload low. That sensor ran in buffer solution for more than 3 months at 25 °C without a significant drop in signal or sensitivity [5].

Neither sensor has been worn by a person yet in those reports. Both teams name the same next step: testing in vivo, inside the body [2][5].